Digital PDFs

EK-DMP11-TM

September 1981

210 pages

Original

8.4MB

Document:	DMP11 Synchronous Controller Technical Manual
Order Number:	EK-DMP11-TM
Revision:	000
Pages:	210
Original Filename:	http://bitsavers.org/pdf/dec/unibus/EK-DMP11-TM-001_DMP11_Synchronous_Controller_Technical_Manual_Sep81.pdf

OCR Text

EK-OMP11-TM-001

DMP11
Synchronous Controller
. Technical Manual

EK-DMP11-TM-001

DMP11
Synchronous Controller
Technical Manual

Prepared by Educational Services
of
Digital Equipment Corporation

1st Edition, September 1981

Copyright © 1981 by Digital Equipment Corporation
The material in this manual is for informational purposes and is
subject to change without notice.
Digital Equipment Corporation assumes no responsibility for any
errors which may appear in this manual.
Printed in U.S.A.

This document was set on DIGITAL's DECset-8000 computerized
typesetting system.

The following are trademarks of Digital Equipment Corporation, .
Maynard, Massachusetts:
DIGITAL
DEC
PDP
DECUS
UNIBUS

DECsystem-l0
DECSYSTEM-20
DIBOL
EDUSYSTEM
VAX
VMS

MASSBUS
OMNIBUS
OS/8
RSTS
RSX
lAS

CONTENTS

Page
PREFACE
CHAPTER 1

SYSTEM OVERVIEW

1.1
1.2
1.2.1
1.2.2
1.2.3
1.3
1.4
1.4.1
1.4.1.1
1.4.1.2
1.4.1.3
1.4.1.4
1.4.2
1.4.2.1
1.4.2.2
1.4.2.3
1.5
1.6
1.7

INTRODUCTION.................................................................................................
SYSTEM CONCEPT ............................................................................................
The M8207 Microprocessor............................................................................
The M8203 Line Unit.....................................................................................
DMPll Microcode .........................................................................................
DMPII-VAX/PDP-ll INTERFACE ...................................................................
SUMMARY OF DMPII COMMAND AND RESPONSE STRUCTURES.....
User Program Commands...............................................................................
Microprocessor Control Command...................................... ..................
Mode Definition Command....................................................................
Control Command..................................................................................
Buffer Address/Character Count Command.........................................
DMPll Responses..........................................................................................
Buffer Disposition Count Response........................................................
Control Response....................................................................................
Information Response.............................................................................
PROTOCOL SUPPORT........................................................................................
SYSTEM APPLICATIONS..................................................................................
APPLI CABLE DOCUMENTS........................................ .....................................

CHAPTER 2

INSTALLATION

2.1
2.2
2.3
2.4
2.4.1
2.4.1.1
2.4.1.2
2.4.2
2.5
2.5.1
2.5.2
2.5.3
2.6
2.6.1
2.6.2
2.7
2.7.1
2.7.2
2.7.3
2.7.4

SCOPE.................................................................................................................... 2-1
UNPACKING AND INSPECTION .................................................................... 2-1
INSTALLATION CONSIDERATIONS............................................................. 2-1
PREINSTALLATION CONSIDERATIONS..................................................... 2-3
System and Device Placement........................................................................ 2-3
System Placement........................................... ....................................... 2-3
Device Placement ................................................. '" ..... ........ ...... ... ......... 2-3
System Requirements.. ................................ ... .......................................... ...... 2-7
MICROPROCESSOR INSTALLATION ........................................................... 2-11
Backplane Considerations ............................................................................... 2-11
M8207-Y A Considerations ............................................................................. 2-11
M8207-Y A Insertion ....................................................................................... 2-12
M8203 LINE UNIT INSTALLATION ............................................................... 2-12
M8203 Considerations......................... ......... ..... .................. ........................... 2-18
M8203 Insertion.............................. ............................. .................................. 2-18
DMPII SYSTEM TESTING ................................................................................ 2-34
Functional Diagnostic Testing ........................................................................ 2-34
DECXll System Exerciser ............................................................................ 2-34
Final Cable Connections ................................................................................. 2-42
DMPl1 Link Testing ...................................................................................... 2-42

111

1-1
1-1
1-3
1-3
1-3
1-4
1-4
1-4
1-4
1-6
1-6
1-6
1-6
1-6
1-7
1-7
1-7
1-7
1-8

CONTENTS (Coot)

Page
CHAPTER 3

COMMAND AND RESPONSE STRUCTURES

3.1
3.2
3.3
3.3.1
3.3.1.1
3.3.2
3.3.3
3.3.3.1
3.3.4
3.3.4.1
3.4
3.4.1
3.4.1.1
3.4.2
3.4.2.1
3.4.2.2
3.4.2.3
3.4.2.4
3.4.3
3.4.3.1

INTRODUCTION................................................................................................. 3-1
COMMAND/RESPONSE STRUCTURE.......................................................... 3-1
USER PROGRAM COMMANDS....................................................................... 3-1
Microprocessor Control Command................................................................. 3-3
Microprocessor Control Command Field Descriptions .......................... 3-4
Mode Definition Command ............................................................................ 3-8
Control Command.......................................................................................... 3-8
Control Command Field Descriptions .................................................... 3-10
Buffer Address/Character Count Command ................................................. 3-18
Buffer Address/Character Count Command Field Descriptions ........... 3-19
DMPll RESPONSES ........................................................................................... 3-19
Buffer Disposition Response ........................................................................... 3-20
Buffer Disposition Response Field Descriptions .................................... 3-22
Control Response ............................................................................................ 3-22
System Events ......................................................................................... 3-25
Protocol Events ....................................................................................... 3-26
Network Errors ....................................................................................... 3-27
Procedural Errors ................................................................................... 3-28
Information Response ..................................................................................... 3-29
Information Response Field Descriptions .............................................. 3-30

CHAPTER 4

PROGRAMMING METHODS

4.1
4.2
4.2.1
4.2.2
4.2.3
4.3
4.3.1
4.3.2
4.3.2.1
4.3.2.2
4.3.3
4.4
4.4.1
4.4.2
4.4.3
4.4.4
4.5
4.5.1
4.5.2

INTRODUCTION................................................................................................. 4-1
CSR INTERFACE DISCIPLINE ........................................................................ 4-1
Issuing Commands.......................................................................................... 4-2
Reading Responses .................... ...................... .......... ...................... ............... 4-2
CSR Interface Interactions ............................................................................ 4-2
DMPll CONFIGURATION AND STARTUP ................................................... 4-3
Configuration Procedure ................................................................................ 4-3
Specifying User-Defined Parameters ............................................................. 4-4
Specifying TSS Parameters ................................................................... 4-4
Specifying GSS Parameters ................................................................... 4-7
Protocol Operation.......................................................................................... 4-9
CRITERIA FOR DETERMINING USER-DEFINED PARAMETERS .......... 4-9
Setting the Selection Interval Timer.............................................................. 4-10
Setting the Babbling Tributary Timer ............................................................ 4-12
Setting the Streaming Tributary Timer .......................................................... 4-12
Criteria for Determining a Value for XDT Preset. ......................................... 4-12
ERROR COUNTER ACCESS ............................................................................. 4-13
Reading the Counters................... ........................ ........................... ..... ..... ..... 4-13
Counter Skew .................................................................................................. 4-13
iv

CONTENTS (Coot)

Page
4.6
4.6.1
4.6.1.1
4.6.1.2
4.6.2
4.6.2.1
4.6.2.2
4.6.2.3
4.7
4.7.1
4.7.2
4.7.3
4.7.4
4.7.4.1
4.7.4.2
4.7.4.3

ERROR RECOVERY PROCEDURES ............................................................... 4-14
Recovery from Network Errors .... ....... ... ...... ....... .... ... ....................... .......... ... 4-14
Recovery from Threshold Errors............................................................ 4-14
Recovery from Babbling and Streaming Tributary Errors .................... 4-14
Recovery from Procedural Errors......... ................... .............. ...... ...... ...... ....... 4-14
Recovery from Non-Existent Memory Error ......................................... 4-15
Recovery from a Receive Buffer Too Small Error ................................. 4-15
Recovery from a Queue Overflow Error ................................................ 4-16
BOOTING A REMOTE STATION ..................................................................... 4-16
Steps Leading to a Remote Load Detect Boot ................................................ 4-17
Steps Leading to a Power On Boot ................................................................. 4-18
Steps Leading to an Invoke Primary MOP Boot ............................................ 4-18
Line Unit Switch Settings for the Boot Functions .......................................... 4-18
Switch Settings for the Power On Boot Function ................................... 4-18
Switch Settings for the Invoke Primary MOP Boot Function ................ 4-18
Switch Settings for the Remove Load Detect Boot Function ................. 4-18

CHAPTERS

ASPECTS OF DMPl1 MICROCODE OPERATION

5.1
5.2
5.2.1
5.2.2
5.2.3
5.2.3.1
5.2.3.2
5.2.3.3
5.2.3.4
5.3
5.3.1
5.3.1.1
5.3.1.2
5.3.1.3
5.4
5.4.1
5.4.1.1
5.4.1.2
5.4.1.3
5.4.2
5.4.3
5.4.3.1
5.4.3.2

INTRODUCTION................................................................................................. 5-1
DMPll DYNAMIC POLLING ALGORITHM.................................................. 5-1
Polling States ..... ........ ...... ..... ........................ ...... ..... ... ............. ........ ........... .... 5-1
Computation Urgency .................................................................................... 5-2
Criteria for Determining Polling Parameters ................................................. 5-4
Criteria for Determining DELTA T....................................................... 5-5
Criteria for Determining Values of Q and R.......................................... 5-5
Determining a Value for Poll Delay ....................................................... 5-5
Determining a Value for DEAD T ......................................................... 5-6
ERROR RECORDING......................................................................................... 5-8
Error Counter Structures... ....... .... ............... ................ ................................ ... 5-8
Data Link Error Counters ...................................................................... 5-10
Threshold Error Counters....................................................................... 5-12
Station Error Counters ........................................................................... 5-13
STRUCTURE OF DMPII INTERNAL DATA BASE ...................................... 5-16
Linked Lists.................................................................................................... 5-16
The Free Linked List .............................................................................. 5-16
The Response Linked List ...................................................................... 5-17
Buffer Linked Lists .............. ....... .......... .............. .............. .......... ........... 5-17
Slot Mapping Table ........................................................................................ 5-19
TSS and GSS Structures ................................................................................ 5-19
The Global Status Slot ........................................................................... 5-19
Tributary Status Slot .............................................................................. 5-20

CONTENTS (Cont)

Page
CHAPTER 6

SERVICE

6.1
6.2
6.3
6.3.1
6.3.2
6.3.2.1
6.3.2.2
6.3.2.3
6.3.3
6.4
6.4.1
6.4.2
6.4.2.1
6.4.2.2
6.4.2.3
6.4.2.4
6.4.2.5
6.4.3
6.5
6.6
6.6.1
6.6.2
6.6.3
6.6.4
6.6.5
6.6.6
6.6.6.1
6.6.6.2
6.6.7
6.7
6.7.1
6.7.2
6.7.3
6.7.4

SCOPE.................................................................................................................... 6-1
MAINTENANCE PHILOSOPHy....................................................................... 6-1
TROUBLESHOOTING METHODS FOR MULTIPOINT ............................... 6-1
Approach ........................................................................................................ 6-2
Error Counters ................................................................................................ 6-2
Data Link Counters ................................................................................ 6-2
Station Counters ..................................................................................... 6-2
Threshold Error Counters............... ........ .............. ..... ..... .............. .......... 6-13
Error Counter Analysis .................................. ........................ ........... .............. 6-14
MAINTENANCE FUNCTIONS/MAINTENANCE MODES ......................... 6-17
Microprocessor Control Command Register.................................................. 6-17
Maintenance Modes....... .......... ..... ...... ....... ........ ..... .... ....... .... ....... ............ ...... 6-18
Maintenance Mode...................................................................... ........... 6-18
System Test ................................... ....... ........ ............. ....... ...................... 6-18
Single Step Internal Maintenance Mode ................................................ 6-18
System Test Internal Maintenance Mode .............................................. 6-18
External Maintenance Modes ................................................................ 6-20
Maintenance (LED) Indicators .... ................ ..... .............. ...... ......................... 6-21
PREVENTIVE MAINTENANCE (PM) ............................................................. 6-21
CORRECTIVE MAINTENANCE ON A PDP-II PROCESSOR ..................... 6-22
CZDMP* /DZDMQ* ..................................................................................... 6-24
CZDMR */CZDMS* ..................................................................................... 6-24
CZDMT* ........................................................................................................ 6-24
DMPll Microdiagnostic Error Reporting ..................................................... 6-28
Data Communications Link Test CZCLM* (DCLT) .................................... 6-30
DEC/XII DMPll Modules .......................................................................... 6-31
DMD* .................................................................................................... 6-32
DME* ..................................................................................................... 6-33
Soft Error Reports Under DEC/XII ............................................................. 6-33
CORRECTIVE MAINTENANCE ON A VAX-II /780 ..................................... 6-34
EVDXA *. * COMM Microprocessor Repair Level Diagnostics ................... 6-34
EVDMA REV *.* M8203 Line Unit Repair Level Diagnostics .................... 6-34
EVDMB* VAX-II /780 DMPll Functional Diagnostic (Level 2) ................ 6-35
EVDMD* VAX DMPll DCLT Program ...................................................... 6-35

CONTENTS (Cont)

Page
APPENDIX A

DDCMP IN A NUTSHELL

A.l
A.1.1
A.1.2
A.l.3
A.1.4
A.l.5
A.2
A.3

DDCMP ..................................................................................................................
Controlling Data Transfers ....................................................................... ......
Error Checking and Recovery... ....... ..... ...... ..... ..... .... ... ....... ...... ......... ...... ......
Character Coding ...........................................................................................
Data Transparency .........................................................................................
Data Channel Utilization................................................................................
PROTOCOL DESCRIPTION ..............................................................................
MESSAGE FORMAT ...........................................................................................

APPENDIX B

FLOATING DEVICE ADDRESSES AND VECTORS

B.l
B.2
B.3

FLOATING DEVICE ADDRESSES ................................................................... B-1
FLOATING VECTOR ADDRESSES.................................................................. B-4
EXAMPLES OF DEVICE AND VECTOR ADDRESS ASSIGNMENT ......... B-4

APPENDIX C

LINE UNIT AND MODEM CONTROL REGISTER FORMATS

C.l
C.2

LINE UNIT AND MODEM CONTROL REGISTER FORMATS .................. C-l
RS-449 VERSUS RS-232-C................................................................................... C-6

APPENDIX D

MODEM CONTROL

D.l
D.1.1
D.1.1.1
D.1.1.2
D.1.2

MODEM CONTROL ............................................................................................ D-l
Line Unit Modem Control .............................................................................. D-l
Modem Ready Lockout of RTS ... ... .... ...... .... ......... ........ ...... ..... ...... .... ... D-l
Half-Duplex Mode ................................................................................. D-l
Modem Control Implemented by the DMPll Microcode ............................. D-l

APPENDIX E

INTERFACE DIAGNOSTIC COMMAND FLOWCHARTS

A-I
A-I
A-I
A-2
A-2
A-2
A-2
A-3

FIGURES
Figure No.
1-1
1-2
1-3
1-4
2-1
2-2
2-3

Title

Page

Simplified Block Diagram. ...... .... ..... ........... ......... ..... ...... .... ....... ....... ..... ........ .... .... 1-2
General Summary of DMPll Command Response Structure............................... 1-5
Block Diagram of a Multipoint Network with Single Assigned Addresses ............ 1-9
Block Diagram of a Multipoint Network with Multiple Tributary Addresses ....... 1-10
Example of a Topology Diagram for a Local Multipoint Network......................... 2-4
Example of a Topology Diagram for a Remote Multipoint Network ............ ......... 2-5
M8207-Y A Microprocessor Switch/ Jumper Locations.......................................... 2-8
vii

FIGURES (Cont)

Figure No.
2-4
2-5
2-6
2-7
2-8
2-9
2-10
2-11
2-12
2-13
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
3-10
4-1
4-2
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
6-1
6-2
6-3
6-4
6-5
6-6
A-I
A-2
B-1

Title

Page

M8203 Switch and Jumper Locations .................................................................... 2-20
Microprocessor/Line Unit Installation ................................................................... 2-25
DMPII Cable Drawings ......................................................................................... 2-26
DMPII Turnaround Test Connectors ..................................................................... 2-29
DMPII Remote System Cabling Diagram ............................................................. 2-35
DMPII /DMRII to DMPII Integral (Local) Modem
Cabling Diagram for Point-to-Point Network ......................................................... 2-37
Multipoint Cable Routing Example Half-Duplex ................................................... 2-38
Multipoint Cable Routing Example Full-Duplex Control Station End Node ......... 2-39
Multipoint Cable Routing Example Full-Duplex
Control Station Not an End Node .......................................................................... 2-40
Multipoint Cable Routing Example Full-Duplex
Tributary Station End Nodes .................................................................................. 2-41
Relationship Between DMPll CSR's Byte and Word Symbolic Addresses .......... 3-2
Command Response Fixed and Variable Formats .................................................. 3-3
Bit Layout of Initialization Register ....................................................................... 3-3
Flowchart for a Routine to Intialize DMP11.......................................................... 3-5
Mode Definition Command Format........................................................................ 3-9
Control Command Format. .......... ........ ................ ..... ..................................... ......... 3-9
Buffer Address/Character Count Command Format. ............................................ 3-18
Buffer Disposition Response Format ...................................................................... 3-20
Control Response Format ........................................................................................ 3-22
Information Response Format ................................................................................ 3-29
CSR Interface Control Bits .................................................................................... 4-2
CSR Access Window.............................................. ...................... ........ ............ ...... 4-3
State Diagram of Polling State Transitions ................................... \0........................ 5-3
Relationship Between Polling Parameters Q, R, and DELTA T ............................ 5-4
Relationship Between Polling Parameters Q, R,
and the Minimum Polling Interval. ..... ...... .......... ......................... ....... ...... ..... ......... 5-6
Relationship Between the Default Values for Q
and R for the Three Polling Levels .... ..... ....................... .................... ............ ......... 5-7
TSS Map of Error and Statistical Counters ........................................................... 5-9
Memory Map .................................................................................................. 5-10
Map of DMPll Data Memory ............................................................................... 5-17
DMPII Linked List Structure Format ................................................................... 5-18
Typical Isolation Troubleshooting Flowchart ..... ........................ ............................ 6-3
Data Memory ............................... ,..... ,. .......... ..... ..... .... .................................... ....... 6-8
Tributary Error Counters........................................................................................ 6-9
Station Error Counters ............................................................................................ 6-10
Full-Duplex Seven Tributary Multipoint Network ................................................. 6-16
M8203 Maintenance LED Locations ..................................................................... 6-23
DDCMP Data Message Format ............................................................................. A-I
DDCMP Message Format in Detail ....................................................................... A-2
UNIBUS Address Map .......................................................................................... B-1

ass

viii

TABLES
Table No.
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
2-10
2-11
2-12
2-13
2-14
2-15
2-16
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
4-1
4-2
4-3
4-4
6-1
6-2
6-3
6-4
6-5
6-6
6-7
6-8

Title

Page

DMRII Option Packing List.................................................................................. 2-2
Typical Control Station Options of a Bell 208A Compatible Data Set
(4800 b/s) Full-Duplex Operation .......................................................................... 2-6
Typical Tributary Options of a Bell 208A or a Compatible Data Set
(4800 b/s) Full-Duplex Operation .......................................................................... 2-6
DMRII Voltage Chart ........................................................................................... 2-7
M8207-YA DMRII Jumper Chart ........................................................................ 2-8
Switch Pack E 127 Selections .................................................................................. 2-9
Switch Pack E28 Selections .................................................................................... 2-10
Switch Selectable Features ..................................................................................... 2-11
DMPII Diagnostics ................................................................................................ 2-13
Normal M8203 Configuration ................................................................................ 2-14
M8203 Jumper Functions ....................................................................................... 2-15
Switch Pack E39 Selections .................................................................................... 2-17
Switch Pack E12l Selections .................................................................................. 2-19
Switch Pack E 134 Selections ....... ........................................................................... 2-21
Cable Description ................................................................................................... 2-22
Modem Option Jumper Functions .......................................................................... 2-33
Symbolic Byte and Word Address ..... .............. .............. ......................................... 3-2
User Program Command Codes.............................................................................. 3-4
DMPII Internal Diagnostic Execution Codes (Octal) ........................................... 3-4
Mode Field Codes and Functions............................. ........................................... .... 3-9
Correspondence Between Polling Levels and the
Binary Code in BSEL4, Bits 0 and 1....................................................................... 3-11
Octal Codes for Request Key Control Functions .................................................... 3-13
Buffer Disposition Response Type Codes.. .............................................. ............... 3-21
Event/Error Names, Categories, and Corresponding Codes (in Octal) ................. 3-23
DMPII Procedural Errors and Corresponding Codes (in Octal) ........................... 3-24
User-Defined TSS Parameters................................................................................ 4-5
User-Defined ass Parameters ............................................................................... 4-8
Recommended Selection Interval Timer Values .................................................... 4-11
Mode Switch Settings ............................................................................................. 4-19
BSEL 1 Bit Descriptions ......................................................................................... 6-19
Maintenance Indicators .......................................................................................... 6-22
CZDMP* Diagnostic Summary ............................................................................. 6-25
DZDMQ* Diagnostic Summary ............................................................................ 6-25
CZDMR * Diagnostic Summary ............................................................................. 6-26
CZDMS* Diagnostic Summary ............................................................................. 6-27
CZDMT* Diagnostic Summary ............................................................................. 6-29
Internal Diagnostic Results.......... .... .... ............................................ ....................... 6-30

PREFACE

This manual describes in detail the installation requirements, programming considerations, and servicing procedures, including diagnostic support, for the DMPII Synchronous Controller. Appendices
are also provided as an addition to the above.
Other publications which support the DMPII Synchronous Controller are:

•

M8207 Microprocessor Technical Manual (EK-M8207-TM)

•

M8203 Line Unit Technical Manual (EK-M8203-TM)

•

DMPll Print Set (MP-00795)

•

Electronic Industries Association (EIA) specifications

CHAPTER 1
SYSTEM OVERVIEW

1.1 INTRODUCTION
The multipoint DDCMP-DMPII intelligent communications line controller is a general purpose device
designed to manage efficient, high-speed, synchronous, multipoint networks using VAX or PDP-II systems as control and tributary station hosts. This device is designed to be managed by a VAX or PDP-II
system with minimum demand on system resources.
NOTE
This manual should be used with the M8207 Microprocessor Technical Manual (EK-M8207-TM) and
the M8203 Line Unit Technical Manual (EKM8203-TM).
1.2 SYSTEM CONCEPT
DMPII operation is based on a high-speed microprocessor that is optimized for multipoint networks. In
multipoint networks, a DMPII can operate as either a control or a tributary station. When operating as
a control station, a DMPII can control up to 32 other tributary stations. These tributary stations can
consist of additional DMPIls or devices compatible with Digital Data Communications Message Protocol (DDCMP).
A DMPII can be configured to operate with both half-duplex and full-duplex communications lines.
For full-duplex lines, line speeds can range up to 500K bits per second (b/s). For half-duplex lines, line
speeds can range up to 1M b/s.
Operation of each DMPII in a multipoint network is managed by an internal microprogram that includes a dynamic polling algorithm. This polling algorithm changes the polling frequency at each tributary in order to accommodate the actual resource needs of that tributary. This polling frequency
change is based upon the resporise history. Because system resources are allocated as the need arises,
overall network efficiency is increased.
DMPII communications line control and data transfer activity is directed by a user program through a
command/response structure.
NOTE
In this manual, the term user program is synonymous with device driver.
Using this command/response structure, a user program can manage data traffic between a control
station and selected tributaries. Also, with this command/response structure, the user program can control protocol operation, tributary polling priorities, transmit and receive buffer assignments, and network status and error recording and reporting.

1-1

The multipoint DDCMP-DMPll intelligent communications line controller is made up of two devices:
the M8207 microprocessor and the M8203 line unit. The communications functions performed by the
DMPll are carried out by a microprogram contained in ROM and executed by the M8207 microprocessor. Physically, a DMPll includes two standard-height, hex boards inserted in a standard
UNIBUS hex SPC slot. A simplified block diagram of the DMPll is shown in Figure 1-1.

VAX/PDP11
MAIN CPU

UNIBUS

I~~----~~~----------~

M8207

CONT.ROL AND STATUS
REGISTERS AND NPR
INTERFACE REGISTERS
(MULTIPORT RAM)

DATA
RAM

CJ)

CONTROL
ROM
(MICROCODE)

MICRO·
PROCESSOR

::::>

.....- - - - I. . ...J

<
Z

a:
w ~
z

INTERNAL
REGISTERS

LINE UNIT
INTERFACE

M8203

LINE

UNIT

_---,--.J r-'--__
1
REC
SILO

XMIT
SILO

XMIT LOGIC
PARALLEL -

SERIAL

REC LOGIC
SERIAL - PARALLEL

Ir--------J

1
EXTERNAL MODEM
INTERFACE

INTEGRAL MODEM

Figure 1-1

Simplified Block Diagram

1-2

1.2.1 The M8207 Microprocessor
The M8207 microprocessor is designed to support networks formed by VAX and PDP-II systems. It
operates in parallel with the associated system and has an architecture optimized for data movement,
character processing, performing address arithmetic, and other functions necessary to controlling I/O
devices, formatting data, and carrying out protocol operation. The DMPII resides as a non-processor
request (NPR) device on both the VAX and PDP-II UNIBUS.
Operation of the D MP 11 as an intelligent multipoint or point-to-point network controller is managed by
a ROM-based microprogram executed by the M8207 microprocessor. User-defined parameters controlling the dynamic aspects of DMPll operation are in the microprocessor data memory.
Data transfers by the microprocessor are performed under link control of DDCMP over synchronous
multipoint or point-to-point networks. In multipoint networks, tributary addresses can be switch-selected at installation time or software-assigned by the user program during actual operation. The microprocessor has an 8-bit internal data path, a 4K, 8-bit random access data memory, and a control ROM
having a maximum addressable storage capacity of 6K, I6-bit micro-instructions.

1.2.2 The M8203 Line Unit
Functionally, the M8203 line unit serves as an interface between the M8207 microprocessor and the
physical communications line. In this role, the line unit provides the user with a variety of operational
options. For example, line speeds from 56K to 1M b/s can be selected through the line unit integral
modem. Alternatively, the line unit can be configured to connect the microprocessor to the communications line through an external modem. Also, both the integral and external modem interfaces can
accommodate both full-duplex and half-duplex communications lines.
A major advantage of the M8203 line unit is that it connects directly to the communications line, bypassing the UNIBUS. As a consequence, communications activity handled through the DMPII has
little impact on UNIBUS bandwidth.
Data silos within the line unit provide for the buffering of transmit and receive messages. The buffering
capacity of these silos is optimized to eliminate, as much data overrun as is possible under normal operating conditions.

1.2.3 DMPII Microcode
The DMP11 microcode is a 6K, 16-bit, ROM-encoded microprogram executed by the M8207 microprocessor. This microprogram is supported by a 4K, 8-bit, data RAM that stores user-defined communications and protocol parameters as well as the internal data base. These parameters, along with the
internal data base, allow the DMPll to manage data traffic between stations and to respond dynamically to changing communications line activity.
In addition to handling all transmit and receive message processing under DDCMP, the DMPIl microcode manages buffer assignments and the storage and retrieval of messages in main memory, maintains
the internal data base, and performs a wide range of error/event surveillance, detection, reporting, and
recovery activities.
A major feature of the microcode involves the ability to down-line load a remote main CPU through the
DMP1! at an associated multipoint or point-to-point station, and to initiate execution of the loaded
program. As a result, the operational integrity of network stations at inaccessible locations can be readily maintained. The only restriction on this process is that the main CPU at the remote station must be
in the PDP-II family. However, the main CPU at the control station can be one of either the VAX or
PDP-II family.
The operational efficiency of DMPII-implemented multipoint ,networks is substantially increased by
the dynamic polling algorithm contained in the DMPll microcode. This algorithm adjusts the current

1-3

polling priority of a tributary according to the past polling response history of that tributary. The relative priority of each tributary at the three polling activity levels (active, inactive, and unresponsive) can
be established through user specified parameters. In addition, the user program can change the priority
of a tributary and its polling state during actual operation.

1.3 DMPII-VAX/PDP-ll INTERFACE
Communication between the user program and the DMPll is accomplished over the VAX/PDP-II
UNIBUS through four control and status registers (CSRs). These four 16-bit registers serve as a bidirectional port to pass user program commands to the DMPll and DMPll responses to the user program. Each of these registers is byte and word addressable by the user program.
In this group of four CSRs, the first two have a fixed format that serves as a preface to all user program
issued commands and DMPll resonses. The format of the second two registers is variable and determined by the specific command or response. Functionally, the four CSRs form a port for the exchange
of commands and responses between the user program and the DMPll. The actual content of the data
port is specified by an identification field in the fixed format portion. Other fields in the fixed format
portion provide for system initialization, interrupt enabling, reading and execution of maintenance instructions, line unit maintenance, data transfer set up and execution, and tributary addressing.
A user program issues a command to the DMPll by transferring the command to the CSRs. Specific
bits in the fixed format portion of the command alert the DMPll to the existence of the command and
its identity. The DMPll then performs the functions specified by the command. If a response is required, the DMPll stores the appropriate response in the CSRs and informs the user program that a
response is pending.
Message data received or transmitted by the DMPll is respectively written into or read from user program allocated buffers in main memory. These buffers are accessed by the DMPll through execution
of NPRs to the UNIBUS address of the buffer. A UNIBUS address is defined as an 18-bit address
used by an NPR device to access a device on the UNIBUS or a location in main memory.
1.4 SUMMARY OF DMP11 COMMAND AND RESPONSE STRUCTURES
A user program can issue four types of commands to be executed by the DMPl1. In general, the
DMPll replies to these commands with three types of responses. Figure 1-2 provides a general summary of the functions performed by the DMPll command/response structure. The functions performed
by DMPll commands and responses are described in detail in Chapter 3.
1.4.1 User Program Commands
The four types of commands issued by a user program permit the program to control all necessary aspects of DMPll operation. These are:
1.
2.
3.
4.

The microprocessor control command,
The mode definition command,
The control command, and
The buffer address/character count command.

The general functions performed by these commands are described in the sections that follow.
1.4.1.1 Microprocessor Control Command - This command has two functions, one operational and one
diagnostic. For the operational function, the microprocessor control command is used to clear all condition-sensitive logic in the DMPll. At startup time this command is the first issued by the user program
in order to place the DMPl1 in the run state in preparation for system initialization.
On receiving this command, the DMPll executes an internal diagnostic microprogram to verify its
operational state. When the diagnostic has executed, the user program is notified of success or failure
through a code in the CSRs. If successful, the DMPll is placed in an operational state.

1-4

VAX/PDP11
USER PROGRAM

USER PROGRAM ISSUES
MICROPROCESSOR CONTROL
COMMAND TO MASTER
CLEAR DMP11

INITIALIZE
DMP11

NOTIFY
USER
PROGRAM

FAILURE

USER
PROG. VERIFIES
INTERNAL
DIAG.

SUCCESSFUL,
DMP11 RUNNING
SET MODE TO HALF
OF FULL DUPLEX,
ESTABLISH STATION
AS POINT-TO-POINT
NODE OR MULTIPOINT
CONTROL OR TRIBUTARY
STATION

USER PROGRAM ISSUES
MODE DEFINITION
COMMAND

ESTABLISH TRIBUTARIES,
SET USER DEFINED
PARAMETERS, INITIATE
PROTOCOL STARTUP

USER PROGRAM ISSUES
CONTROL COMMANDS

ASSIGN TRANSMIT
AND
RECEIVE BUFFERS

USER PROGRAM ISSUES
BUFFER ADDRESS /
CHARACTER COUNT
COMMANDS
RECEIVE

TRANSMIT

THE DMP11 PERFORMS THE FOLLOWING
RECEIVE MESSAGE FUNCTIONS:
1. PROCESS RECEIVE HEADERS
2. CHECK CRC's
3. ACKNOWLEDGE (ACK) PROPERLY
RECEIVED MESSAGES
4. NEGATIVE ACKNOWLEDGE (NAK)
ERRONEOUSLY RECEIVED MESSAGES

THE DMP11 PERFORMS THE FOLLOWING
TRANSMIT MESSAGE FUNCTIONS:
1. CREATE DDCMP MESSAGE
HEADERS
2. GENERATE MESSAGE AND
HEADER CRCs
3. TRANSMIT MESSAGES

THE DMP11 PERFORMS THE FOLLOWING
MESSAGE TRAFFIC CONTROL FUNCTIONS:
1. MESSAGE SEQUENCING
2. ERROR RECORDING AND REPORTING
3. PROTOCOL SUPPORT
4. LINK MANAGEMENT

DMP11 ISSUES BUFFER
ADDRESS / CHARACTER
COUNT RESPONSES, CONTROL
RESPONSES AND INFORMA T/ON
RESPONSES

MK-2394

Figure 1-2

General Summary of DMPll Command Response Structure

1-5

For diagnostic purposes, the microprocessor control command permits the execution of maintenance
instructions, the single stepping of both the microprocessor and the line unit, and operation of the line
unit in loopback mode. This microprocessor control command function is normally used when the
DMPll is not connected to the communications line.

1.4.1.2 Mode Definition Command - A user program issues this command to an associated DMPll
once at system startup time to initialize the DMPll and define the operational state of that device. For
a DMPII, this command defines its function as that of either a control or tributary station in a multipoint network or as a station in a point-to-point network. It also serves to define the station as full- or
half-duplex. This command can also provide DMPll/DMCll line compatibility when these devices
are connected point-to-point.
1.4.1.3 Control Command - This command allows the user program to structure the internal data
base, specify user-defined parameters, and initiate protocol operation. In addition, the control command
provides the user program with access to the status, statistical, and error information maintained in the
internal data base by the DMPll.
With this command, the user program at a control station can change the protocol state of an associated
tributary. Tributaries can also be established and deleted with this command.
Various aspects of dynamic polling, as implemented by the DMPII polling algorithm, can also be controlled by the user program through this command. For example, dynamic polling for individual tributaries can be enabled and disabled and polling parameters can be selectively modified.

1.4.1.4 Buffer Address/Character Count Command - This command has two forms, one for initiating
transmission of a message, and one for assigning a buffer for anticipated message reception.
The form for initiating a transmission directs the DMP11 to transmit a message having a specified
character count from a main memory buffer identified by a designated UNIBUS address to the tributary addressed by the command. If the DMPll is a control station, the tributary address identifies the
tributary that is to receive the message. If the transmitting DMPll is a tributary station, the tributary
address in the command serves to identify the sending tributary to the receiving control station.
The form for allocating buffers for received messages directs the DMPll to store an expected receive
message from a specified tributary in the main memory buffer of designated length at the memory
location pointed to by the UNIBUS address contained in the command.

1.4.2 DMPll Responses
A DMPl1 provides three response messages to the user program arising from message traffic initiated
by the user program. The role of responses in the operation of a DMPll-implemented network is summarized in Figure 1-2. The three responses posted to the user program are:
1.
2.
3.

The buffer disposition count response,
The control response, and
The information response

The general nature of these responses is described in the sections that follow.

1.4.2.1 Buffer Disposition Count Response - This DMP11 response provides the user program with
information on the disposition of messages transmitted and received by the DMPII. For example, when
a message is successfully transmitted and the receipt is acknowledged by the receiving station, the
DMPII issues this response to inform the user program of that fact.
Similarly, when a message is received and stored in the assigned buffer, this response informs the user
program of the buffer completion and identifies the received message. Other forms of this response are
used to inform the user program of unacknowledged transmissions and of transmissions and receptions
not completed because the logical link was disconnected.

1-6

1.4.2.2 Control Response - This response message is issued by the DMPII to report system events and
error conditions encountered during transmission, reception, and diagnostic operations. It is generally
unsolicited. The types of events and error conditions reported include system and protocol events, and
network and procedural errors.
1.4.2.3 Information Response - Specific actions requested by control commands require responses
from the DMPll confirming that the requested action has been performed. The DMPII returns such
responses to the user program by issuing an information response containing the confirming response.
For example, when the user program requests, through a control command, that the contents of a specific error or statistical counter be read, the contents of that counter is returned by an information
response.
In most cases, the DMPII issues an information response in reply to a control command issued by the
user program. However, the DMPll microcode can issue unsolicited information responses to the user
program when the state of the protocol changes. In such cases, the DMPII generally returns all unused
buffers and issues appropriate information responses to the user program to inform it of the action
taken.

1.5 PROTOCOL SUPPORT
In DMPII implemented multipoint and point-to-point networks, all message-data transfers between stations are conducted under the control of DDCMP. Although knowledge of DDCMP is not necessary to
using the DMPll, it is helpful in understanding events and error conditions related to data traffic in a
DMPl1-implemented network. A brief summary of DDCMP is presented in Appendix A.
The DMPII microcode handles all aspects of DDCMP processing including data messages, control
messages, and maintenance messages. At the user program level, message handling involves only passing buffers to the DMPll for transmission and accepting buffers from the DMPII containing received
messages.
There are no file structure restraints on message data transmitted and received over DMPll-implemented networks. In addition, all data is transmitted and received in transparent form so that there are
no restrictions on the type of data that can be transmitted or received under DDCMP.
In multipoint networks, the control station polls each tributary by sending a valid DDCMP message
having the select bit set (see Appendix A). In response to a poll, a tributary can only send a valid
DDCMP data message to the control station. The nature of a multipoint network is such that a tributary can only transmit to the control station. As a result, any communication between tributary stations
must go through the control station.

1.6 SYSTEM APPLICATIONS
As previously stated, the multipoint DDCMP DMPll intelligent controller is a general-purpose device
designed primarily for implementing high-speed multipoint and point-to-point networks. A point-topoint network can be considered as a multipoint network formed by one control station and one tributary station where the two stations alternately serve as control station and tributary station. In a halfduplex, point-to-point network, the transmitter is the control station and the receiver is the tributary.
When the line turns around, the roles reverse. In a full-duplex, point-to-point network, each station simultaneously serves as both a control station and a tributary station.
Because the DMPII is a general-purpose device, detailing specific applications is not as helpful as actually describing a network configuration. Therefore, this section outlines some multipoint network configurations that can be implemented with the DMPII.
A DMPII is equipped with a group of switches that permit the device to be configured for a wide range
of application situations. One set of switches in this group allows a DMPII to be assigned a single

1-7

tributary address. In this case, a specific set of switches define the numerical value of the tributary
address. Switch assigned tributary addresses have the advantage of data transfer security because a
tributary address, once established in the device switches, cannot be changed by the attendant software.
Figure 1-3 illustrates a multipoint network where each DMPll in the network is a single tributary station. The tributary address at these stations can be either software-assigned or switch-assigned.
Figure 1-4 illustrates a multipoint network where the DMPll at each station supports software-assigned multiple tributary addresses. Multiple tributary addressing is advantageous when network resources are distributed throughout a limited physical area. This 32-tributary limit is imposed by the
availability of storage in the microprocessor data memory.
Figure 1-4 also illustrates the tributary status slot (TSS) structures as part of the DMPII data base. In
multipoint networks, these TSS structures are established by the user program as part of the control
and tributary station data bases. TSS structures are maintained at each tributary station with a separate TSS for each tributary maintained at the control station.
A major capability of DMPl1-implemented multipoint networks lies in the ability of the main CPU at
the control station to down-line load programs to tributary stations and execute those programs without
human intervention. As a result, DMPl1-implemented multipoint networks are particularly suited for
installation at remote locations that are inaccessible under most circumstances. For example, DMPlls
could be placed in difficult to reach locations, such as weather stations at sea, and in environments that
are hazardous to humans.
Note that in Figure 1-3, the blocks representing external modems are shown in dotted lines. This representation refers to the ability of a DMPII to connect to the physical communications line through the
integral modem or through an external modem. In general, an external modem is used when the communications line is a low to medium speed long-distance leased or dial-up line. The integral modem is
used when the communications line is a short distance (up to 16000 feet) high speed line (up to 1M
b/s).

1.7

APPLICABLE DOCUMENTS
1.

M8207 Microprocessor Technical Manual (EK-M8207-TM)

M8203 Line Unit Technical Manual (EK-M8203-TM)

DMPll Print Set (MP-00795)

Digital Data Communications Message Protocol (DDCMP) Specification, Version 4.0 (AAD599A-TC)

Maintenance Operation Protocol (MOP) Specification, Version 2.0 (AA-D602A-TC)

Electronic Industries Association (EIA) specifications

1-8

VAX/PDP11
OPERATING
SYSTEM
USER PROGRAM/
DEVICE DRIVER

TRIBUTARY
STATUS SLOTS

#1
#2

DMP11
CONTROL
STATION
(INTERNAL
DATA
STRUCTURE) t--#-3--f
2

r -

......
I

--,

I MODEM I

--,

: MODEM:
L _ _ ...J

L _ _ .....J

TRIB #2

VAX/PDP11

DMP11
#1

r -...,
I MODEM I

DMP11

VAX/PDP11

L _ _ ....I

•••
VAX/PDP11

OPERATING
SYSTEM

USER PROGRAM
DEVICE DRIVER

USER PROG RAMI
DEVICE DRIVER

USER PROGRAM/
DEVICE DRIVER

DMP11
#32

MK-1632

Figure 1-3

Block Diagram of a Multipoint Network with Single Assigned Addresses

VAX/PDP11
OPERATING
SYSTEM
USER PROGRAM
DEVICE DRIVER

TRIBUTARY
STATUS SLOTS

DMP11
CONTROL
STATION
(INTERNAL
DATA
STRUCTURE)

r--""
I MODEM I

I
I
L __ -J

r---,

: MODEM ~

L __ ..J

DMP11
MULTIPOINT
NODE # 1
OPERATING
SYSTEM
USER PROGRAM/
DEVICE DRIVER

#1
STATUS SLOT
FOR SWITCH
SET TRIB.
ADDRESS

DMP11
MULTIPOINT
NODE #2

#3
#2
. #4

OPERATING
SYSTEM
USER PROGRAMI
DEVICE DRIVER

}

r---,
TRIB/STATUS
SLOTS FOR
MULTIADDRESS
STATION
MULTIPOINT
NODE #3

MODEM II

L __ ~

DIMP11

OPERATING
SYSTEM
USER PROGRAMI
DEVICE DRIVER

MK-1634

Figure 1-4

Block Diagram of a Multipoint Network with Multiple Tributary Addresses

CHAPTER 2
INST ALLATION

2.1 SCOPE
This chapter provides all the necessary information for installing and testing the DMP}} microprocessor subsystem. A checklist, which can be used to check the installation process, is also included.
2.2 UNPACKING AND INSPECTION
The DMP11 is packed according to standard packing practices. When unpacking, remove all packing
material and check the equipment against the shipping list (Table 2-1 contains a list of supplied items
for each configuration). Examine all parts and carefully check the module for cracks, loose components, and separations in the etched paths. Report damages or shortages to the shipper and inform
the DIGITAL representative.
2.3 INSTALLATION CONSIDERATIONS
Installation of the OMP11 microprocessor/line unit subsystem should be done in four phases:
•

Phase I - Preinstallation Considerations
Verify system requirements, system placement, and configuration requirements.
Network Topology Chart
For multipoint networks, it is absolutely necessary to know the configuration of the OMP
(that is, control tributary, HOX, FOX, etc.), locations of tributaries (w /address), and where
in the network they are connected (control, Trib 187, Trib 98, Trib 208), or else troubleshooting will be impossible.

•

Phase II - Microprocessor Installation
Configure, install, and verify the microprocessor module via the appropriate diagnostics.

•

Phase III - Line Unit Installation
Configure the line unit module for the customer application and install, cable, and verify it
via appropriate diagnostics.

•

Phase IV - OMP}} System Testing
Verify the DMP}} microprocessor subsystem operation with the functional diagnostics and
system exerciser programs.

2-1

Table 2-1
Option

Parts List

DMPII-AD

DMR 11 Option Packing List
Description
DMPII basic subsystem unit containing:

M8203
M8207-YA
BC08S-I
H3254
H3255

Line unit module
Microprocessor module with DMPII microcode ROMS
Module connection cable
V.35 and integral test connector
RS-232-C/RS-422-A/RS-423-A module test connector

EK-DMPII-UG

DMP11 User's Guide

MP-00795

Customer print set

ZJ-289-RB

Diagnostic set

DMPII-AA

RS-232-C/RS-423 interface configuration containing:
DMPII-AD
BC55C-IO
H325I
H325

DMPII-AB

Basic DMPII unit
EIA RS-232-C/RS-423-A cable
Cable turnaround test connector
Cable turnaround test connector
CCITT V.35 interface configuration containing:

DMPII-AD
BC05Z-25
H3250
DMPII-AC

Basic DMPII unit
CCITT V.35 cable
Cable turnaround test connector
Integral modem interface configuration containing:

DMPII-AD
BC55A-IO
H3257/H3258
DMPII-AE

Basic DMPII unit
Integral modem cable
BC55A terminators
RS-422-A interface configuration containing:

DMPII-AD
BC55B-IO
H325I

Basic DMPII unit
RS-422-A cable
Cable turnaround test connector

It is strongly recommended that a topology diagram be drawn at installation time and maintained
throughout the life of the installation. The topology diagram should provide the information shown below. Figure 2-1 shows a local network topology and Figure 2-2 shows a remote network topology.

2-2

Cable routing

Shows the actual physical location of the cable trough and indicate any equipment which may cause interference, such as an
X-ray room.

Machine type

Shows whether the CPU is a PDP-II/70, PDP11/34, VAX11/780, etc.

Type of station

Identifies whether the station is a control or tributary station.

Physical address

OOCMP address can range from 1-255.

Location

Shows by room number, or other appropriate means, the actual
location of the equipment.

Node name

The name given to the tributary if applicable.

Operating system and
version

The name of the software operating system, such as RSX-ll M
V3.2.

D ECnet version

DECnet software version, such as OECnet-11 M V3.0.

Transmit and recieve

Shows transmit and receive lines. Shows end nodes and terminations. If a patch panel is used, it shows the line numbers between patch panels.

NOTE
The use of patch' panels and the numbering of the
lines is recommended.
2.4 PREINSTALLATION CONSIDERATIONS
Table 2-1 should be considered prior to ordering a OMP11 communications interface to ensure that the
system can accept the OMPll and that it can be installed correctly. Sections 2.4.1 through 2.4.2 should
also be verified at installation time.
2.4.1

System and Device Placement

2.4.1.1 System Placement - On systems that contain many high-speed, direct memory access (DMA)
devices, there is a chance of adverse bus latency. To help prevent this, the closer the physical placement
of the OMP11 to the processor, the higher the OMA device priority. Customer applications using
speeds greater than 250K bits per second (b/s) require UNIBUS placement before all UNIBUS repeaters and before all devices that have a lower NPR rate on the UNIBUS.
2.4.1.2 Device Placement - The DMP11 requires two hex-height, small peripheral controller (SPC)
backplane slots (preferably two adjacent slots). Any SPC backplane [0011-B (REV E) or later] can
accept the DMPl1. The 0011-0 can accommodate a maximum physical configuration of three
OMP11s (see CAUTION).
CAUTION
Each DMPll requires approximately 8.5 amperes
from the + 5 volt source. For example, three
DMPlls in a 0011-0 can overload the "744 supply. Therefore, there is a restriction to 2 per 0011o with each supply providing 24 amps of + 5 V. The
"7441 supply , however, can supply 32 amperes.
2-3

I
I
RX

I TX

?ORSHAFT

DELTA
11/70
DMP11
TRIB 3
RSTS XX.X
DE - XX.X

SUSPENDED CEILIN G
ROOM 515

,.
T

I
I

I TX

11
1J

BETA
11/34
OMP 11
TRIB 1
RSX11 M
OM - XX.X

GAMA
11/23
OMV 11
TRIB 2
RSX 11 M
OM - XX.X
ROOM 412

ROOM 430

.....

ALPHA
VAX 11/780
OMP 11
CONTROL
VMS XX.X
OV - XX.X
ROOM 111
MK-2S02

Figure 2-1

Example of a Topology Diagram for a Local Multipoint Network

2-4

MERRIMACK
MK1-1/K37
VAX 11/780
VAX VMS XX.X
DV-XX.X
*NOTE 1
*NOTE 3
CONTROL
NOTE 1
PATCH PANELS ARE
RECOMMENDED BUT MAY
NOT ALWAYS BE USED IFTHEY
ARE USED, THEIR PHYSICAL
LOCATION SHOULD BE
INDICATED.

NASHUA
NU
4800

NASHUA
EXCHANGE

11/60
RSTS XX.X
DE-XX.X
*NOTE 1
*NOTE 3

4800

NOTE 2
MODEM - 208A
CONTROL STATION OPTIONS
SEE TABLE 2-2

TRIB 1

*NOTE 3
MODEM - 208A
TRIBUTARY OPTIONS
SEE TABLE 2-3

TEWKSBURY
TW

.--_-11 LOWELL
EXCHANGE

~LnVAX 11/750

----...

IUVAX VMS XX.X
DV-XX.X
*NOTE 1
*NOTE 3
TRIB 2

4800

MARLBORO
MR

MA YNAR D
EXCHANGE

4800

11/34
RSX-11 M
V.3.2
DM XX.X
*NOTE1

r-__

MARLBORO
1--+---1

*NOTE 2
TRIB 4

EXCHANGE

MAYNARD
PK3
11/70
RSX-11 M
V.3.2
*NOTE 1
TRIB 3

MK-2501

Figure 2-2

Example of a Topology Diagram for a Remote Multipoint Network

2-5

Table 2- 2 Typical Control Station Options of a Bell 208A Compatible Data Set
(4800 b/s) Full-Duplex Operation
Data Set Options

DEC Recommended
Settings

Transmitter Timing Provided

Data set (internal)

Carrier Control

Continuous

Request-to-Send Operation
in Continuous Carrier Mode

Continuous (CB constantly on)

One Second Holdover at
Receiver on Line Dropouts

Not provided

New Sync Option to Squelch
Receiver Clock

Not used - NS is strapped OFF
within the data set

Data Set Ready Lead Option
for Analog Loopback Testing
by Data Terminal

CC is on when the AL button (only)
is pressed

Grounding Option

AB connected to AA

Table 2-3

Typical Tributary Options of a Bell 208A or Compatible Data Set
(4800 b/s) Full-Duplex Operation

Data Set Options

DEC Recommended
Settings

Transmitter Timing Provided

Data set (internal)

Carrier Control

Switched (48.5 ms CA-CB delay)

Request-to-Send Operation
in Continuous Carrier Mode

Continuous (CB constantly on) N I A
Switched (8.5 ms CA-CB delay)
NjA

One Second Holdover at
Receiver on Line Dropout~

Not provided

New Sync Option to Squelch
Receiver Clock

Not used - NS is strapped off
within the data set

Data Set Ready Lead Option
for Analog Loopback Testing
by Data Terminal

CC is on when the AL button (only)
is pressed

Grounding Option

AB connected to AA

2-6

2.4.2

System Requirements
I.

UNIBUS Loading (Recommended)
M8207-Y A microprocessor
1 UNIBUS dc load
5 UNIBUS ac loads
M8203 line unit
No UNIBUS loads

Power Requirements
Check the power supply before and after installation to ensure against overloading. The microprocessor /line unit total current requirement for the + 5 volt supply is approximately 8.5
amperes. Additionally, the unit requires ± 15 volts for the silos, level conversion logic, and
integral modem. Power requirements for the microprocessor/line units are listed in Table 24.
Table 2-4 DMRll Voltage Chart

Module

Voltage Rating
(Approximate Values)

Maximum
Voltage

Minimum
Voltage

Back Plane
Pin

M8207-YA

+ 5 Volts @ 5.0 A

+ 5.25

+ 5.0

CIA2

M8203

+ 5 Volts @ 3.5 A
+ 15 Volts-@.1 A

+ 5.25

CIA2

+ 15.75

+ 5.0
+ 14.25

CIUI

-15 Volts @ .2 A

-15.75

-14.25

CIB2

Interrupt Priority
The interrupt priority is selected by priority plug E77 on the M8207-Y A microprocessor
module. This plug is preset to select priority five (BR5). This corresponds to IPL 21 on VAX
systems. Refer to Figure 2-3 for the priority plug location.

Device Address Assignment
The DMPII is in the floating address space of the input/output (I/O) page of memory.
The selection of the device address is accomplished by switch pack E127 on the M8207-Y A
microprocessor module. Refer to Figure 2-3 for the switch pack placement. The DMPII has
a ranking of 21 in the floating list.
Refer to Appendix B if more information is needed on the floating address allocation.

Device Vector Address Assignment
The DMPII is in the floating vector space of the reserved vector area of memory. The ranking assignment of the DMPII is ranking number 42. The selection of the device vector address is accomplished by switch pack E28 on the M8207-Y A microprocessor module. Refer
to Figure 2-3 for the location of the switch pack. Appendix B contains more information on
floating vector allocation.
2-7

DEVICE ADDRESS SELECTION

A3--- - - - - - - - - - - - - A12

SWITCH PACK E127

SWITCH OFF = LOGICAL 1
(REFER TO TABLE 2-6)

c::=

~PRIORITY

~PLUG

SWITCH PACK E85

SWITCH PACK E28

L---...J
SPARES

VECTOR ADDRESS
SELECTION

BSEL 1
LOCKOUT

RUN
INHIBIT
MK-2527

Figure 2-3
6.

CSR
DISABLE

SWITCH ON = LOGICAL 1
(REFER TO TABLE 2-7)

M8207-YA Microprocessor Switch/Jumper Locations

Customer Application Considerations
The DMPll is configured according to the customer's application and configuration requirements. Ask the customer to determine these requirements for the DMPll being installed (refer to Tables 2-5 through 2-8).
Table 2-5

M8207- YA DMPll Jumper Chart

Jumper
Number

Normal
Configuration

Always In

Microprocessor Clock Enable - When removed it disables the microprocessor clock. Removed only for automatic module testing
at the factory.

Always In

Bus ac Low Enable - When removed it disables a program asserted ac low signal passed onto the UNIBUS.

Function

2-8

Table 2-6 Switch Pack El27 Selections
Switches

Function

1-10

Device Address Selection:

LSB

MSB
15 14

SWITCH
NUMBER

12111 110 1 9

I 8 1 7 I 6 15 I 4 I 3

SWITCH PACK E127

••

••
••
••
•
•
• DEVICE
•
S10 S9 S8 S7 S6 S5 54 53 52 51
I

ADDRESS

OFF
OFF
OFF OFF
OFF
OFF
OFF
OFF OFF
OFF OFF OFF
OFF

760010
760020
760030
760040
760050
760060
760070
760100

---

OFF

760200

OFF OFF

760300

---

OFF

760400

OFF

---

OFF

760500

----760700
--761000
--762000
--763000
-764000

OFF OFF

760600

OFF OFF OFF
OFF
OFF
OFF OFF
OFF

NOTE: SWITCH OFF RESPONDS TO LOGICAL ONE ON THE UNIBUS.

SWITCH PACK E127
DEVICE ADDRESS
SELECTION
A3 - - - - - - - - - - - - - - - - -- -- A12

o 1

~~~~~~~~88~
2-9

Switch Pack
EI27

Table 2-7
Switches

Switch Pack E28 Selections

Function
NOTE
Switch ON equals a logical one (t) on the UNIBUS.

1-6

Vector Address Selection:

MSB

LSB

15 14 13 12 11
0
0 0 0 0

I I I I

S6 S5 S4 S3 S2 S1

VECTOR
ADDRESS

••
•
SWITCH

NUMBER

0
0

8 J 7 6 5 4 3
SWITCH PACK E28

ON
ON
ON
ON
ON
ON
ON
ON

ON
ON
ON ON
ON
ON
ON
ON ON
ON ON ON

1
0

300
310
320
330
340
350
360
370
400

--ON

500

--ON ON

600

ON ON ON

700

--NOTE: SWITCH ON PRODUCES LOGICAL ONE ON THE UNIBUS.
SWITCH PACK E28
V3 - - - - - - - - - - - V8

;~~~~e~~Q
\...--

VECTOR ADDRESS
SELECTION

RUN INHIBIT*.J
CSR DISABLE*

* Refer to Table 2-6 for switch functions.

Table 2-8

Switch Selectable Features

Switch
Location
and Number

Normal
Configuration

E28
Switch 7

Run Inhibit - Under normal conditions, (switch ON) the initialization of the microprocessor directly sets the run flip-flop
which allows the microcode to begin execution immediately. If an
internal malfunction or execution of erroneous microcode occurs
during power up, it is possible for the microprocessor to hang the
UNIBUS. Placing the run inhibit switch in the OFF position
clears the run flip-flop and allows the diagnostics to be loaded to
determine the fault. Once diagnostics have been executed and the
problem corrected, the run switch must be returned to the ON
position.

E28
Switch 8

OFF

CSR Disable - When the switch is ON it keeps the device from
responding to its address. Used in special applications on other
options.

E85
Switch 1

BSEL 1 Lockout - When the switch is OFF, it inhibits the use of
all maintenance test features and keeps the run bit asserted at all
times. This allows the DMPII microcode to run with the ability
to detect a boot request message at all times. This is used at unattended computer sites. This switch must be ON to execute all diagnostics.
NOTE
Although BSEL 1 maintenance functions are inhibited, master clear (bit 6) is still functional.

E85
Switches
2, 3,4, and 5

2.5

OFF

Function

Not Used (see note in Table 2-6)

MICROPROCESSOR INSTALLATION

2.5.1 Backplane Considerations
Perform the following on the SPC slot that will contain the DMPII M8207-Y A microprocessor module
(selected at preinstallation).
1.

Verify that the backplane voltages are within the specified tolerances listed in Table 2-4.

Turn system power off and remove the NPR grant (NPG) wire that runs between CAl and
CBI on that backplane slot for the M8207-YA module.
NOTE
Be sure to replace this jumper if the microprocessor
is removed from the system.

Perform resistance checks on the backplane voltage sources to ground to ensure that no short
circuit conditions exist. Refer to Table 2-4 for backplane pin assignments.
2-11

2.5.2

M8207-YA Considerations

Perform the following on the DMPII M8207-Y A microprocessor module.
I.

Ensure that the module version number is an M8207-Y A, which shows DMPll microcode.

Verify that M8207-Y A jumpers WI and W2 are installed correctly (refer to Table 2-5).

Configure switch pack E 127 to implement the correct device address for the DMP II, as determined from the floating address allocation. Refer to Table 2-6 for the correlation between
switch number and address bit. A switch OFF (open) responds to a logical one on the
UN I BUS. Refer to Appendix B for additional information on floating address allocation.

Configure switch pack E28 to implement the correct vector address for the DMPII as determined from the floating vector allocation. Refer to Table 2-7 for the correlation between
switch number and vector bit. A switch ON (closed) responds to a logical one on the
UN I BUS. Refer to Appendix B for additional information on floating vector allocation.

Verify that the switch selectable features of the M8207-Y A are configured as follows (Table
2-8 provides a summary of switch selectable features):
•

Run Inhibit - Switch pack E28, switch 7; always ON.

•

CSR Inhibit - Swtich pack E28, switch 8; always OFF.

•

Byte Select/Lockout - Switch pack E85, switch 1; normally ON (allows all functions in
BSEL 1 to be used). If the switch is OFF, the Run bit is always asserted and will not
allow diagnostic testing.

NOTE
For switch pack E85, only switch 1 is used.
6.

2.5.3

Verify that the priority plug is a BR5 and is installed correctly in location E77.

M8207 -YA Insertion

Carefully insert the M8207-Y A microprocessor module into the selected SPC slot and perform the following tests:
I.

Perform resistance checks on the backplane voltage sources to ground to ensure that no short
circuit conditions exist on the module. Refer to Table 2-4 for backplane pin assignments.

Turn system power ON and verify that the backplane voltages are within the specified tolerances listed in Table 2-4.

Load and execute the M8207 static diagnostics, parts one and two (no test connectors are
required). See Table 2-9.
On obtaining a minimum of five error free end passes, proceed to the M8203 line unit installation section.

2.6 M8203 LINE UNIT INSTALLATION
The M8203 line unit is a universal module with various types of interface capabilities. The M8203 line
unit does not present any ac or dc loads to the UNIBUS and only draws power from the backplane slot
in which it is. All data and control signals flow into and out of the line unit via a berg port to the
microprocessor. Because of the various M8203 applications, the configurations for each may be differ-

2-12

ent and are selected via switches, jumpers, and different cables. To provide a better understanding of
these variations, a number of tables describing each switch pack, jumper, and cable function (as listed
below) has been created for reference. Table 2-10 lists the normal M8203 line unit configurations for
the different types of OMPll options. Also, refer to the following:

NOTE
Modem interface signal jumpers are included on the
BCSSC panel and, therefore, do not have to be configured on the M8203 module.
•

Table 2-11 Jumper Functions - These jumpers are used to select various interface standard
parameters and modem interface signals, depending on application and modem type. Additional jumpers are av::tilable on the BC55C (panel) cable for additional interface signal selection.

•

Table 2-12 Switch Pack E39 Functions - This switch pack allows proper selection of interface driver and receiver control logic and different line speeds for various applications.

Table 2-9

DMPII Diagnostics

PDP-II
Diagnostic

Description

CZOMP**

M8207 microprocessor static diagnostic I

CZOMQ**

M8207 microprocessor static diagnostic 2

CZOMR**

M8203 line unit static diagnostic 1

CZOMS**

M8203 line unit static diagnostic 2

CZOMT**

OMPl1 functional diagnostic

CZCLM**

DMPII OCLT program

CXOMO**

OMPll DEC/XII master module

CXOME**

DMPll DEC/XII slave module

VAX-II
Diagnostic

Description

EVOXA
REV 3.0 and Above

CO MM microprocessor repair level
diagnostics (Level 3)

EVOMA
REV *.*

M8203 line unit repair level
diagnostics

EVOMB
REV *.*

VAX 0 MP II functional diagnostic

2-13

Table 2-10

Normal M8203 Configuration

Option Type:

DMRll-AA
EIA RS-232-C
or RS-423-A

DMRII-AB DMRII-AE
CCITT V.35 EIA
RS-422-A

DMRll-AC
Integral
Modem
56K b/s

DMRII-AC
Integral
Modem
I MEG b/s

DMRII-AA
Null Modem
Clock
9.6K b/s

Jumper
Configuration·
(refer to Table 2-11)

WI-W6, WII
WI4-WI7

OUT

W7-W10
W12,W13

Switch Pack E39

SI-4
SS
S6
S7
S8
S9
SIO
SI-4

OFF
OFF
OFF
ON
OFF
OFF
OFF
OFF
ON
ON
ON
ON
OFF
OFF
Unit Number Selection

OFF
OFF
OFF
ON
ON
ON
OFF

OFF
OFF
OFF
OFF
OFF
OFF
ON

OFF
OFF
OFF
OFF
ON
ON
ON

OFF
OFF
OFF
OFF
OFF
ON
OFF

SS-7

Mode Definition Selection

Mode Enable

Remote Load Detect

S10
SI-8
S9
S10

Power On Boot
Tributary Address/Password
Not Used
High Speed/Low Speed
BCSSC-IO·
BCOSZ-2S
and
BC05D-25

BCS5B-IO
and
BC55D-33

BC55A-IO
and Twinax
Cables

BC5SA-IO
and Triax
Cables

BCSSC-10*
and Null
Modem

H3254
IN Jl

H3254
IN J2

M8203
Configuration

(refer to Table 2-12)

Switch Pack E 121 t
(refer to Table 2-13)

Switch Pack EI34 t
(refer to Table 2-14)
Cables Required
(refer to Figure 2-6)
Module Turnaround
Test Connector
(refer to Figure 2-7)
Cable Turnarouna
Test Connector
(refer to Figure 2-7)

H3255
IN J2

H3254
IN 11

H3255
IN 12

EIA RS-232-C
H325

H32S0

H3251'

EIA RS-423-A
H3251

*Modem variable
tCustomer application variable (there is no normal setting for these switches)

EIA RS-232-C
Half-duplex
Half-duplex
H325
Switch ON and Switch ON and
EIA RS-423-A
Cables
Cables
H3251
Removed
Removed

Table 2-11
Jumper

M8203 Jumper Functions

Normal
Configura tion

Function

OUT

Clear to send EIAjV.35

OUT

Data mode EIAjV.35

OUT

Receive data EIA

OUT

Receive clock EIA

OUT

Receive ready EIA

OUT

Transmit clock EIA

NOTE
These jumper positions can be used to insert capacitors or resistors, depending on interface standard.
•

RS-423-A - Capacitors can be inserted in noisy
environments. Refer to RS-423-A Standard,
Section 4.1.6, Wave Shaping, for capacitor selection.

•

RS-422-A - Resistors are required to properly
terminate the receivers in order to obtain distances of 1219.2 meters (4000 feet). Unterminated receivers only obtain distances of 12.2 meters (40 feet). For ease of installation, these
terminating resistors (100 ohms) have been installed in the RS-422-A cable BC55B-I0.

Signal Rate Indicator - When removed, opens signal to interface
in RS-422-Aj423-A and RS-232-C configurations. Remove when
a BC05C-XX* cable is used as this signal is presented to cinch
pin nine, which has a positive test voltage on some modems.

Data Mode (Data Set Ready) - When removed, opens signal to
interface in RS-422-Aj 423-A configurations. It has no effect in
RS-232-C. Remove when a BC05C-XX* cable is used, as the signal is presented to cinch pin 18, which is dibit clock on some
modems.

Null Modem Clock - When removed the signal amplitude is
lowered below the interface standards so as not to create interference in some modems.

*DMPII does not support the use of a BCOSC-XX cable.

2-15

Table 2-11
Jumper

Normal
Configura tion

M8203 Jumper Functions (Cont)

Function

WIO

Terminal Ready - When removed, it opens the signal to modem
in RS-422-A/423-A configurations. Remove when a BC05C-XX*
cable is used, as this signal is presented to cinch pin 10, which is
negative test voltage on some modems.

WII

OUT

Receiver Ready (Carrier Detect) - When installed, it allows this
signal to be on at all times. This could cause a problem with the
microcode
since
the
Universal
Synchronous
Receiver/Transmitter (USYRT) will be enabled all the time.

WI2

Terminal in Service (Make Busy) - When removed, it opens this
signal to the modem. Some modems will not answer the phone
and will be put in analog loopback when this signal is asserted.
When a BC05C-XX* cable is used, this signal is presented at
cinch pin 25.

WI3

Oscillator Enable - To be removed only for factory automatic
testing. Jumper should always be installed in the field.

W14 and
W15

OUT
OUT

WI6

OUT

Switched RTS-CTS Enable - When jumper is installed, it enables the request to send and clear to send interlock in the M8203
line unit which inhibits asserting RTS, until CTS is dropped. This
jumper should never be installed when the DMPII is operating
with a modem that has the constant CTS option installed.

WI7

OUT

Half-Duplex Lockout Enable - When this jumper is installed, it
enables the M8203 line unit half-duplex lockout feature when
half-duplex mode is selected. The lockout feature disables the
transmitter or receiver when the other is active. This jumper applies only to half-duplex applications. It must not be installed for
full-duplex applications.

56K Bandpass Filter Enable - With these jumpers installed, the
bandpass filter is limited to 56K b/s. Used in special applications
only.

NOTE
Jumpers W16 and W17 are mutually exclusive. Only
one or the other may be installed, not both. Also,
these jumpers are provided only on M8203 modules
REV E or higher. For modules up to REV D, refer
to ECO-M8203-MK-007 for details of similar
jumpers.

*DMPII does not support the use of a BC05C-XX cable.

2-16

Table 2-12

Switch Pack E39 Selections

NOTE
Switch off equals a logical one (1).
Switches

Function

1-4

Not used in DMPII.

5-7

Interface Selection - Selects proper drivers and receivers for each interface
type:
•

SW5, 6, and 7 all OFF - Selects EIA RS-232-C and RS-423-A
interface logics which are single-ended (unbalanced) drivers and
receivers. The integral modem logic is selected with the BC55A-IO
cable when installed in Jl of the M8203.

•

SW5 ON - Selects V.35 interface logic which has balanced drivers
and receivers.

•

SW6 ON -

•

SW7 ON - Selects EIA RS-422-A interface logic with balanced
drivers and receivers.

Reserved.

Interface Type

SW5

SW6

SW7

RS-232-C, RS-423-A,
Integral Modem *
V.35
RS-422-A

OFF
ON
OFF

OFF
OFF
OFF

OFF
OFF
ON

(Switches 8-10)

Line Speed Selection - Selects modem speed for integral modem applications,
null modem applications, and diagnostic testing.

Switch

Speed

1 MEG
500K
250K
56K

ON
OFF
ON
OFF

ON
ON
OFF
OFF

ON
ON
ON
ON

19.2Kt
9.6K
4.8K
2.4K

ON
OFF
ON
OFF

ON
ON
OFF
OFF

OFF
OFF
OFF
OFF

* Integral modem is selected by BeS5A cable when installed in J 1 of the M8203 line unit. Module connector J2 must not have
a cable or test connector installed.

tNormal switch setting unless the integral modem or null modem clock features are used.

2-17

•

Table 2-13 Switch Pack E 121 Functions - This switch pack is provided for the selection of
the mode of operation of the DMPII (for example, Pt-to-Pt, multipoint control, trib, etc.)
and for various microcode features. See Table 2-11 for details.

•

Table 2-14 Switch Pack E134 Functions - This switch pack is provided for the selection of
the tributary address and remote load detect password. See Table 2-7 for details.

•

Table 2-15 Cable Description - This table lists the functions and uses of each cable used with
the DMPII.

•

Figure 2-4 shows the jumper and switch pack placements on the M8203 line unit.

•

Figure 2-5 shows the jumper and switch pack placement on the M8203 line unit.

•

Figure 2-6 shows the outline drawings of DDMPII cables.

2.6.1 M8203 Considerations
Configure all appropriate switch settings and jumpers on the M8203 line unit module according to the
recommendations in Table 2-15.

NOTE

If the customer has additional requirements because

of modem restrictions or a bootstrapping feature, be
sure to configure the line unit to these requirements
using the information contained in Tables 2-11
through 2-14.
2.6.2 M8203 Insertion
With system power OFF, carefully insert the M8203 line unit module into the proper backplane slot
(usually adjacent to the microprocessor) and perform the following:
1.

Connect the line unit and the microprocessor using the BC08S-1 cable. One end of the cable
is connected to Jl of the M8207-Y A microprocessor module and the other end to J3 of the
M8203 line unit module. Carefully fold the cable back to the right, tightly against the component side of either the microprocessor or line unit module, so as to fit it into the mounting
box. Refer to Figure 2-5 for connector layouts.

Insert the appropriate module test connector into the correct line unit connector as specified
in Table 2-10. Be sure to insert with "SIDE 1" (etched on the test connector) visible from the
component side of the line unit. See Figure 2-5 for test connector orientation.
Schematics and outline drawings of each test connector used with the DMPII are provided
in Figure 2-7.

Turn system power ON and perform voltage checks on the line unit backplane slot. Ensure
that the voltages are within the specified tolerances as listed in Table 2-4.

Load and execute the M8203 static diagnostics, parts one and two, with external maintenance mode selected. (See Table 2-9.)
Chapter 6 provides detailed information on these diagnostic routines. Upon obtaining a minimum of five error free end passes, with the module turnaround test connector installed, proceed with step 5.

Remove the module turnaround test connector and connect the appropriate cable (see Table
2-10) to the proper berg connector for the DMPII option selected. Refer to Table 2-15 for
detailed information on cable requirements.
2-18

Table 2-13
Switch

Switch Pack EI21 Selections

Function

NOTE
Switch off equals a logical one (I).
1-4

Unit Number Selection - These switches are physically connected to IBUS register
16 Bits 0-3 with switch 1 being the least significant bit (LSB) and switch 4 the most
significant bit (MSB). The unit number is used to identify each unique DMPII if
several are installed on the same system. This is particularly important for systems
utilizing the remote load detect and booi features of the DMPII. The boot program
must know which of the DMPIIs in the floating address space is performing the
down line load.

5-7

Mode Definition Selection - These switches are physically connected to IBUS register 16 Bits 4-6, and are used in conjunction with switch 8 to define the mode of operation for the DMPII.
SWITCH

MODE

HDX PT-TO-PT IDMC Line Compatible
FOX PT-TO-PT IDMC Line Compatible
HDX PT-TO-PT*
FOX PT-TO-PT*
HDX MULTIPOINT CONTROL STATION*
FOX MULTIPOINT CONTROL STATION*
HDX MULTIPOINT TRIBUTARy*
FOX MULTIPOINT TRIBUTARY*

ON
OFF
ON
OFF
ON
OFF
ON
OFF

ON
ON
OFF
OFF
ON
ON
OFF
OFF

ON
ON
ON
ON
OFF
OFF
OFF
OFF

*DDCMP V4.0
8

Mode Enable - This switch is physically connected to IBUS register 16 bit 7, and is
used to indicate that the mode definition is actually selected in switches 5-7.
Switch in off Position: Mode defined in switches Switch in on Position: Mode not defined in switches defined by software (overridden by software.

Remote Load Detect - This switch is physically connected to IBUS register 11 bit 2,
and is used to enable the remote load detect feature of the DMPII.
Switch in off Position: Enables
Switch in on Position: Disables

Power on Boot - This switch physic'ally connected to IBUS register 11 bit 1, and is
used to enable or disable the power ON boot feature of the DMPIl.
Switch in off Position: Disables
Switch in on Position: Enables

NOTE
If the remote load detect and power on boot features
are used, then switches I through 10 must be appropriately set. Remote load detect and power on boot
features are only applicable to PDP-II processors
and do not apply to VAX-II processors.

2-19

NOT USED
HIGH SPEED/LOW SPEED

SWITCH ON
SWITCH OFF

MODE ENABLE
REMOTE LOAD DETECT

SWITCH PACK E134
UNIT
NUMBER
~

=0
=1
SWITCH

POWER ON BOOT

MODE
DEFINITION

r---"-l

SWITCH PACK E121

W3 W4

t MODE DEFINITION

HDX PT-TO-PT/DMC LINE COMPATIBLE
FOX PT-TO-PT/DMC LINE COMPATIBLE
HDX PT-To-prxFDX FT-TO-PT·j(HDX MULTIPOINT CONTROL STATION'xFOX MULTIPOINT CONTROL STATlON·j(HDX MULTIPOINT TRIBUTARy·j(FDX MULTIPOINT TRIBUTARy·j(-

ON
OFF
ON
OFF
ON
OFF
ON
OFF

ON
ON
OFF
OFF
ON
ON
OFF
OFF

ON
ON
ON
ON
OFF
OFF
OFF
OFF

WG
*DDCMP V4.0

tv
I
tv

SWITCH
SPEED

W160

1MEG

500K

OFF ON

250K

OFF

56K

OFF OFF

·)(-19.2K

OFF

9.6K

OFF ON

OFF

4.8K

2.4K

OFF OFF OFF

W17

* NORMAL SWITCH SETIING UNLESS THE
INTEGRAL MODEM OR NULL MODEM CLOCK
FEATURES ARE USED.

OFF OFF

RS-232-C OR RS-423-A
OR INTEGRAL*')(-

SELECTS DATA RATE
(SEE TABLE 2)

RESERVED

SELECTS RS-422 INTERFACE
RESERVED
SELECTS V-35 INTERFACE

]

SW5

SW6

SW7

OFF

V.35

OFF

RS-422

OFF

** INTEGRAL MODEM IS SELECTED BY BC55A
CABLE WHEN INSTALLED IN J1 OF THE
M8203 LINE UNIT. MODULE CONNECTOR
J2 MUST NOT HAVE ANY CABLE OR TEST
CONNECTOR INSTALLED.
MK·2524

Figure 2-4

M8203 Switch and Jumper Locations

Table 2-14
Switch

Switch Pack E134 Selections

Function

NOTE
Switch off equals a logical one (1).
1-8

Tributary Address/Password - These switches are physically connected to IBUS register 15 bits 0-7, with switch 1 being the least significant bit (LSB) and switch 8 the
most significant bit (MSB). The tributary address switches allow the user to define
the address in switches which the device will respond to, for message traffic over the
communication link. Valid addresses are: 1 for PT-to-PT stations, and 1 through 255
for multipoint tributary stations; address of zero is illegal. If the power on boot and
remote load detect features are used, these switches must be set to the tributary address. In addition, for remote load detect this address also serves as a password for the
maintenance message ENTER MOP MODE. (Tributory address and Password must
be equal.

Not used in DMPII - Switch is physically connected to IBUS register 11, bit 3.

High Speed/Low Speed - This switch is physically connected to IBUS register 11,
bit 5. The setting of this switch indicates to the microcode whether the data rate over
the communication line is greater or less than 250K b/s.
Switch in off Position: High speed (250K b/s)
Switch in on Position: Low speed «250K b/s)

NOTE
If the boot feature is selected, it is mandatory that
the switches in switch packs E121 and E134 be appropriately configured according to the following
charts:
LSB

E121

MSB

OFF

'v"""

'v"'"
EXAMPLE:

UNIT NUMBER

= 3, MODE = FOX TRIB.
MK·2731

LSB

E134

MSB

OFF

7
OFF

X
J

= DON'T CARE

EXAMPLE: TRIB ADDR
REPRESENTATION (123 10

10
X
HIGH/LOW
SPEED

= 123 (DECIMAL), MUST BE SET IN OCTAL
= 173 8 )
MK·2732

2-21

Table 2-14

Switch Pack E134 Selections (Cont)

LSB
E121

MSB

OFF

EXAMPLE:

UNIT NUMBER

= 4, MODE = HDX TRIB.
MK·2733

MSB

LSB
E134
X

OFF

4
OFF

OFF

= DON'T CARE.

8
ON

X
HIGH/LOW
SPEED

= 123 (DECIMAL), MUST BE
= 173 8 )

EXAMPLE: TRIB ADDRESS AND PASSWORD
SET IN OCTAL REPRESENTATION (123 10

Table 2-15
Interface

MK·2734

Cable Description
Description

RS-232-C

•

Cable Assembly:
BC55C-IO (Refer
to Figure 2-6A)

•
•

M8203 Connector: J2
Test Connector: H325

A 3M (10 feet) cable with a 40 pin, berg
connector at one end which is installed
into J2 of the M8203 so that the ribbed
side of the cable faces out. This
creates a half twist in the cable and is
required for proper pin connections. The
other end has a panel bracket that includes three different cinch connectors,
J I, J2, and J3. Connector J2 is used for
RS-232-C and is connected to the modem
with external cable BC05D-25. The bracket
must be mounted on the rear mounting rail
of the cabinet to ensure proper grounding
and for easy access to external cable
connections.
The BC55C panel has several jumpers.
Depending on the modem option selected,
certain jumpers must be installed. Refer
to Table 2-13 for detailed jumper configurations.

External Cable

•

BC05D-25 (Refer to
Figure 2-6E)

A 7.5 M (25 feet) external cable that
connects J2 of the BC55C panel to an
RS-232-C modem.

2-22

Table 2-15
Interface
•

Cable Description (Cont)
Description

Test Connector:
H325

RS-422-A

•

Cable Assembly:
BC55B-IO
(Refer to Figure
2-6B)

A 3m (10 feet) cable with a 40 pin, berg
connector at one end which is installed
into J2 of the M8203 so that the ribbed
side of the cable faces out. This creates
a half twist in the cable and is required
for proper pin connections.
This cable is similar to the BC55C in
that it has a panel bracket at the other
end. There is, however, only one connector on the panel (J2) that is used
with cable BC55D-33 for external connection to the modem. The bracket must be
mounted on the rear mounting rail of the
cabinet to ensure proper grounding and
easy access to external cable connections. The BC55B panel has two jumpers
(WI & W2) for shield grounding connections. Normally, WI is always out
(special application only) and W2 normally in. For RS-449 applications, W2
can be removed to place a 100 ohm resistor between circuit ground and frame
ground to dissipate ground currents.

External Cable

•

BC55D-33 (Refer to
Figure 2-6 F)

•

Test Connector:
H3251

A 10m (33 feet) external cable that
connects J2 of the BC55B panel to an
RS-422-A modem.

RS-423-A

•

Cable Assembly:
BC55C-IO (Refer to
Figure 2-6A)

Same cable as used for RS-232-C except
that panel connector J 1 is used with
external cable BC55D-33 for connection to
the modem. The bracket must be mounted
on the rear mounting rail of the cabinet
to ensure proper grounding and easy
access to external cable connections. The
BC55C panel contains several jumpers.
Depending on the modem selected, certain
jumpers must be installed. Refer to
Table 2-13 for detailed jumper configurations.

2-23

Table 2-15

Cable Description (Cont)
Description

Interface
External Cable

•

BC55D-33 (Refer to
Figure 2-6F)

•

Test Connector:
H3251

A 10m (33 feet) external cable that
connects Jl of the Be55C panel to an
RS-423-A modem.

V.35

•

Cable Assembly:
BC05Z-25 (Refer
to Figure 2-6G

•

Test Connector:
H3250

A 7.5 m (25 feet) modem cable with a 40
pin, berg connector at one end that connects to Jl of the M8203. A 37 pin, Data
Phone Digital Service (DDS) connector is
installed at the other end and connects
to the modem.

Integral Modem

•

Cable Assembly:
BC55A-I0 (Refer
to Figure 2-6C)

•

M8203 Connector:
Jl

•

Test Connector:
NONE (Place panel
HDX Position for
turnaround)

A 3 m (10 feet) cable with a 40 pin, berg
connector at one end that plugs into Jl
of the M8203 with the cable strain relief
tab facing out. A BC55A connector panel
is installed at the other end. This panel
contains four connectors, two female and
two male. The panel also switch to includes a toggle switch to select either
full-duplex or half-duplex. The panel is
mounted on the rear mounting rail of the
cabinet to ensure easy access to external
connections and for proper grounding.
Appropriate terminator connectors H3257
or H3258 must be used. See Figures 2-7
and 2-8.

NOTE
Ensure that all cables mounted in the M8207 and
M8203 are properly installed and seated in the berg
connectors.
External Cables

•

BC55N-98 (Refer to
Figure 2-6D)

•

Test Connector:
NONE

•

BC55M-98 (Refer to
Figure 2-41 )

A 29.4 m (98 feet) external twinax cable
used to connect a DMCII to a DMRll or a
DMRII to a DMRll for a selected data rate
of 56K b/s.
A 29.4 m (98 feet) external triaxial
cable used for the same purpose as the
BC55N, but for data rates above 56K b/s.

2-24

Table 2-15

Cable Description (Cont)

Interface

Description

•

For full-duplex, connect the male end of
one BC55N or BC55M cable to the receive
out of the connector panel. Connect the
female end of the same cable to the
transmit in of the connector panel at the
other station.
Connect the male end of the other cable
(BC55N or BC55M) to the transmit out and
the female end to receive in at the other
station.
For half-duplex, connect the male end of
the cable (BC55N or m) to the receive out
and connect the other end of this cable
to the receive in at the other station.

Test Connector:
NONE

CONNECT CABLE BC05Z FOR V.35 INTERFACE OR
BC55A FOR INTEGRAL MODEM.

H3254

I~ ~

H3255

I~ ~ - - -

TEST CONNECTOR

TEST CONNECTOR ~

---

I
I
C;J

M8203
LINE UNIT

.,
I

CONNECT CABLE BC55C-10
FOR RS-232-C OR RS-423-A
INTERFACE, OR BC55B-10 FOR
RS-422-A INTERFACE

I;'J
&.,
I

..,

MAINTENANCE LED'S

r;J

TRANSMIT DATA (D13)
RECEIVER DATA (D12)

REQUEST TO SEND (D14)

CARRIER (D11)
MICRODIAGNOSTIC (D15)

I
I

SIGNAL QUALITY (D1 0)

r;J

SEE TABLE BELOW FOR DESCRIPTION

BC08S-1

'--_ _ _ _ _ _

DESIGNATION

M8207
MICROPROCESSOR

~'-_

.J._ _ _

JIk~ ___ _
~

----- = ----

DESCRIPTION

D13

ON INDICATES DMP11 IS TRANSMITTING A STEADY STREAM OF 1 ·s.

D12

ON INDICATES DMP11 IS RECEIVING A STEADY STREAM OF 1 'so

D14

ON INDICATES THE USYRT IS READY TO TRANSMIT WHEN CTS IS DETECTED.

D11

ON INDICATES CARRIER IS PRESENT AT THE RECEIVER.

D15

ON FOR EXECUTION OF MICRODIAGNOSTICS

D10

ON INDICATES CARRIER PRESENCE AND OFF INDICATES CARRIER ABSENCE.

Figure 2-5

Microprocessor/Line Unit Installation

2-25

MK-1570

VIEW A

NOT USED
IN DMP11

RS-423-A

RS-232-C

c:::::JW8

c:::::JW9
c:::::J W10
c:::J W11
J3

0
W1
c:::::J

c:::::J W13

IT[]

c::::J W14

lliJ

c::::J W15

<
<

c:::J W1 2

c:JW2
c:::::JW3

c:JW4

c:J W16

c:::::JW5
c:::::JW6

c::::J W17
c::::J W18

em
[IT]

c:::J W19
0

c:::::JW7

c:::::JW20

ruJ c:::::JW21

CBIJ
BC55C-10(RS-232-C/RS-423-A) INTERFACE PANEL CABLE
MK-1571

VIEWB

::..

t
~

00 0

~ !.- °00
0

I.:.

BC55B-10 (RS-422-A INTERFACE) PANEL CABLE
MK-2136

Figure 2-6

DMPll Cable Drawings (Sheet 1 of 3)

2-26

VIEW C

FEMALE CONNECTORS
OUT

OUT

D:©l
BC55A-1 0 (INTEGRAL MODEM) PANEL CABLE

MALE CONNECTORS
CONNECTOR PANEL
(FRONT VIEW)
MK-2137

H3258

H3257

TERMINATOR

TERMINATOR
MK-2244

VIEW D

BC55N TWINAX CABLE

ru leflU~~6

-OJ!! [II

I.--P

BC55M TRIAX CABLE

MK-2168

Figure 2-6

DMPII Cable Drawings (Sheet 2 of 3)

2-27

VIEW E

25 PIN
CINCH

BC05D-25 (RS-232-C INTERFACE) MODEM CABLE

25 PIN
CINCH
MK-2139

VIEW F

37 PIN
CINCH

BC55D-33 (RS-422-NRS423-A INTERFACE) MODEM CABLE

37 PIN
CINCH

MK-2187

VIEWG

BC05Z-25 (V.35 INTERFACE) MODEM CABLE
MK-2138

Figure 2-6

DMPll Cable Drawings (Sheet 3 of 3)

2-28

J1
NUll ClK +
REC ClK +
TX ClK +

H_H

NUll ClK REC ClK TX ClK TX DATA R DATA +
TX DATA +

SS
-.
A~

S
...

TER RDY

TXINT

RECINT

+
+

fll3
..
D2
..f

SIDE 1

H3254

TXINT
RECINT

T
....

Y..

DATA MODE

CTS
REC RDY

K...K

R DATA
RTS

T.J

N-N

FF
...;:

M-M

VIEWA

H3254 MODULE TEST CONNECTOR (J1 ON M8203)

000

MK-2143

AUX ClK +
TX ClK DIFF +

y-....

Y..

RX ClK DIFF +

..-

W
:..;:

RX ClK DIFF-

OROTOPOS

TX DATA DIFF +

OVOxOU
OW
OZ
OY
OdObOcoa
OlO 9
OP

RX DATA DIFF +

OBOOOAOC

AUX ClK -

OFOJOEOH

TX ClK DIFF -

OLONOKOM

TX DATA DIFFRX DATA DIFF RTS

onOkomOj

CTS
CDET

000

VIEWB

DTR
DSR

H3250

Figure 2-7

DMPII Turnaround Test Connectors (Sheet 1 of 4)
2-29

•

..
..

--a

..f

-B
-I
§

--C12...
..-f!:!
..
~

MK-2123

J2
~

SEND COMMON
REC COMMON
TER IN SER
INCOMING CALL
TER RDY +

...

REC DATA +
SEND DATAREC DATA-

...

SEND TIMINGREC TIMING'RTS +

...
...
...

CTS +
REC RDY +

....

NEW SIGNAL
SIGNAL QUALITY
SEL SIGNAL RATE
SIGNAL RATE IND

*
*

SEC SEND DATA
SEC REC DATA

*
*

SEC RTS
SEC CTS

LOCAL LOOP
TEST MODE

...

....

REC RDY DATA MODE -

TER RDY-

§.

-R
:.;.

E...E

55
A

y.--

H_H
M-M

+
+

SIDE 1

-...
-Y

R_R

+
+

THl

L.,.L

N_N

TH2

H3255

...P

-Q
K

;..:

VIEW C

M
':':

A-A

.
.....

RTS CTS -

J<..K

C-C

SEL STAND BY
STAND BY IND

f
J...

SEND TIMING +

NULL CLK-

NULL CLK +
REC TIMING +

DATA MODE +
SEND DATA +

....

PP
'
.~

...

U
...

H3255 MODULE TEST CONNECTOR (J2 ON M8203)

NOT REQUIRED FOR DMP11
MK-2122

Figure 2-7

DMPll Turnaround Test Connectors (Sheet 2 of 4)
2-30

J1
1

SHIELD GROUND
SIGNAL RATE INDICATION

SEND DATA +

SEND TIMING +

1j)

REO TO SEND +
REO TIMING +

LOCAL LOOP

DATA MODE +
TERMINAL READY +

RECEIVER READY +
REMOTE LOOP
INCOMING CALL

SIGNAL RATE SEL

TERMINAL TIMING +

TEST MODE

SIGNAL GROUND

REO TO SENDREO TIMING-

CLEAR TO SENDTERMINAL IN SERVICE

DATA MODETERMINAL READYRECEIVER READYSELECT STANDBY
SIGNAL OUALITY
NEW SIGNAL
TERMINAL TIMINGSTANDBY INDICATION
SEND COMMON

Figure 2-7

~
1-2

13
14

•

....

F
~i

<D~

...
1J5
u~

18
...
:.;;;:

(!)i

•
2i
•
~

REO DATA-

-..

SEND TIMING-

~
~

RECEIVE COMMON

SEND DATA-

-~2.

REO DATA +

CLEAR TO SEND +

•2
'3
•1

...

~
26
A

~
is

...
...;.

30
3i-:....
3J
33

H3251

34
....

...

VIEWD

35
3_6

37
H3251 CABLE TEST CONNNECTOR
BC55B BC55C A NO BC550)

DMPII Turnaround Test Connectors (Sheet 3 of 4)
2-31

MK-2121

H325

SCE~24
SCT

SCR

SEC XMIT

SECRCV~

CUT TO TEST
SYNC
NEW

SIDE

•

XMITDATA~

RCVDATA~
RS
CS
CO
NEW SYNC

•• • ••••••• •••
•••••• ••• •••

•

8
14

H325

DATA SET RDY-...._~_6--t
CUT TO TEST*
NEW SYNC

VIEW E

DTR-.....- - - i

RI-"'-~--

H325 CABLE TEST CONNECTOR (BC55C AND BC05D)

Figure 2-7

DMPII Turnaround Test Connectors (Sheet 4 of 4)

2-32

MK-2124

NOTE
When installing panel cables Be55A, Be55B, or
Be55e, it is important that the panel be properly
mounted on the rear mounting rail of the cabinet to
ensure adequate grounding.
When connecting integral modem options, check the BeSSA connector panel to verify that
the appropriate modem jumpers on the panel are properly configured for the option selected.
Table 2-16 lists each of these options and required jumper configurations.

Table 2-16

Modem Option Jumper Functions

& ~

7~'"

!"')

J2 Jumper

1
2
3
4
5
6
7
8
9
10
11

12
13

14
15
16
17
18
19
20
21
22
23
24
25

WI
W7

W19

W14
W3
IN
W2
IN
W6
IN
W20 IN
W5
IN
W18 IN
W17
W4
IN
W16
W13

~
~~

c:-

'" ~
t:
t:
~I:: . .~.#
,Sj
.
8 8
4,~

IN
IN
IN
IN
IN
IN

IN
IN
IN
IN
IN

IN
IN
IN
IN
IN
IN

INt
IN
IN
IN
IN
IN
IN

(,j'"

.Q:"

101

BA
BB
CA
CB
CC
AB
CF

SD
RD
RS
CS
DM
SG
RR

103
104
105
106
107
102
109

SCF
SCB
SBA
DB
SBB
DD

SF
SRR
SCS
SSD
ST
SRD
RT

126
122
121
110
114
119
115
141
120
108
110
140
125
111
112
116
113
117
142

SCA SRS
CD TR
CG SQ
CE
CH
CI

IC
SR
SF
SS

IN
IN

(,j

.Q:"

W21 IN·
W12
•
W15
WI0 IN
Wll
W9
W8

~
~q,j

I
"'V

SB
TM
Make Busy

*RS-232-C defines both signals for this pin
t CCITT modem A only

2-33

Integral modem options require that a 75 ohm terminator be connected to end each transmit
line (BC55A panel) at each end of a full-duplex and a half-duplex network. The terminators
are used to prevent reflections at the end of a transmission line. These terminators are available in both male (H3257) and female (H3258) types to accommodate different integral
modem cabling. Selection of the appropriate terminator type is dependent upon which type
of unused panel connector is available on the receive line at the BC55A-I0 panel. Refer to
Figure 2-8 for DMPll to DMRll/DMCll cabling and to Figure 2-9 for DMPll to
DMPll/DMRll cabling. Various multipoint cabling examples are shown in Figures 2-10, 211, and 2-12.
6.

2.7

Insert the appropriate cable turnaround test connector in the end of the cable. Refer to Table
2-10 for the specific test connector. Load and execute the M8203 static diagnostics specified
in step 4 using the external maintenance mode selected to verify the module and cable. Upon
obtaining a minimum of five error free passes, proceed to the DMPll system test procedures,
Section 2.7. Figure 2-7 shows the various test connectors used in the DMPl1.

DMPII SYSTEM TESTING

The final step in the installation of a DMPll subsystem is to exercise the microprocessor and line unit,
as one complete unit, on the UNIBUS and over the communications link. This is the first testing that
will use the DMPll microcode.

2.7.1

Functional Diagnostic Testing

Ensure that the specific cable turnaround test connector for the selected DMPll option is still installed
at the end of the cable. Load and execute the DMPll functional diagnostics with the external mode
selected.
A.

PDP-II Systems
CZDMT** DMPl1 Functional Diagnostics

VAX-II Systems
EVDMB
REV *. * - VAX D MP 11 Functional Diagnostic

On obtaining a minimum of five error free end passes, proceed to Section 2.7.2.

2.7.2

DECXII System Exerciser

There are two DEC/XII modules for the DMPll: DMD* and DME*. Together these two modules can
operate:
•

Up to 16 DMPI1 devices in loopback, or in point-to-point.

•

A single DMPII configured as a multipoint control station communicating with up to 32
tributaries.

•

Up to 16 devices configured as multipoint tributaries on the same PDP-II UNIBUS.

2-34

RS-232-C INTERFACE
J2
BC08S-1

RS-232-C
MODEM

BC05D-25
J1

CABLE TEST
CONNECTOR
H325

J2 TEST
CONNECTOR
H3255

M8207 MICROPROCESSOR

MK·2131

RS~422~A INTERFACE

J2
BC08S-1

RS-422-A
MODEM

BC55D-33

H3251

'IInm' M8207 MICROPROCESSOR

MK·2132

Figure 2-8

DMPII Remote System Cabling Diagram (Sheet 1 of 2)

BC08S-1

RS-423-A
MODEM

S
BC55D-33
H3251

Jl
H3251

J2 TEST
CONNECTOR
H3255

Il!I!i!D M8207 MICROPROCESSOR

MK-2133

N
I

BC08S-1

VJ
0"1

V.35 INTERFACE
BC05Z-25

Jl TEST
CONNECTOR
H3254

Il!I!i!D M8207 MICROPROCESSOR

n::I
~

CABLE TEST
CONNECTOR
H3250

MK-2134

Figure 2-8

DMPII Remote System Cabling Diagram (Sheet 2 of 2)

Figure 2-8

DMPII Remote System Cabling Diagram (Sheet 2 of 2)

FULL-DUPLEX
BCOSS-1
INSTALL 7SQ
TERMINATOR
CONNECTOR (H3257)

J1
MS200-XX
MICROPROCESSOR

BC08S-1
MK-2129

HALF-DUPLEX
BCOSS-1
INSTALL 7SQ
TERMINATOR
CONNECTOR (H3257)

mmm

M8207 MICROPROCESSOR

MICROPROCESSOR
BC55N-98
BC08S-1
MK-2130

Figure 2-9

DMCII to DMPII Integral (Local) Modem Cabling Diagram for Point-to-Point Network

FULL-DUPLEX

BC55A-l0

INSTALL 75Q
TERMINATOR
CONNECTOR (H3257)

mmm

M8207 MICROPROCESSOR

Jl
M8207 MICROPROCESSOR
ffill!ilJ

*
BC08S-1

BC55A-l0

* FOR 56K bps USE CABLE BC55N-98
FOR 250K bps to 1 M bps USE CABLE BC55M-98
MK-2127

tv
I

HALF-DUPLEX

W
00

BC08S-1
INSTALL 75
TERMINATOR
CONNECTOR (H3257)

mmm

M8207 MICROPROCESSOR

Jl
M8207 MICROPROCESSOR
I!!ffi!J

BC55A-l0

BC08S-1

* FOR 56K bps USE CABLE BC55N-98
FOR 250K bps to 1 M bps USE CABLE BC55M-98
MK-2128

Figure 2-10

Multipoint Cable Routing Example Half-Duplex

HDX NETWORK
IiiiffiiI

DE)

M8207
I1!El!l
MICROPROCESSOR

IllEIEJ

CONTROL
STATION

M8203 LINE UNIT

TERMINATOR
(H3257)

BC55A-10

I
I
I
I
L.._

BC08S-1

G!!iI

DE)

M8207
I1!El!l
MICROPROCESSOR

J1
RCV

XMIT

HDXO

TRIBUTARY

r --oFDXO
BC55A-10

BC08S-1

G!!iI

DE)

M8207
I1!El!l
MICROPROCESSOR

TRIBUTARY

Eii!Eil

DE)

M8207
MICROPROCESSOR§ITI)

TRIBUTARY

BC55A-10

BC08S-1
MK-2433

Figure 2-11 Multipoint Cable Routing Example
Full-Duplex Control Station End Node

2-39

FOX NETWORK

IiiiiiJ

CJI!!!J

M8207
IillliliJ
MICROPROCESSOR
J1
TERMINATOR
(H32S7)

RCV
XMIT
--"""'·"---"'HDXO

@
FDX

I!!ilii!]

CONTROL
STATION

M8203 LINE UNIT
J1 J2
J3

Ii!i!iliJI!!ilii!]

BCSSA-10

BC08S-1

IiiiiiJ

CJI!!!J

M8207

MICROPROCESSOR~
J1

mmm
M8203 LINE UNIT
J3

BCSSA-10

TRIBUTARY

BC08S-1

IiiiD

CJ I!!!J

M8207

MICROPROCESSOR~
J1

TRIBUTARY

BCSSA-10

BC08S-1

IiiiiiJ

NOTES:

CJI!!!J

M8207
mmm
MICROPROCESSOR
J1

SOLID LINE REPRESENTS
CONTROL STATION TRANSMIT TO
TRIBUTARY RECEIVE.
DASH LINE REPRESENTS CONTROL
STATION RECEIVE FROM TRIBUTARY
TRANSMIT.

TRIBUTARY

TERMINATORS
BOTH ENDS OF THE TRANSMIT LINE FROM THE TRIBUTARIES NEED TERMINATION IN ADDITION TO THE ONE
TRANSMIT LINE FROM THE CONTROL STATION
MK·2435

Figure 2-12

Multipoint Cable Routing Example Full-Duplex Control Station End Node

2-40

FDX NETWORK

Wliill

o I!!!]

M8207
mmm
MICROPROCESSOR
J1

TRIBUTARY

BC08S-1

GElI

I
I
I
I
L..

0 I!!!]

M8207
mmm
MICROPROCESSOR
J1
RCV
XMIT
QHDX

CONTROL
STATION

@
QFDX

BC08S-1

L_
GElI

o I!!!]

M8207
mmm
MICROPROCESSOR
J1

TRIBUTARY

BC55A-10

NOTES:
SOLID LINE REPRESENTS
CONTROL STATION TRANSMIT TO
TRIBUTARY RECEIVE.

BC08S-1

ElllI

0 I!!!]

M8207
mmm
MICROPROCESSOR
J1

DASH LINE REPRESENTS CONTROL
STATION RECEIVE FROM TRIBUTARY
TRANSMIT.

TRIBUTARY

TERMINATORS
SINCE THE CONTROL STATION IS NOT AN END MODE, THERE IS NEED TO TERMINATE TWO SETS OF
TRANSMIT LINES FROM THE CONTROL STATION, ADDITIONALLY BOTH ENDS OFTHETRANSMIT LINE FROM
THE TRIBUTARIES
MK·2434

Figure 2-13

Multipoint Cable Routing Example Full-Duplex Tributary Station End Nodes
2-41

2.7.3

Final Cable Connections

The final step in the installation process is to return the DMPII to its normal cable connections, either
to the appropriate modem or to the distribution panel. Cable Routing instructions are included for each
cable in the description column of Table 2-15. The DMPII system cable routing diagrams in Figures 28, 2-9, and 2-13, have been included to help show overall cable routing for the various DMPII-XX
options. References to specified locations of the various test connectors during diagnostic testing are
also included. After the cables are connected to the appropriate modem or distribution panel, it is suggested that the data communications link test program (DCL T) be exercised.

2.7.4

DMPll Link Testing

The DMPII can be exercised over a communications link by the data communications link test
(DCL T). It is suggested that DCLT be configured to run first on a cable test connector and then on a
modem with the modem analog loopback test feature selected, if the modem includes this feature.
Next, the overall communications link should be exercised with the remote computer systems that contain DMPIIs and/or DMRIIs or a DMCII.
PDP-II DCLT is CZCLM**
VAX DCLT is EVDMD**

2-42

CHAPTER 3
COMMAND AND RESPONSE
STRUCTURES
3.1 INTRODUCTION
The interaction between a user program and a DMPII includes the issuing of commands to the DMPII
and the processing of responses from the DMPll. Specific DMPII responses are solicited by issuing
commands. Other DMPII responses are unsolicited and result from the detection of various hardware
and software events and error conditions and changes to the operational status of tributaries.
Each DMPII command and response has a highly structured format, where the issuing of commands
by the user program and responses by the DMPII requires that a specific programming sequence be
followed. This chapter defines these formats in detail and describes the relationship of commands and
responses to the operation of a DMPII in the network environment.
3.2 COMMAND/RESPONSE STRUCTURE
The DMPII is a nonprocessor request (NPR) device residing on a VAX/PDP-II UNIBUS. Communication between a DMPII and the user program is done through the UNIBUS control and status registers (CSRs). DMPII CSRs are assigned the symbolic byte addresses BSELO through BSEL 7, and
the four symbolic word addresses SELO, SEL2, SEL4, and SEL6. Table 3-1 lists the symbolic byte and
word addresses assigned in the VAX/PDP-II input/output (I/O), page floating address space along
with the octal addresses.
NOTE
Each letter x in the I/O page addresses listed in
Table 3-1 represents an octal value that is switch assigned at each DMPl1 in a network at installation
time.
The relationship between the symbolic byte and word addresses for DMPII CSRs and the actual CSR
layout is shown in Figure 3-1. Figure 3-2 shows the fields in CSR bytes BSELO, BSEL2, and BSEL3
that make up the fixed format portion of both user program commands and DMPII responses. This
fixed format portion serves to identify the command/response type, the tributary address that the command/response applies to, and to coordinate ownership of the CSRs between the DMPII and the user
program.
The four bytes that make up SEL4 and SEL6 contain the fields pertinent to each user program command and DMPII response. In-depth descriptions of the SEL4 and SEL6 fields making up each user
program command and DMPll response are presented in the sections which follow.
3.3 USER PROGRAM COMMANDS
A user program has a set of four commands that it can issue to a DMPII for execution by that device.
These are:
I.
2.
3.
4.

Microprocessor control,
Mode definition,
Control, and
Buffer address/character count
3-1

Table 3-1

Symbolic Byte and Word Address

Symbolic Address

Octal Value

SELO

76xxxO

BSELO

76xxxO

BSELI

76xxxl

SEL2

76xxx2

BSEL2

76xxx2

BSEL3

76xxx3

SEL4

76xxx4

BSEL6

76xxx4

BSEL5

76xxx5

SEL6

76xxx6

BSEL6

76xxx6

BSEL7

76xxx7

8 7
BSEL1

BSELO

SELO

BSEL3

BSEL2

SEL2

BSEL5

BSEL4

SEL4

BSEL7

BSEL6

SEL6

MK·1635

Figure 3-1 Relationship Between DMP11
C SR' s Byte and Word Symbolic Addresses

3-2

6
I

5
I

BSEL3

3
I

, 1

2
I

lEO

I lEI

BSELO

RDO

RDI

COMMAND TYPE
CODE

BSEL2
SEL2

BSEL5

COMMAND/RESPONSE
FIELDS
I

BSEL7

COMMAND/RESPONSE
FIELDS
1
I

o
I

TRIBUTARY ADDRESS

, 2

RQI

BSEL4
SEL4
I

BSEL6
SEL6

o
MK·1636

Figure 3-2

Command and Response Fixed and Variable Formats

BSEL1

MK·1637

Figure 3-3

Bit Layout of Initialization Register

With the exception of the microprocessor control command, which is used to set up the DMPII initialization register and involves only BSEL 1 of the CSRs, user program commands need an identification code in the first three bits of BSEL2 (see Figure 3-2). These codes identify each command and
are listed in Table 3-2.
3.3.1 Microprocessor Control Command
This single-byte command has two functions:
1.
2.

To initialize and start the DMPII running, and
To execute the maintenance functions implemented by the DMPII.

At startup time, under typical operating conditions, this command is the first issued by the user program.
The format for the DMPII initialization register (BSELI) is shown in Figure 3-3. To set the master
clear bit and start the DMPII running, the user program moves a byte having an octal value of 100 to
BSELI. As a result, all condition sensitive logic in the DMPII is reset for startup and the internal
startup diagnostic is executed. The internal startup diagnostic indicates the result of execution by placing a code in BSEL4 and BSEL6 of the CSRs to be read by the user program. The codes returned by
the internal diagnostic are listed and defined in Table 3-3.
When the diagnostic completes successfully, as indicated by the value 77 in BSEL4 and 305 in BSEL6,
the Run bit in BSEL 1 is set to one.
Figure 3-4 presents a flowchart describing how to set master clear. A timeout counter is set to avoid an
endless loop catching the user program in the event that the internal diagnostic does not complete successfully. A timeout counter is necessary because this process does not generate an interrupt. This timeout counter should be set to a value that results in a delay of from one-half to one second.
3-3

Table 3-2

User Program Command Codes

Command Name

Binary Code

Mode Definition

010

Control Command

001

Buffer Address/Character
Count Command (Receive Buffer)

000

Buffer Address/Character
Count Command (Transmit Buffer)

100

Table 3-3

DMPll Internal Diagnostic Execution Codes (Octal)

BSEL4

BSEL6

Meaning

305

The internal diagnostic executed successfully.

(Random)

143

The internal register test failed; the microcode is in a loop.

100

135

The microprocessor data RAM test failed; the microcode is in
a loop.

100

125

The microprocessor ALU test failed; the microcode is in a
loop.

100

264

The line unit test failed; the microcode is in a loop.

100

305

The line unit test failed; however, the microprocessor is running. When returned, this code usually indicates that the electrical cable connecting the line unit to the microprocessor disconnected. In this circumstance, the microprocessor must be
running in order to troubleshoot the line unit.

As shown in Figure 3-3, bits zero through four of BSELI are used to perform maintenance and diagnostic functions on the M8207 microprocessor and the M8203 line unit. When these maintenance bits are
used by a programmer through the main CPU console or by a system diagnostic running in the main
CPU, both the run bit and the master clear bit must be set to zero. The only exception to this is the LU
loop bit which is not affected by the state of the run bit.
3.3.1.1 Microprocessor Control Command Field Descriptions - Figure 3-3 shows that the DMPl1 initialization register (BSEL 1) has eight one-bit control fields, with three used for startup in normal operation and five for microcode diagnostic and maintenance purposes. These five bits are used by diagnostic programmers and generally have no use in the development and operation of DMPll user programs.
One of the switch controlled features of the DMPII is the enabling and disabling of the five diagnostic/maintenance bits by the BSELI lockout switch (see Section 2.5.2). When the BSELI lockout
switch is ON, the various maintenance bits can be used for diagnostic and maintenance purposes. When
OFF, all diagnostic and maintenance functions performed by these bits are disabled. The functions performed by each of the BSELI single bit fields are described in detail below.
3-4

SET MASTER
CLEAR BIT
IN BSEL1
(BSEL= 1 OOs )

SET
TIME OUT
COUNTER
TO
~ 0.5 SEC

DIAGNOSTIC
ERROR. EXIT
TO ERROR
CONTINGENCY

EXIT TO
COMPLETE
START UP

MK-2697

Figure 3-4 Flowchart for a Routine to Initialize the DMPII

3-5

Master Clear (Bit Six of BSELl)
This bit controls the DMPII initialization process. When set by the user program, all condition sensitive logic in the microprocessor and the line unit is cleared. In addition, the microprocessor program
counter is cleared, the microprocessor clock is enabled, and the microprocessor starts to run and begins
to execute the internal diagnostic. Since master clear is a write-only bit, the user program can only set
this bit.
In addition, when the BSELI lockout switch (see Section 2.5.2) is in the OFF position, the master clear
bit is the only bit in BSELI that can be set by the user program. This feature prevents an inadvertent
setting of the maintenance/diagnostic bits by the user program.
Run (Bit Seven of BSEL 1)
Run is a flag bit that serves to inform the user program that the DMPII is ready for configuration and
normal operation. Execution of the internal diagnostic routine is started by setting master clear. When
the internal diagnostic routine successfully executes, it sets the run bit, clears the master clear bit, and
sets BSEL4 and BSEL6 to the code for successful completion (see Table 3-3). At this point, the microcode is running and the DMPII is ready to accept startup and operational commands and to issue responses.
In normal operation, the run bit serves to flag only the transition of the microprocessor from off to the
running state, not from running to off. With the BSELI lockout switch OFF, the run bit can only be
read. However, with this switch ON, the run bit can be set and cleared by the user program.

NOTE
Clearing the run bit halts the M8207 microprocessor.
Step Microprocessor (Step MP, Bit Zero, BSELl)
Setting the step microprocessor bit causes the microprocessor to perform one instruction cycle to execute the CROM microinstruction addressed by the program counter. To use this bit, the BSELI lockout switch must be on. The step microprocessor function is useful only for maintenance and diagnostic
operations.
Control ROM Input (ROM I, Bit One, BSELI)
This bit permits a microinstruction to be entered through the CSRs for execution by the microprocessor. In actual practice, the ROM I bit is set and the microinstruction is placed in SEL6 of the
CSRs (see Figure 3-2,). Step MP is then set and the instruction in SEL6 is executed as if it were a
microinstruction in the M8207 control ROM.
If the microinstruction entered in SEL6 is a move class instruction, the microprocessor program
counter is increased by one. However, if the microinstruction executed is a branch class instruction, the
program counter is set to the CROM address branched to by the microinstruction.

NOTE
It is good programming practice to clear the ROM I
bit immediately after the microinstruction in SEL6
is executed.

To use this bit, the BSEL 1 lockout switch must be ON. The function performed by the ROM I bit is
useful only for maintenance/diagnostic operations.

3-6

Control ROM Output (ROM 0, Bit Two, BSELI)
Setting this bit places the contents of the CROM location addressed by the program counter in SEL6
of the CSRs. The contents of any CROM location can be examined in SEL6 by using the ROM 0 bit
and by executing an unconditional branch instruction to the desired location using the ROM I bit. This
sets the microprocessor program counter to address this location. Then, by setting the ROM 0 bit, the
contents of that location are moved to SEL6 for examination.
The general sequence listed below describes the use of the ROM I and ROM 0 bits for examining
CROM locations:
1.

Set ROM I;

Move the appropriate branch class instruction to CSR SEL6;

Clear ROM I; (If an additional microinstruction is needed to set the program counter to the
desired CROM address, repeat steps one to three.);

Set ROM 0;

Read information from CSR SEL6; and

Clear ROM O.

To use these bits, the BSELI lockout switch must be ON. The function performed by the ROM I and
ROM 0 bits is useful only for maintenance and diagnostic operations.
Line Unit Loop (LU Loop, Bit Three, BSELl)
This bit is used to implement internal loopback testing without the necessity for a loop back cable or
connectors. Serial output prior to the transmitter level conversion logic is fed back to the receiver serialto-parallel conversion logic so that internal logic levels are maintained throughout the feedback loop.
When the LU loop bit is set in conjunction with the run bit being cleared, data is clocked through the
transmit/receive loop at the rate the step LU bit is toggled. However, if the LU loop bit is set in conjunction with run being set, data is clocked through the transmit/receive loop at the internal maintenance clock rate which is 48K b/s. Clearing the LU loop bit halts the internalloopback process.
To use this bit the BSELI lockout switch must be ON. The functions performed by the LU loop bit are
useful only during maintenance/diagnostic operations.
Step Line Unit (Step LU, Bit Four, BSELI)
Step LU is used in conjunction with the LU loop bit to single-step through the transmit/receive loop.
With the LU loop bit set, and the run bit cleared, each setting of step LU causes one transmit bit to be
shifted for loopback. In the same manner, each clearing of step LU shifts the transmitted bit for reception. By alternately setting and clearing this bit, a succession of bits can be transmitted and looped
back, one at a time, for reception.
Invoke Primary MOP Mode (BOOT, Bit Five, BSEL 1)
When set to one, this bit causes the DMPll at this multipoint station to request that the control station
start the primary maintenance operation protocol (MOP) boot procedure. In point-to-point networks, a
DMPll with this bit set requests the other station to start the primary MOP boot procedure.

3-7

NOTE
Master clear must be set at the requesting DMPll
in order to use the primary MOP boot procedure.
3.3.2 Mode Definition Command
The mode definition command is used to create both the hierarchy of a network and the characteristics
of the communications line serving that network. As shown in Figure 3-5, the mode definition command
has two fields: the command type code field in BSEL2 (which is common to all DMPII commands)
and the mode field in BSEL6 (which contains a code defining the function to be performed by the
command).
With the mode definition command, the user program can name a DMPII as a control tributary station
in a multipoint network or as a station in a point-to-point network. In addition, the characteristics of the
physical communications line connecting a network can be defined.
The actual mode field codes and the functions implemented by each code are listed in Table 3-4.
Under typical operating conditions, the mode definition command is issued by the user program at startup time after the internal diagnostics are executed successfully (see Table 3-3) and the run bit is set.
Network discipline needs the user program associated with each DMPII in the network to issue a mode
definition command. For example, in a half-duplex multipoint network compiled solely of DMPlls:
1.

The user program at the control station issues a mode definition command having the mode
field set to four (control station half-duplex).

The user program at each tributary station issues a mode definition command with the mode
field set to six (tributary station half-duplex).

This network discipline also applies to DMPlls operating in point-to-point networks with other
DMPlls or DMClls.
The functions performed by the mode definition command can also be implemented by the mode selection switches on the DMPII line unit module (see Section 2.6). The mode definition function at a tributary station having a switch-assigned tributary address must be done using these switches. However,
when tributary addresses are software assigned, the mode definition command must be used at the control and tributary stations to configure the network and assign line characteristics. The switch settings
for performing the mode definition functions compare with the BSEL6 codes listed in Table 3-4 (see
Section 2.6 and Table 2-10).
3.3.3 Control Command
This command provides the primary device for controlling the operation of DMPII-implemented networks. The format for the control command is shown in Figure 3-6. At startup time, after issuing the
microprocessor control and the mode definition commands, the user program at the control station
DMPII must issue a control command for each tributary address supported in the multipoint network.
Also, the user program at each tributary station must issue a control command for each tributary to be
established at that station. At a control station, these commands result in the creation of a tributary
status slot (TSS) in the microprocessor data memory for each tributary address in the network. At a
tributary station one TSS is created for each tributary address created at that station.
The DMPlls at the control station and at all tributary stations use these TSS structures to coordinate
protocol operation over the network between the control station and a tributary. User programs at control and tributary stations access these structures to obtain functional information, such as counts of
messages transmitted and received, selection interval counts and timeouts, reply timeouts, and pertinent error counts.

3-8

RDO

lEI

BSELO

1 I 0
0
CMD .TYPE fODE

BSEL2
SEL2

TRIBUTARY ADDRESS

BSEL3

lEO

RQI

RDI

BSEL5

.
I

BSEL4
SEL4

BSEL7

MODE FIELD
I

Figure 3-5

MK-1639

Mode Definition Command Format

Table 3-4

Mode Field Codes and Functions

BSEL6 Bit
Positions
2 1 0

Line
Characteristics

Network
Configuration

DMCll
Compatibility

Half-Duplex

Point-to-Point

Yes

Full-Duplex

Point-to-Point

Yes

Half-Duplex

Point-to-Point

Full-Duplex

Point-to-Point

Half-Duplex

Control Station

N/A

Full-Duplex

Control Station

N/A

Half-Duplex

Tributary Station

N/A

Full-Duplex

Tributary Station

N/A

6
I

BSEL3

5
I

2
I

RDI

RDO

TRIBUTARY ADDRESS

lEO

RQI

2
I

BSELO

1
0
0
CMD TYPE CODE

BSEL2
SEL2

BSEL7
15

,I
13

I
12

I
11

WRI
TSS

I I

READ/I READ
CLR
TSS
TSS
5
6

REQUEST KEY OR
TSS ADDRESS
4

LATCH
POLLING
STATE
UNLATCH
POLLING
STATE

BSEL4
SEL4

DATA

o
lEI

BSEL5

BSEL6
SEL6

DISABLE
COMMON
BUFFER
POOL
ENABLE
COMMON
BUFFER
POOL

Figure 3-6

Control Command Format

3-9

BSEL6
SEL6
0

At the startup of tributary stations having multiple software-assigned tributary addresses, the user program at each station issues as many control commands as there are tributaries created at that station.
However, in networks where tributary addresses are switch-assigned, only one control command specifying the switch-assigned address need be issued as part of the startup sequence. Also, at a tributary
station with the tributary address assigned by means of switches, any non-zero tributary address in a
control command is overridden by the switches.
In point-to-point networks, a control command having the tributary address field set to one must be issued
at both stations. This results in the creation of a single TSS structure at each station.
The user program at the control station has access to parameters for controlling the operation of the
polling algorithm. It also has the option of specifying a unique set of polling parameters for each tributary in the network by issuing control commands.
During a typical operation, the control station, through the polling algorithm, determines the polling
level of any tributary in the network and adjust the polling frequency of that tributary as necessary (see
Section 5.2). Based on the short-term polling response history of a given tributary, that tributary can be
placed in one of four polling frequency levels:
1.

Active - a tributary is active when it responds to most polls with data messages.

Inactive - a tributary becomes inactive when it responds to a consecutive number of polls
with control messages instead of data messages. This number is a user-defined parameter
referred to as the number of no-data messages needed to go to inactive (NDM-INACT).

Unresponsive - a tributary becomes unresponsive when it does not respond to a consecutive
number of polls. This number is a user-defined parameter referred to as the number of timeouts to go unresponsive (TO-UNRESP).

Dead - a tributary becomes dead when it continues to be unresponsive for a consecutive number of polls after a TO-UNRESP timeout. This number is a user-defined parameter referred
to as the number of timeouts to go to dead (TO-DEAD). All dead tributaries are polled at a
rate designated by the user-defined polling parameter DEAD T (see Section 5.2.3.4).

3.3.3.1 Control Command Field Descriptions - The Control Command permits the user program to
perform a number of control functions using a single command format. In general, each function implemented by this command needs the issuing of a single correctly formatted control command. Exceptions to this are identified in the descriptions for the field or fields implementing each command function. These fields are described in detail in the sections that follow.
Enable/Disable Common Pool Buffers (Bits 0 and 1 of BSEL 7)
In multipoint networks, user programs at control and tributary stations have the ability to set up a common pool of receive buffers and to selectively enable and disable access to buffers in this pool. Buffers
in this common pool are allocated on a quota basis with each tributary assigned a specific quota. A
common buffer pool is created by the user program through the buffer address/character count command (see Section 3.3.4). These control bits apply only to multipoint networks.
Enable Common Pool

When set to one, this control bit enables the use of receive buffers
from the common buffer pool for the tributary designated by the
address in BSEL3. In addition, the quota for that tributary is set by
the octal value in BSEL4. If the enable common pool bit is set and
the value in BSEL4 is zero, the effect of setting this bit is nullified.

3-10

If the octal value placed in BSEL3 is 377, the buffer quota available to the tributary is without a limit. During network operation,
when a common pool receive buffer is completed by a tributary, the
quota for that tributary is decremented by one. The user program
can increase the quota for a tributary by issuing a control command
with the enable common pool bit set and a value in BSEL4. However, in this event, the value in BSEL4 is added to the current quota
to establish the new quota for that tributary. If the increase causes
the quota for that tributary to equal or exceed the octal value 377,
the quota is reset to 376 and a procedural error is posted to the user
program (see Section 3.4.2.4).

When set to one, this bit disables the use of common pool receive
buffers by the tributary addressed by BSEL3. In addition, the
quota previously created for that tributary is reduced to zero.

Disable Common Pool

Latch/Unlatch Polling State (Bits 4 and 5 of BSEL 7)
In DMPl1-implemented multipoint networks, the control station DMPII uses the microcode polling
algorithm to automatically poll each tributary at a frequency determined by the current polling level
and user-defined polling parameters. In addition to specifying polling parameters for each tributary by
means of the control command (see the write TSS function), the user program at the control station can
also change the polling level of a given tributary. That tributary then stays at the designated polling
level, regardless of its ongoing response history, until the user program returns polling control of that
tributary to the polling algorithm.
Latch Polling State

When set to one, this bit causes the polling level of the tributary
addressed by BSEL3 to be set according the binary code specified
by bits 0 and 1 of BSEL4. The relation between the coded value of
these bits and polling levels is defined in Table 3-5.

Unlatch Polling State

When set to one, this bit causes the polling level of the tributary
addressed by BSEL3 to be restored to polling algorithm control at
the active level. During operations which follow, the polling algorithm adjusts the tributary to a polling level determined by its
short-term response history.

Table 3-5 Correspondence Between Polling Levels
and the Binary Code in BSEL4, Bits 0 and 1
Binary Code

Polling Level
Active

Inactive

Unresponsive

Dead

3-11

Request Key Field (Bits 0 to 4 of BSEL6)
The codes in the request key field provide for the control functions related to the startup and operation
of DMPlls by the user programs at all stations. When the request key field is used, bits five through
seven of BSEL6 must be cleared. The octal code for each control function is listed in Table 3-6. Each of
the control functions listed in Table 3-6 are defined in detail below.
No Request

This code allows the issuing of a null control command for the purpose of returning control of the CSRs to the DMPll when RDI is
set and there is no command to issue. The effect of using this
request key is to issue a NOP command. Note that the enable/disable common pool and latch/unlatch polling state bits can
be used with this request key.

Establish Tributary

This control function allows user programs at all stations to start
creation of the tributary status slot (TSS) data structure. To establish tributary addresses in a mUltipoint network the user programs
at both the control and tributary stations must issue one establish
tributary control command for each tributary supported. In these
commands. the establish tributary code must be set in the request
key field, with the correct tributary address set in BSEL3.
NOTE
In a point-to-point network, this control command,
with a tributary address of one in BSEL3, should be
issued at each station to create the needed TSS.
At the control station, the user program can create a TSS structure
for up to 32 tributary addresses. In addition, the total number of
tributary addresses created and maintained by a mUltipoint control
station at any given time cannot exceed 32. Correct network discipline states that tributary addresses created at the control station
also be created at the tributary stations.
As a result of creating one or more tributary addresses at a station,
the DMPII automatically creates a global status slot (GSS) as part
of the internal data structure at that station. The GSS maintains the
status of DMPll operation and specific data that is global to the
device. The GSS is accessible to the user program by specifying a
tributary address of zero in BSEL3 of the accessing control command.
NOTE
This control command function can also be used
during the typical operation of a multipoint network
to recreate previously deleted tributaries. In such
events, the tributary address must be reestablished at
both the control station and the pertinent tributary
station.

Delete Tributary

This control command function allows the user programs at both
control and tributary stations to remove a tributary from functional
status by deleting the TSS. Prior to issuing this command, the user
program must first halt the tributary being deleted (see request key
request halt state). Although a DMPII-implemented multipoint

3-12

Table 3-6

Octal Codes for Request Key Control Functions

Control Function

Octal Code

No Request

Establish Tributary

Delete Tributary

Request Startup State

Request Maintenance State

Request Halt State

Reserved

Read Modem Control

Write Modem Control

Run Interface Diagnostic

Write Modem Test Register

Reserved

37
network can support a maximum of 32 tributary addresses, the tributary address space provided by the DMPII CSR BSEL3 allows
up to 255 tributary addresses. As a result, it is possible for the tributary stations in a network to support more than 32 tributary addresses, with the restriction that a maximum of 32 tributaries can
be created at anyone time.

Request Startup State

This control function initializes the tributary addressed by BSEL3
and starts the the DDCMP startup sequence for that tributary (see
Appendix A). The user program is informed by a control response
when the startup sequence is complete (see Section 3.4.2.2). When
the control response, tributary in run state, is received by the user
programs at both the control and tributary stations, then message
traffic can start between the control station and that tributary. Before issuing control command with this request key to an assigned
tributary, the tributary must be in the DDCMP halt state. If this
command is issued to a tributary not in the halt state, a procedural

3-13

error occurs (see Section 3.4.2.4). A control command with this
function can be issued by the user program at any time after the
tributary is established.
NOTE
The enable/disable common pool and/or the
latch/unlatch polling state bits in BSEL 7 can be
combined with this control function in a single control command.
Request Maintenance State

This control function places the tributary addressed by BSEL3 into
the DDCMP maintenance state. A tributary placed in this state can
only transmit and receive maintenance messages. In order for maintenance message traffic to occur between a control station and a tributary, the control and tributary station must be in the maintenance
state. Issuing a control command with this control function initializes the TSS for the pertinent tributary. Before issuing a control
command with this request key the tributary must be in the
DDCMP halt state. To return to the usual operation from the maintenance state, a tributary must first be in the halt state. Then the
request startup state control command can be issued.

NOTE
The enable/disable common pool and/or the
latch/unlatch polling state bits in BSEL7 can be
combined with this control function in a single control command.
Request Halt State

This control command function places the tributary addressed in
BSEL3 into the DDCMP halt state. When a tributary is halted at
the control station, that tributary is no longer polled. When a tributary is halted at the tributary station, that tributary no longer responds to polls transmitted from the control station. In the process
of halting a tributary at both the control and tributary stations, all
private receive buffers allocated to that tributary are returned.
However, the TSS for that tributary at both control and tributary
stations remains the same. A tributary in the halt state can be restarted by issuing a control command containing the request key for
request startup state.

NOTE
The request key request halt state can be used
globally to return all common pool buffers at the
control and tributary stations. However, the common buffer pool quota assigned to each tributary remains the same. The user programs at the control
and tributary stations must issue Control commands
with the code for request halt state (05) with a tributary address of zero in BSEL3. If a common pool
buffer is in use when this global command is issued,
a procedural error results (see Section 3.4.2.4). If
the common buffer pool is empty when this command is issued, no buffers are returned.

3-14

Read Modem Control

The user program issues this control function to read the contents of
the DMPII line unit registers 13 and 17. The DMPll responds to
this command by issuing an information response with the contents
of line unit registers 13 and 17 in SEL4 (see Section 3.4.3 and Appendix C). The functions of the eight bits comprising each of these
registers are described in detail in Appendix C.

Write Modem Control

The user program issues this control function to write the contents
of BSEL4 (modem control information) into the DMPll extended
line unit register 13 (see Appendix C).

Run Interface Diagnostic

This control function starts execution of the DMPII interface diagnostics. These diagnostics test the CSR interface and related data
paths between the DMPII and the user program. These routines,
which are designed for use with a cooperating diagnostic program
executed out of main memory, are detailed in Appendix E. (This
request key is designed for diagnostic use only.)

Write Modem Test Register

The diagnostic programmer issues this control function to write the
contents of BSEL4 (modem test information) into the extended line
unit register AX3-1S. This register is a single byte register (See Appendix C). (This request key is designed for diagnostic use only.)

Access Tributary Status Slot and Global Status Slot (BSEL6)
Figure 3-6 shows that the control command field defined by bits zero to four of BSEL6 have either a
request key code or provide the user program with access to TSS and GSS locations. Using this field in
a control command for implementing a request key control function needs bits five through seven of
BSEL6 cleared to zero. However, when using BSEL6 in a control command to access a specific TSS or
GSS location, one of the BSEL6 bits five through seven must be set to one. These bits specify the type
of TSS/GSS access.
These bits specify that the TSS/GSS location addressed by BSEL6, bits zero to four, be either written
into, read from, or be read and cleared. The TSS accessed through these control command fields is
defined by the tributary address in BSEL3. The GSS is accessed in the same way except that a tributary address of zero is used. Both the TSS and GSS are addressed on word boundaries so that each access
results in two bytes being read or written. The detailed functions performed by the BSEL6 control bits
five through seven along with the name and octal address of the specific TSS/GSS locations accessed
are shown below.

NOTE
TSS and GSS addresses are specified in octal with
all data to be written. Also, data read from a TSS or
GSS is in octal.
Write TSS/GSS

This bit (bit seven, BSEL6), when set to one, enables the user programs at all stations to write into the TSS or GSS locations addressed by bits zero to four of BSEL6. The control command to
write into a TSS or GSS location must have the octal address of
that location in bits zero to four of BSEL6. The address of the tributary whose TSS is being accessed is included in BSEL3 of the
command. When writing into a GSS, the contents of BSEL3 (tributary address) must be zero. In both events, the data to be written
is in BSEL4 and BSELS of the control command. There are 13 TSS
parameters that can be written by a control command (see Table 41).

3-15

NOTE
Some parameters are eight bits in length with two
included in a single TSS location.
If the user program tries to write into any other TSS location, a procedural error is generated (see Section 3.4.2.4). The names of these
parameters, the TSS octal addresses, and the octal values of BSEL6
are listed below:

Transmit delay timer (XDT) preset value: address = 30 and
the value of BSEL6 = 230.

Initial polling urgency (Q) and polling rate (R) for active
state: address = 31 and the value of BSEL6 = 232.

Initial polling urgency (Q) and polling rate (R) for inactive
state: address = 32 and the value of BSEL6 = 232.

Initial polling urgency (Q) and polling rate (R) for
unresponsive state: octal address = 33 and the octal value of
BSEL6 = 234.

Number of selection intervals during which no data
messages were received, causing a tributary to go inactive
(NDM-INACT) and number of time outs to go unresponsive
(TO-UNRESP): address = 34 and the value of BSEL6 =
234.

Number of time outs to go dead (TO-DEAD) and maximum
message count: address = 35 and the value ofBSEL6 = 235.

Selection interval timing counter: address = 36 and the value
of BSEL6 = 236.

Babbling tributary timing counter: address = 37 and the
value of BSEL6 = 237.

There are five GSS parameters that can be written by a control
command (see Section 4.3.2. and Table 4-2). If the user program
tries to write into any other GSS location, a procedural error occurs
(see Section 3.4.2.4). The names of these parameters, the ass octal
addresses, and the octal values of BSEL6 are listed below:
1.

2.
3.
4.
5.

Number of sync characters to precede non-abutting
messages (NUM SYNC): address = 33 and the value of
BSEL6 = 233.
Preset value for carrier with timing counter: address = 34
and the value of BSEL6 = 234.
Polling urgency update interval (DELTA T): address = 35
and the value of BSEL6 = 235.
Polling rate for dead tributaries (DEAD T): address = 36
and the value of BSEL6 = 236.
Poll delay timer (POLL DELAY): address = 37 and the
value of BSEL6 = 237.
3-16

Read and Clear TSS/GSS

This bit (bit six of BSEL6), when set to one, allows the user programs at all stations to read and clear specific locations in a TSS
and GSS. If a user program tries to read and clear any other than
the specified TSS and GSS locations, a procedural error is generated (see Section 3.4.2.4). The control command to read and clear a
TSS or GSS location must include the octal address of the location
within the TSS or ass being accessed in bits zero to four of
BSEL6. The address of the tributary assigned to the TSS is in
BSEL3. However, if a ass location is to be read and cleared, the
contents of the tributary address field, BSEL3, must be zero. When
the DMPII receives a control command to read and clear a TSS or
ass location, it passes the information is passed, not the request to
the user program through an Information response. The TSS and
ass locations accessible to the user program for reading and clearing are made up of local and global error counters and supporting
statistics. For the TSS, the names of these counts, the octal addresses, and the octal values of BSEL6 are listed below:

Number of data messages transmitted: address = 07 and the
value of BSEL6 = 107.

Number of data messages received: address = 10 and the value
of BSEL6 = 110.

Number of selection intervals completed: address = 11 and the
value of BSEL6 = 111.

Number of data errors outbound and associated error types:
address = 12 and the value of BSEL6 = 112.

Number of data errors inbound and associated error types:
address = 13 and the value of BSEL6 = 113.

Number of local buffer errors and associated error types:
address = 14 and the value of BSEL6 = 114.

Number of remote buffer errors and associated error types:
address = 15 and the value of BSEL6 = 115.

Number of selection interval time outs and associated error
types: address = 16 and the value of BSEL6 = 116.

Number of local reply timeouts and associated error types:
address = 17 and the value of BSEL6 = 117.

There are four station error counts maintained in the ass that can
be read and cleared through a control command and each of these is
described in detail in Section 5.3.l.3. The names of these counts,
the ass octal addresses, and the octal values of BSEL6 are listed
below:
1. Remote station error counter and associated error types:
address = 15 and the value of BSEL6 = 115.

Local station error counter and associated error types: address
16 and the value of BSEL6 = 116.

3-17

Global header block-check error count and maintenance data
field block-check error count: address = 17 and value of
BSEL6 = 117.

When set, this bit (bit five of BSEL6) permits user programs at all
stations to read any location in either the TSS or the
The control command to read a TSS or
location must contain the octal
address of the location within the TSS or
in bits zero to four of
BSEL6. The address of the tributary with the TSS being accessed
is in the tributary address field (BSEL3). However, if a
location is to be read, the contents of BSEL3 must be zero. When the
DMPll receives a control command to read a TSS or GSS location, it passes the information to the user program through an information response (see Section 3.4.3). The word boundary octal addresses for TSS and
locations are in the range 0 to 37. The
corresponding octal values for BSEL6 of a control command to read
these TSS and
locations range from 40 for location zero to 77
for location 37.

Read TSS/GSS

ass.

ass

3.3.4 Buffer Address/Character Count Command
In DMP II-implemented networks, user programs at all stations control the allocation of transmit and
receive buffers through the buffer address/character count command. This command has two forms to
help separate management of transmit and receive buffers. These two forms are identified by the type
code in bits zero to two of BSEL2. The format for the buffer address/character count command is
shown in Figure 3-7.
7
6

.
.

BSEL3
BSEL5
BSEL7

5
I

, 3

2
I

1
I

, 5

RQI

ROO

RDI

B/A
17

I I

BUFFER ADDRESS
HIGH BYTE

B/A
16

CHARACTER COUNT
~IGH SIX BITS
12

lEO

TRIBUTARY ADDRESS

BSELO
BSEL2
SEL2

BSEL4
SEL4

CODE~

CHARACTER COUNT
I LOW BYTE I

lEI

CMD TYPE

BUFFER ADDRESS
LOW BYTE

BSEL6
SEL6

*TYPE CODES
BUFFER ADDRESS/CHARACTER COUNT COMMAND - RECEIVE
000
BUFFER ADDRESS/CHARACTER COUNT COMMAND - TRANSMIT = 100

Figure 3-7

Buffer Address/Character Count Command Format

The type code for commands to allocate receive buffers is zero, and for commands to allocate transmit
buffers, the type code is four. Except for the type code field, the command formats for allocating receive and transmit buffers are exactly the same.
When a user program has a message to transmit, it issues a buffer address/character count command to
inform the DMPII of the address and size of the message buffer to be transmitted, with the tributary
address of the receiving station. For a DMPII in a point-to-point network, the receiving station address
is always one. In all networks, a zero in the tributary address field (BSEL3) of a buffer address/character count command for a transmit buffer will result in a procedural error (see Section
3.4.2.4).
In multipoint networks, user programs at both the control and tributary stations can handle the allocation of receive buffers in two ways:

3-18

The first method is the allocation of receive buffers from a common pool of buffers. With the
common buffer pool enabled (see Section 3.3.3.1), receive buffers are assigned to the pool
through the buffer address/character count command, based on one buffer for each command issued. Each buffer address/character count command used to assign a buffer to the
common pool must have a zero in BSEL3. Although this command assigns buffers to a common pool, actual allocation to a tributary is done through the control command by enabling
access to the pool and assigning quotas (see Section 3.3.3.1).

With the second method, the user program can directly allocate private receive buffers based
on expected message traffic using one buffer address/character count command for each private buffer allocated. In this way, a private buffer is defined as a receive buffer assigned to a
specific tributary for its exclusive use. In all events, the address of that tributary must be in
BSEL3 of the assigning command.

Private buffers and buffers from a common pool can be used together at control and tributary stations
in a multipoint networks. Under these conditions, the benefits provided by both methods are available
to the user program. For example, common pool buffers can be set up for messages that are not expected and/or for large messages, although private buffers can be allocated for expected receive message traffic and/or receive messages of optimum size.
3.3.4.1 Buffer Address/Character Count Command Field Descriptions - Figure 3-7 shows that the value of the type code field (bits zero through two of BSEL2) serves to identify the transmit and receive
buffer address/character count commands. Except for the type code field, the other fields are identical
for both command types. These fields are described in detail below.
Buffer Address (BSEL4, BSEL5, and Bits Six and Seven of BSEL7)
The four fields (two bytes and two bits) make up the 18-bit buffer address used by a DMPII to transfer
data to or from a location in main memory through non-processor request execution. BSEL4 forms the
low-order eight bits of the buffer address with BSEL5 the next eight higher-order bits. Finally, the two
high-order bits 16 and 17 of the buffer address are created by bits six and seven of BSEL 7.
Character Count (BSEL6 and Bits Zero through Five of BSEL7)
The two fields that make up the 14-bit character count provide for a maximum buffer size of 16,383
bytes. The eight low-order bits of the character count is in BSEL6, with the six high-order bits in bits
zero to five of BSEL 7. A character count is always in positive binary notation. If any other notation is
used, the results are undefined. In addition, a character count of zero results in a procedural error (see
Section 3.4.2.4).
3.4 DMPll RESPONSES
The DMPll has a set of three response structures that it can use to respond to specific user program
commands as well as to inform the user program of functional events and error conditions. These responses are:
1.
2.
3.

Buffer disposition response,
Control response, and
Information response.

Although a one-to-one communication does not exist between user program commands executed by the
DMPII and responses issued to the user program, commands often get responses. For example, the
disposition of transmit and receive messages is specified by the buffer disposition response. Information
requested through control commands for TSS, ass, and request key data is provided by the information response. However, error messages and messages about functional events, which are generally unsolicited, are returned to the user program through the control response. Like user-program commands,
DMPll responses are identified by a type code in bits zero through two of BSEL2.

3-19

NOTE
In multipoint networks, each response issued includes the address of the tributary that the response concerns in the tributary address field (BSEL3). However, in point-to-point networks equivalent responses
always have a one in BSEL3.
3.4.1 Buffer Disposition Response
Management of transmit and receive buffer completions uses five different types of buffer disposition three for transmit buffers and two for receive buffers. As shown in Figure 3-8, the format for these
responses is the same for all transmit and receive buffer dispositions. However, the types of transmit
and receive buffer dispositions are identified by the type code in bits zero through two of BSEL2. A
summary of the relationship of buffer disposition to type code is found in Table 3-7.
6

,6 , 5 , 4 , 3 , 2 , 1

BSEL3

RQI
ROO

BfA
17

I I

BfA
16

BUFFER ADDRESS
HIGH BYTE
CHARACTER COUNT
HIGH piX BITS
13

RDI

CMD TYPE
COOE*

BUFFER ADDRESS
LOW, BYTE

lEI

BSELO
BSEL2
SEL2
BSEL4
SEL4

CHARACTER COUNT
LOW-.t BYTE J

BSEL5

TRIBUTARY ADDRESS
I

BSEL7

lEO

BSEL6
SEL6

* BUFFER DISPOSITION RESPONSE TYPE CODES:
1. RECEIVE BUFFER COMPLETE = 000
2. RECEIVE BUFFER UNUSED
011
3. TRANSMIT BUFFER. COMPLETE = 100
4. TRANSMIT BUFFER SENT NOT ACK'd
5. TRANSMIT BUFFER NOT SENT = 111

Figure 3-8

= 110
Buffer Disposition Response Format

Receive Buffer Completed
When a message is successfully received and stored in the assigned main memory buffer, the DMPll
informs the user program of this successful completion by issuing a buffer disposition response with the
type code for receive message completed. Data in a receive buffer is not valid until such a response is
given for that buffer. This response is issued for both receive buffers from the common pool and for
private buffers.
Receive Buffer Unused
When the protocol at the control or tributary station is halted for a specified tributary, the DMPI I
automatically returns all private receive buffers allocated to that tributary whether used or not. Also
included are buffers for messages received but not acknowledged when the protocol was halted~ As a
result, unacknowledged message buffers could have been written into, and have invalid information.
The DMPll issues one buffer disposition response for each buffer returned.
When the user program at the control or tributary station issues a control command having the request
halt state request key, with the tributary address field (BSEL3) set to zero, the DMPll returns all
common pool buffers. The DMPl1 returns these buffers by issuing one buffer disposition response having the receive buffer unused type code for each common pool buffer left unused. When all common
pool buffers are returned, the DMPl1 issues an information response indicating that the completion
process is at an end. If the common buffer pool is empty when the request halt state control command is
issued, the DMPll issues the same information response (se~ Section 3.4.3).
3-20

Table 3-7

Buffer Disposition Response Type Codes
Binary Code

Buffer Disposition
Receive Buffer Completed

000

Receive Buffer Unused

all

Transmit Buffer Completed

100

Transmit Buffer Sent But Not
Acknowledged

110

Transmit Buffer Not Sent

111

Transmit Buffer Completed
A Buffer Disposition Response having the type code for transmit buffer completed serves to inform a
user program that the message from the specified buffer has been successfully transmitted to the designated tributary. Successful transmission means that the receiving station has acknowledged receiving
the message through the protocol.
Transmit Buffer Sent but not Acknowledged
When the protocol for a tributary is halted, either at a control or tributary station, the DMP11 automatically returns all transmit buffers being processed by that tributary to the user program. For each transmit message sent but not acknowledged at the time the protocol was halted, the DMP11 issues one
buffer disposition response having the tributary address and this type code.
NOTE
During protocol operation, after seven unacknowledged transmissions of a message, the transmit
threshold error is exceeded and the DMPll issues a
control response indicating this error (see Section
3.4.2). The DMPll continues to transmit the message and responsibility for terminating the transmission belongs to the user program.
Transmit Buffer Not Sent
At both the control station and at tributary stations, the DMP11 handles transmit message traffic by
maintaining a queue of buffers to be transmitted for each tributary address created in the network.
When the protocol for a given tributary is halted at the control or tributary station, the DMP11 automatically returns all unused buffers allocated to that tributary to the user program. The DMP11 issues
one buffer disposition response having the tributary address and this type code for each transmit buffer
staying in the queue of transmit buffers.
The protocol can be halted for a tributary in one of three ways:
1.

The user program issues a control command halting the tributary.

A DDCMP STRT message is received while the tributary is in the RUN state.

A DDCMP maintenance message is received, temporarily halting the protocol, while receive
buffers are being returned.

3-21

3.4.1.1 Buffer Disposition Response Field Descriptions - Except for the type code field, whose function has been defined in the preceding section, this response has six other fields that apply equally to all
five response changes. These fields, which serve to name the size and main memory address of the
buffer being completed are described in detail below.
Buffer Address (BSEL4, BSEL5, and Bits Six and Seven of BSEL7)
These four fields (two bytes and two bits) make up the I8-bit buffer address used by a DMPII to transfer data to or from a location in main memory through NPR execution. BSEL4 forms the eight loworder bits of the buffer address with BSEL5 the next eight higher-order bits. Finally, the two high-order
bits 16 and 17 of a buffer address are created by bits six and seven, of BSEL 7.
Character Count (BSEL6 and Bits Zero Through Five of BSEL7)
The two fields which make up the 14-bit character count provide a count of the actual number of bytes
which make up the message having in the buffer being returned. The eight low-order bits of the character count is in BSEL6, with the six-high order bits in bits zero to five of BSEL 7. A character count is
always in positive binary notation.
3.4.2 Control Response
A control response is an unsolicited response issued by the DMPII when an error is detected or when
certain functional events occur that need the user program be notified. There are four types of information passed to the user program by control responses:
1.
2.
3.
4.

System events,
Protocol events,
Network errors, and
Procedural errors.

The format for the control response is shown in Figure 3-9. Figure 3-9 shows that except for the tributary address and type code fields, a control response has five information fields. Four of these fields are
used to include the IS-bit UNIBUS address. The fifth field is dedicated to the error/event code.
The UNIBUS address fields, which are used to identify undersize buffers and non-existent memory
locations, are pertinent only to certain error conditions. The event/error code is in BSEL6 of a control
response and is an octal number in the range of two through 376. The name of each event/error and the
corresponding octal code are listed in Table 3-S. Procedural errors are listed in Table 3-9.
NOTE
The event/error octal codes are all even values to
make the use of these codes for dispatch table entry
easier.
7

5
I

3
I

-'-

ROO

B/A
17
15

BSEL2
SEL2
BSEL4
SEL4

BSEL6
SEL6

COUNT LOW BYTE

EVENT/ERROR
C0...LDE
.~
~

*ONLY APPLICABLE FOR NON·EXISTENT MEMORY
AND BUFFER TOO SMALL ERRORS

Figure 3-9

BSELO

1
0
CMD TYPE CODE

* BUFFER ADDRESS OR CHARACTER

I I
B/A
16

RDI

o
lEI

* BUFFER ADDRESS OR CHARACTER
COUNT HIGH BYTE

BSEL5

lEO

ROI

0
I

TRIBUTARY ADDRESS

BSEL3

BSEL7

1
I

2
I

Control Response Format

3-22

Table 3-8

Event/Error Names, Categories, and Corresponding Codes (in Octal)

Event/Error

Information Category

Code

Receive Threshold Error

Network Error

Transmit Threshold Error

Network Error

Select Threshold Error

Network Error

Start Message Received
While Running

Protocol Event

Maintenance Message
Received While Running

Protocol Event

Maintenance Message
Received While Halted

Protocol Event

Start Message Received
While in Maintenance Mode

Protocol Event

Reserved

N/A

Tributary Polling State
Changed to Dead

System Event

Tributary in Run State

Protocol Event

Babbling Tributary

Network Error

Streaming Tributary

Network Error

Modem Ring Detect

System Event

Reserved

N/A

Reserved

N/A

Procedural Errors

(See Table 3-1)

100

Procedural Errors

(See Table 3-9)

302

3-23

Table 3-8

Event/Error Names, Categories, and Corresponding Codes (in Octal) (Cont)
Event/Error

Information Category

Code

Modem Disconnect

System Event

304

Queue Overflow

System Event

306

Modem Carrier Loss

System Event

310

Reserved

N/A

312

Reserved

N/A

376

Table 3-9

DMPll Procedural Errors and Corresponding Codes (in Octal)

Octal Code

Procedural Error

100

A command other than a mode definition command is issued before the mode
created. (See Section 3.3.2.)

102

Invalid type code used in a command.

104

Invalid mode change (for example the mode of a tributary station is changed
to point-to-point).

106

A non-global command is issued to a tributary which has not been created.

110

A non-global command is issued with a tributary address of zero.

112

Try to delete or place an unhalted tributory in the start or maintenance
protocol state.

114

Try to create more than 32 tributaries.

116

Try to create a tributary that is currently created.

122

An invalid request key is used in a control command.

124

Try to assign a buffer to a halted tributary.

126

Try to assign a buffer with a byte count of zero.

130

Try to assign a transmit buffer with the tributary address field (BSEL3) set
to zero.
3-24

Table 3-9

DMPl1 Procedural Errors and Corresponding
Codes (in Octal) (Cont)

Octal Code

Procedural Error

132

Try to read and clear or write into a reserved section of a TSS or GSS.

134

Try to use the reserved bits in BSEL 7 of a control command.

136

Try to retrieve common pool buffers while the common pool is in use.

140

Try to raise the common pool buffer quota to a value higher than 376 octal
(see Section 3.3.3.1).

142

Reserved

276

Reserved

300

Buffer too small- recovery from this error condition is detailed in Section
4.6.2.2

302

Non-existent memory error - recovery from this error condition is detailed in
Section 4.6.2.1.

I n general, the functional events and errors described in this section apply to all stations. Exceptions to
this are identified where applicable.
The meaning of each system and protocol event reported by a control response, with a description of
each error condition recorded, are detailed in the sections that follow. The function of the UNIBUS
address fields in specific error reports are defined in the descriptions of those errors.
3.4.2.1 System Events - There are five system events reported by control responses. These concern the
functional state of tributaries and the state of the local modem.
Tributary Polling State
Changed to Dead

This control response code informs the user program at a multipoint
network control station that the polling level of the tributary addressed by BSEL3 has been changed to the dead state by the polling algorithm. A tributary can be removed from the dead state by
the user program through a control command implementing the
latch/unlatch polling function. Also, if a dead tributary responds,
the polling algorithm automatically places the tributary in the active state (see Section 5.2.1). This control response has no meaning
in point-to-point networks.

3-25

Modem Ring Detected

This control response code informs the user program at all stations
that an off-to-on transition of the EIA signal ring, continuing for at
least one millisecond, was detected by the local. modem. When detected, this signal indicates that a remote station is requesting connection of the physical communications line. Modem ring detect ap.;
plies only to networks' connected by a switched line and is not used
in networks based on a dedicated line. Because this is a global response, the contents of BSEL3 are zero.

Queue Overflow

This control response informs the user program at all stations when
the free linked list is empty (see Section 5.4). Often, this response
indicates that operational conditions are creating responses faster
than the DMPII can process them. Because this condition is a global response, the contents of BSEL3 are zero. Once this response is
issued, the user program has three seconds to recover the next response from the CSRs (three seconds per pending response) before
the CSRs are cleared by the internal microdiagnostics. To restore
the DMPII to operation, the complete startup sequence must be executed (see Section 4.3).
NOTE
This is a fatal error. The DMPll will reinitialize itself for subsequent startup.

Modem Disconnected

This control response code informs the user programs at all stations
that an on-to-off transition of the EIA signal data set ready (DSR),
continuing for at least one millisecond, was detected. Such a transition of DSR indicates that the local modem is diconnecting from
the physical communications line. Because this is a global response,
the contents of BSEL3 are zero. This response is also posted to user
programs in point-to-point networks. when there is a failure to connect to the other station. In this event, the response serves as a notification to dial again.

Modem Carrier Loss

This control response code informs the user program that the EIA
signal carrier detect has gone from on to off while the DMPII was
in the process of receiving a message. Because this is a global response, the contents of BSEL3 is always zero.

3.4.2.2 Protocol Events - A DMPII uses the control response to report five different protocol events
to the user program. These reports serve to inform the user program of the state of protocol operations
associated with the tributary specified by the address in BSEL3.
STRT Message
Received While Running

This control response informs the user program at a control or tributary station that a DDCMP STRT message was received 'by a tributary while that tributary was in the DDCMP run state. In the
usual operation, the control station DMPII issues a DDCMP
STRT message to inform a tributary that the user program at the
control station has started protocol operation for that tributary. A
D DCMP STRT message is also used to resynchronize the logical
link between the control station and a tributary because of conditions, such as threshold errors or receive and transmit overruns, that
inhibit efficient message traffic. In this event, the user program at
3-26

the control station (after the tributary is halted at the control station) issues a request startup state control command causing the associated DMPII to send DDCMP STRT messages to the tributary
being resynchronized. As a result, the protocol at that tributary is
halted and the DMPII at that station sends a control response to its
user program having the code for STRT message received while
running. All buffers are then returned through a buffer disposition
response for receive buffers unused. At this point the tributary user
program has the option of restoring the logical link by issuing a
request startup mode control command to the halted tributary. As a
result of this action, the protocol becomes functional at that tributary, allowing it to respond the next time it is polled.
Maintenance Message
Received While Running

This control response informs the user program at the control or tributary station that a DDCMP maintenance message was received
by a tributary while that tributary was in the DDCMP run state. As
a result, the tributary at the receiving station is halted and all buffers (used and unused) are returned to the user program. When all
buffers are returned, the DMPII informs the user program of this
by issuing an information response (see Section 3.4.3) and then
places the tributary in the maintenance state. Once in the maintenance state, the tributary is ready to perform the functions implemented by the maintenance operation protocol (MOP). The actual
maintenance message that caused this event is not returned to the
user program.

Maintenance Message
Received While Halted

This control response informs the user program that at a control or
tributary station a DDCMP maintenance message was received by
a tributary while it was in the DDCMP HALT state. After receiving a maintenance message under these conditions, the tributary is
placed in the DDCMP maintenance state ready to operate under
MOP. The actual maintenance message that caused this event is
not returned to the user program.

STRT Message Received
in Maintenance State

This control response informs the user program at a control or tributary station that a DDCMP STRT message was received by a tributary while it was in the DDCMP maintenance state. After informing the user program of this event, the D MP 11 does not take
any further action. Any action taken in response to this protocol
event is up to the user program.

DDCMP RUN State Entered

This control response code informs the user program at the control
or tributary station or at a point-to-point station that the tributary
addressed by BSEL3 has entered the DDCMP RUN state. The
DMPll issues this response as an answer to a request startup state
control command from the user program. When this response is received, the user program is informed that the protocol is functional
for the tributary or station and that message traffic can begin (see
Section 3.3.3.1).

3.4.2.3 Network Errors - The DMPll provides control responses for five different network error conditions. All network errors, whether pertaining to a tributary or point-to-point station, are capable of
being recovered.
3-27

Receive Threshold Exceeded

When the number of consecutive receive errors recorded for a given
station equals seven, this network error is reported to the user program by a control response. Each time a receive threshold exceeded
error is reported, the receive threshold error counter is reset to zero.

Transmit Threshold Exceeded

When the number of consecutive transmit errors recorded for a given station equals seven, this network error is reported to the user
program by a control response. Each time a transmit threshold exceeded error is reported, the transmit threshold error counter is reset to zero.

Select Threshold Exceeded

When the selection interval timeout counter has timed out seven
consecutive times, this network error is reported to the user program at a multipoint control station, or a half-duplex, point-to-point
station by a control response. Each time a select threshold exceeded
error is reported, the counter is reset to zero. The programmable selection interval period is the time allocated to a tributary for responding to a poll, or to a station for responding to a transmission. A
selection interval is terminated either by a timeout of the counter
resulting in this error condition or by a message which has the selection bit set.

NOTE
For further details on threshold counters see Section
5.3.1.2.
Babbling Tributary Detected

When the babbling tributary timeout period is exceeded, this network error is reported to the user program at a multipoint control
station or a half-duplex, point-to-point station by a control response.
A babbling tributary is a tributary that continues to transmit valid
DDCMP messages without deselecting itself until the programmable babbling tributary timeout counter expires. Detection of this
error condition usually indicates that the transmitting station is malfunctioning, or that the period of the babbling tributary timeout
counter is too short (see Section 4.4.2). Recovery from this error
condition usually requires human intervention at the remote station.

Streaming Tributary Detected

When the streaming tributary timeout period is exceeded, this network error is reported to a user program at a multipoint control station or a half-duplex, point-to-point station by a control response. A
streaming tributary is a tributary or point-to-point station that has
failed to relinquish the physical communications line at the end of
the selection interval for that station. The streaming tributary timeout period, which is determined by the programmable streaming tributary timer, starts at the end of the selection interval. Detection of
this error condition usually indicates that the modem at the remote
station has malfunctioned, or that the period of the streaming tributary timer is too short (see Section 4.4.3). Recovery from this error usually requires human intervention at the remote station.

3.4.2.4 Procedural Errors - Procedural errors are those errors generated when the user program violates the procedures used to interface with the DMPl1. With three exceptions (see Table 3-9, error
codes 140, 300, and 302), these errors, when detected, are placed at the beginning of the DMPII internal response queue and passed immediately to the user program as control responses. Also, with one
exception (see Table 3-9 error code 302), these errors are not fatal.

3-28

NOTE
For this, a fatal error is an error that results in protocol shutdown.
Octal codes contained in BSEL6, in the range 100 to 144 and 300 and 302, serve to define the specific
procedural violations. The name of each procedural error and its corresponding octal code are listed in
Table 3-9.
One of the exceptions to immediate posting of procedural errors by the DMPll is that of raising the
common buffer pool quota for a tributary to a value higher than 376 (procedural error code 140). In
this event, the DMPII automatically resets the quota to 376. There is no urgency to the posting of this
error. A control response with this error code is queued for reporting in the same manner as all other
control responses.

NOTE

With the exception of the buffer too small and nonexistent memory errors, the only control response
fields used by the DMPll when posting a procedural error are type code field in BSEL2 and output code field in BSEL6. All other fields stay as set
by the user program in the command that originally
generated the procedural error (see Figure 3-9).
A buffer-too-small error occurs when the size of the buffer allocated to a receive message is not large
enough to store the message. As shown in Figure 3-9, the control response returned when this error
occurs has the character count of the message that caused the error. The DMP11 does not look for a
buffer large enough. Instead, it requests retransmission of the message and expects that the user program will provide a correct buffer. Recovery from this error is detailed in Section 4.6.2.2.
A non-existent memory procedural error is reported to a user program by a control response when the
DMP11 tries to access a main memory location that does not exist. As shown in Figure 3-9, the
UNIBUS address that causes the error is returned to the user program in BSEL4, BSEL5, and BSEL 7
of the reporting control response. This error can be fatal to the tributary or point-to-point station that
initiated the memory access causing the error. Section 4.6.2.1 describes a recommended procedure for
recovering from this error.
3.4.3 Information Response
Information responses are issued by the DMP11 in direct response to a request for information by the
user program. As described in Section 3.3.3.1, an information response can be generated by control
commands having specific request key codes. Also, buffer return complete responses caused by certain
protocol events reported by control responses (see Section 3.4.2.2) are passed to the user program by
means of the information response. The format for the information response is shown in Figure 3-10.
7
7

6
I

5
I

BSEL3

0
I

•I

•

j
I

I
I

DATA

READ1READ
CLR TSS
TSS

NOT USED
I

•

lEI

DATA
_L

BSEL7

r
1
0
CMD TYPE CODE

RDI

ROO

BSEL5
I

TRIBUTARY ADDRESS
_l

4
lEO

RQI

2
I

BSEL2
SEL2
BSEL4
SEL4
BSEL6
SEL6

'RETURN KEY OR I
.TSS ,\DDREpS
j

BSELO

o
MK-2391

Figure 3-10

Information Response Format

3-29

3.4.3.1 Information Response Field Descriptions,....; Figure 3-10 shows that the data passed to the user
program is contained in BSEL4·and BSEL5, with the fields in BSEL6 naming the type of data. BSEL3
contains the address of the tributary associated with the data in BSEL4 and BSEL5. However, if the
Information response is returning data retrieved from a GSS, BSEL3 has a zero.
Return Key Codes
As shown in Figure 3-10, bits zero to four of BSEL6 of an information response can have either a return
key code or a TSS/GSS address, with the nature of the TSS/GSS access defined by bits five and six of
BSEL6. The return key codes are detailed below:
Return Modem Status
(octal code = 10)

An information response with this return key code in bits zero to
four of BSEL6 has the contents of line unit register 13 and 17 in
BSEL4 and BSEL5. This information response is returned when the
user program issues a control command with the request key read
modem control. The format and data content of line unit registers
13 and 17 are described in Appendix C.

Buffer Return Complete
(octal code = 20)

An information response with this return key code in bits zero to
four of BSEL6 informs the user program that the process of returning buffers has been completed. This buffer returning process is
started under the following circumstances:
1.

The protocol event STRT message received while running
occurred. This causes all private buffers assigned to the
tributary defined by the address in BSEL3 to be returned.

The protocol event maintenance message received while
running occurred. This causes all private buffers assigned
to the tributary defined by the address in B SEL3 to be
returned.

The user program issued a control command with the
request key request halt state. This causes all private
buffers assigned the tributary defined by the address in
BSEL3 to be returned.

The user program issued a global control command
(BSEL3 = zero) with the request key request halt state.
This causes all unused common pool buffers to be
returned.
Information responses with the return key code for buffer return
complete always have zeros in BSEL4 and BSEL5.
TSS /GSS Access
When a user program accesses a TSS or a GSS through a control command, the DMPll reads the
defined location and passes the data to the user program by means of an information response. Reading
a TSS ora GSS location stores two bytes of data from that location in BSEL4 and BSEL5 of the information response, with the lower-order byte in BSEL4 and the high-order byte in BSEL5. If a TSS is
read, BSEL3 in the information response has the associated tributary address. However, data read from
a GSS passes to the user program through an information response having a zero in BSEL3.
As described in Section 3.3.3.1, any TSS or GSS·location can be read and specific locations can be read
and cleared. With the TSS/GSS address, BSEL6 in an information response also has two single bit
fields that indicate a read or a read and clear TSS/GSS. These information response fields are described in detail below:
3-30

TSS/GSS Address

This field (BSEL6, bits zero to four) has the octal address of the
TSS/GSS location from which the data in BSEL4 and BSEL5 was
read.

Read TSS/GSS

When set, this bit (BSEL6, bit five) designates that BSEL4 and
BSEL5 have the data requested by the read TSS/GSS control command.

Read and Clear TSS/GSS

When set, this bit (BSEL6, bit six) designates that BSEL4 and
BSEL5 have the data requested by the read and clear TSS/GSS
control command. In addition to reading the requested TSS/GSS
location, the DMPll also clears that location.

NOTE
The TSS/GSS locations accessible for writing,
reading, and reading and clearing are listed and described in Section 3.3.3.1.

3-31

CHAPTER 4
PROGRAMMING METHODS

4.1 INTRODUCTION
The design of user programs to control the operation of a DMPII needs attention to a number of programming subjects specific to the user program/DMPII interface. These subjects are listed below and
are detailed in the separate sections referenced.
1.

Control and status register (CSR) interface discipline (Section 4.2)

Configuration and startup (Section 4.3)

Criteria for determining user-defined parameters (Section 4.4)

Error counter access (Section 4.5)

Error recovery procedures (Section 4.6)

Booting a remote station (Section 4.7)

4.2 CSR INTERFACE DISCIPLINE
The command/response interface between a DMPII and the user program is the eight-byte DMPII
CSRs that are addressed through the CPU I/O page. Because the DMPII microprocessor runs in a
multiprocessing mode with the PDP-II /V AX-II, the passing of commands and responses through this
interface must be disciplined in order to remove the possibility of a race condition.
NOTE
A race condition results when the user program sets
control bits in the CSRs after the microprocessor
has read those bits).
This interface discipline demands that the user program to follow two separate procedures: one for issuing commands and one for r~trieving responses.
Figure 4-1 shows the control bits in the CSR interface discipline. These bits are located in the OMPII
CSRs BSELO and BSEL2. Figure 4-1 shows that BSELO has two control bits named interrupt enable in
(lEI) and interrupt enable out {lEO), bits zero and four. These bits, when set, enable the M8207 microprocessor to interrupt the main CPU under two circumstances. These are:
1.

The CSRs become available to the user program for the issuing of a command (lEI) when
the ready in bit (ROI) is set by the DMPII.

A response is to be read from the CSRs by the user program (lEO) when the OMP11 sets the
ready out bit (ROO).

A OMPII is capable of operating in both an interrupt and a non-interrupt mode.

4-1

RQI

lEO

ROO

RDI

o
lEI

BSElO
BSEl2
MK·2390

Figure 4;..1

CSR Interface Control Bits

NOTE
For the best system performance, it is recommended
that all system operation be conducted in the interrupt mode with the lEI and lEO bits set to one.
The procedures defining CSR interface discipline, which are described in the sections that follow, are
based on operation in the interrupt mode. As a result, the interrupt enable bits, lEI and lEO, should be
set by the user program before using the CSR interface.
4.2.1 Issuing Commands
A user program issues commands to a 0 MP II in four steps:
1.

The user program sets the request in (RQI) bit (bit seven of BSELO, see Figure 4-1). When
the DMPll sets the RDI bit it also interrupts the main CPU if lEI is set.

If a single command is to be issued, the user program clears RQI. If a series of commands
are to be issued, RQI can stay set until immediately before initiating step three for the last
command issued. By leaving RQI set while issuing a series of commands, the user program is
assured that it will have access to the CSRs after the next response.

The user program loads the command into the CSRs BSEL3, SEL4, and SEL6.

The user program sets the command type code in bits zero to two of BSEL2 and clears RD I.
Clearing RDI informs the DMPII that the CSRs have a command to be read.

4.2.2 Reading Responses
When the DMPII has a response for the user program, it sets the CSR control bit ROO. When the
lEO bit is set, it interrupts the main CPU. This action results in the OMPll giving CSR read access to
the user program.
With the response read, the user program clears the ROO bit. This action results in control of the CSRs
being passed back to the DMPII. Note that the complete response must be read from the CSRs before
clearing ROO because user program control of the CSRs is given up once ROO is cleared. In addition,
to make the operation efficient requires that control of the CSRs by the user program, be kept to a
minimum and all processing of responses must be done after ROO is cleared.
4.2.3 CSR Interface Interactions
User program access to the CSRs is under complete control of the DMPII. Access to the CSRs is given
on request when the user program has a command to issue or when the OMPII has a response for the
user program. Figure 4-2 shows the type of access window available to the user program operating under interrupt control when issuing a command or reading a response.
.
As previously indicated, the user program sets RQI to request use of the CSRs for the purpose of issuing a command. The DMPII makes the CSRs available for issuing a command - only when it is not
using them - by setting ROland interrupting the user program through the floating vector XXO. As a
result, the time between the setting of RQI by the user program and the access to the CSRs through an
interrupt cannot be determined.

4-2

USER

SETS
RQI
_

USER CLEARS RQI
NO FURTHER COMMANDS
TO ISSUE

* DMP11
SETS
RDI

* USER CLEARS RDI TO
, . . - - - - - - - - - - - -..... NOTIFY DMP11 THAT COMMAND
USER HAS TOTAL CONTROL OF CSRs HAS BEEN ISSUED
_ _ _ _ _OJ TO ISSUE COMMAND
(lEI)

*DMP11

* USER CLEARS ROO TO

..I

SETS
t~--------------_R_D_O_--,_
USER HAS TOTAL CONTROL OF CSRs
(lEO)
TO RETRIEVE RESPONSE

NOTIFY DMP11 THAT
RESPONSE HAS BEEN RETRIEVED

*DMP11 HAS OWNERSHIP OF CSRs
MK·2388

Figure 4-2

CSR Access Window

When commands are continuing to issue from a queue and the last command is issued with the user
program owner of the CSRs (RDI is set), there is a way for the user program to return control of the
CSRs to the DMPII. In such events, the user program can issue a control command with the code for
no request in the request key field, informing the DMPII to ignore the contents of the CSRs. In this
way, the possible reading of incorrect data by the DMPll is avoided.
If the user program is to issue a single command to the CSRs, RQ I should be cleared before the command is issued, as indicated in Figure 4-2. However, if a series of commands are to be issued, the user
program can leave RQI set. In this way, when a command is issued and RDI is cleared, the DMPll
will give up the CSRs to the user program after the next response. The time between the clearing RDI
by the user program after completing one command and the access to the CSRs for the next command,
also cannot be determined.

When the DMPll has a response to be passed to the user program, it sets RDO and interrupts the main
CPU through the floating vector XX4. At this point, the user program is the owner of the CSRs and
can proceed by reading the response. Once the response is read from the CSRs, the user program
should immediately clear the RDO bit. User program routines responsible for issuing commands and
recovering responses should be designed to limit CSR access time to that time needed to load a command or recover a response.
4.3 DMPII CONFIGURATION AND STARTUP
When starting a DMPll, the user program needs to perform a series of steps that form three procedures; one to configure the DMPII to operate within the network, one to create user-defined parameters, and one to start protocol operations.
4.3.1 Configuration Procedure
The sequence to configure a control and tributary station DMPll for network operation includes the
following steps:

4-3

Set the master clear bit and wait for the run bit to be set (see Section 3.3.1 and Figure 3-4).

On detecting that the Run bit is set, read BSEL4 and BSEL6.

The DMPll is functional when the contents of BSEL4 equals 77 and BSEL6 equals 305
(octal), indicating that the startup microdiagnostics have executed successfully. Any other
values in BSEL4 and BSEL6 indicate that an error condition was detected by the startup
microdiagnostics. (The values and meanings of these diagnostic error codes are listed in
Table 3-3.)

If the DMPl1 mode is software defined, set the mode for that device by issuing the correct
mode command (see Section 3.3.2). If the mode for a DMPll was setup by internal switches,
this step can be ignored (see Section 2.6).

Setup the tributary addresses for the DMPll by issuing one control command having the
request key establish tributary for each tributary address to be established at this DMPll
(see Section 3.3.3.1).

If the DMPll is a station in a point-to-point network, issue a control command to create a
tributary address of one for that station.

4.3.2 Specifying User-Defined Parameters
After a DMP11 is configured, the user-defined parameters can be specified. User-defined parameters
include parameters used by the polling algorithm and those specific to protocol operations. As described in Section 3.3.3.1, user-defined parameters are specified through control commands configured
to address a tributary status slot (TSS) or a global status slot (GSS). When setting polling and protocol
parameters, the user program has the option of accepting the default for a parameter or setting the
parameter to some predetermined value. The criteria for determining, the best values for the different
polling parameters are detailed in Section 5.2.3. Criteria for the rest of the parameters, which generally
concern the operation of the communications link, are described in Section 4.4.
NOTE
Although most of the user-defined parameters are
16 bits in length, some are single-byte parameters. If
the user program is to accept the default for one parameter and set the other to a unique value, then
both the- default value and the unique value must be
written. This is necessary because both TSS and
GSS user-defined parameters are written on two
byte boundaries.
The process of specifying user-defined parameters includes two basic types:
1.

Parameters for control and tributary station TSS structures

Parameters for the control and tributary station GSS

4.3.2.1 Specifying TSS Parameters - TSS parameters that can be specified by the user program are
listed in Table 4-1 by name, BSEL6 address, size, and default value. A functional summary of of each
parameter- is also given.
NOTE
The octal values listed in Table 4-1 under the heading "TSS Address (octal value of BSEL6") are
based on the setting of bit seven in BSEL6 to start a
write TSS.

4-4

Table 4-1

User-Defined TSS Parameters

TSS Address
(Octal
value
of BSEL6)

Size
(Bits)

Default
Value
(Octal)

XDT
PRESET

230

Q (active)

231

377

Q (inactive
Q (unresponsive)

232

233

R (active)

231

R (inactive)
R (unresponsive)

232

100

233

NDM-INACT

234

Number of no data messages needed to
go inactive. This parameter specifies the
number of consecutive polls to be made
to a tributary without any data messages
being received, before changing to polling
state from active to inactive. This
parameter applies only to TSS structures
at a multipoint control station.

TO-UNRSP

234

Number of timeouts to go unresponsive.
This parameter specifies the number of
consecutive times a tributary will be
polled with no response being received
(consecutive selection timeouts) before
the polling state is changed from active or
inactive to unresponsive. Values for
NDM-INACT and TO-UNRSP can be
specified with a single control command.

Name

(no delay)

4-5

Description
Preset value for the transmit delay timer.
This parameter provides a fixed delay
between transmissions of data and
maintenance messages to the tributary.
The initial polling urgency (U) for
tributary (i). The TSS for a tributary (i)
must be assigned a value for each of the
three polling levels: active, inactive, and
unresponsive. This parameter is
applicable only to TSS structures at the
control station.
The rate (R) by which urgency (U) is
increased for tributary (i). The TSS for
tributary (i) must be assigned a value for
R for each of the three polling levels:
active, inactive, and unresponsive. Both
the Q and R values for a given tributary
are created through a single control
command. Therefore, if one parameter is
to be set to a unique value and the default
is to be accepted for the other, both the
default value and the unique value must
be written. This parameter is applicable
only to the TSS structures at the
multipoint control station.

Table 4-1

Name

TSS Address
(Octal
value
of BSEL6)

User-Defined TSS Parameter (Cont)

Size
(Bits)

Default
Value
(Octal)

Description
If one parameter is to be set to a unique
value and the default is to be accepted for
the other, both the default value and the
unique value must be written. This
parameter applies only to the TSS
structures at a multipoint control station.

TO-READ

235

The number of timeouts to go dead. This
parameter specifies the number of
consecutive polls to be made to an
unresponsive tributary (consecutive
selection timeouts) before the polling
state is changed from unresponsive to
dead. This parameter applies only the
TSS structures at a multipoint control
station.

MXMC

235

Maximum transmitted message count.
This parameter specifies the count of the
maximum number of abutting data
messages to be transmitted by a station
before deselecting itself. This count
applies to the TSS structures at both
control and tributary stations in
multipoint networks as well as point-topoint stations. Both TO-DEAD and
MXMC for a given tributary are created
through a single control command.
Therefore, at the multipoint control
station, when one parameter is to be set to
a unique value and the default is to be
accepted for the other, both the default
value and the unique value must be
written. At tributary and point-to-point
stations, the polling parameter TOO EAD is ignored.

SEL TIMER

236

3 (sec.)
(5670

Selection interval timer. This timer is
started when a message is transmitted
with the select flag set and halted when a
valid response is received or the line is
resynchronized. The selection timer is
used as a reply timer for full-duplex,
point-to-point networks and as a selection
timer at multipoint control stations and
half-duplex, point-to-point stations. This
counter counts in milliseconds, from one
to 65,535 milliseconds.

octal)

4-6

Table 4-1

User-Defined TSS Parameter (Cont)

Name

TSS Address
(Octal
value
of BSEL6)

Size
(Bits)

BAD TIMER

237

Default
Value
(Octal)
6 (sec.)
(13560
octal)

Description
Babbling tributary timer. This timer is
used to detect a babling tributary (see
Section 3.4.2.3). In a multipoint network,
this parameter applies only to the control
station. However, this parameter applies
to both stations in point-to-point networks
operating in half-duplex mode.

The actual order for specifying TSS parameters is arbitrary. The complete command series to specify
these parameters for TSS structures at a multipoint control station are listed below:
1.

Issue a series of control commands to setup the preset value for the transmit delay timer.

Issue a series of control commands to create the polling parameters initial polling urgency
(Q) and polling rate (R) for the three polling levels.

Issue a series of control commands to specify values for the active, inactive, unresponsive,
and dead polling state-change parameters.

Issue a series of control commands to specify values for the maximum transmitted message
count.

Issue a series of control commands to set the selection timers for tributaries (or issue a single
command to set the point-to-point station reply timer).

Issue a series of control commands to set the babbling tributary timers.

4.3.2.2 Specifying GSS Parameters - As previously indicated, when one or more tributary addresses
are setup at a DMPll, the microcode automatically creates a GSS for that control or tributary station.
The GSS parameters that can be specified by a user program are listed in Table 4-2 by name, BSEL6
address, size, and default value. A functional summary of each parameter is also given.
The sequence of control commands necessary to specify GSS parameters for a multipoint control station are listed below:
1.

Issue a control command to specify the number of sync characters (NUM SYNC) that are to
precede non-abutting transmit messages.

Issue a control command to set the streaming tributary timer (STREAM TRIB).

Issue three control commands to assign values for the global polling parameters delta time
(DELTA T), dead tributary (DEAD T), and poll delay.

4-7

NOTE
The octal values listed in Table 4-2 under the heading "GSS Address (octal value of BSEL6)" are
based on the setting of bit seven of BSEL6 to start a
write GSS. Also, when specifying a user-defined
GSS parameter, BSEL3 in the pertinent control
command has a zero.
Specific user-defined TSS and GSS parameters are common to both control and tributary stations.
Note that control commands specifying TSS parameters must have the address of the tributary associated with the TSS in BSEL3. Similarly, each control command specifying a GSS parameter must have
BSEL3 set to zero.
Table 4-2

User-Defined GSS Parameters

Name

GSS Address
(Octal
value
of BSEL6)

Size
(Bits)

Default
Value
(Octal)

NUMSYNC

233

10 (low
speed)
15 (high
speed)

Number of sync characters. This global
value specifies the number of sync
characters that are to precede non-abutting
transmitted messages. This parameter
applies to all stations. Low speed is defined
as less than 250K bls, and high speed as
equal to or larger than 250K b/s.

STREAM
TRIB

234

6 (sec.)
(1350
octal)

Streaming tributary timer. This timer is
used to detect a streaming tributary (see
Section 3.4.2.3). In a multipoint network,
this parameter applies only to the control
station. However, in half-duplex, point-topoint networks, this parameter is
applicable to both stations.

DELTAT

235

50 (ms)

Delta time. This parameter creates the
polling urgency update interval. This
global parameter, which applies only to
multipoint network control stations, is used
by the polling algorithm to compute polling
urgency for each tributary (see Section
5.2.2). This interval is also used for the
updating of the transmit delay timer. The
default value of 50 ms is the minimum
value for this parameter.

(62 octal)

DEADT

236

10 (sec.)
(23420
octal)

POLL
DELAY

237

0
(no delay)

4-8

Decription

Dead tributary. This parameter specifies
the interval between polls for dead
tributaries. This global parameter applies
only to multipoint control stations.
This parameter provides for a fixed delay
between polls for all tributaries in a
network. If the default is accepted, the
next poll for any tributary will occur
immediately following completion of the
current poll.

Previous steps have covered specifying user-defined TSS and GSS parameters (see Table 4-1 and 4-2)
at the control station. The two steps listed below cover the parameter specifying process at the tributary
stations:
1.

Issue a series of control commands at each tributary station to set the maximum transmitted
message count (MXMC) for each assigned TSS at a tributary station. Note that the 8-bit
value for MXMC must be placed in BSEL5 of each command (the tributary station DMPII
ignores BSEL4 in these commands). The procedure for determining the best value for this
parameter for tributary stations is the same as that used for for control stations (see Section
4.4.2).

Issue a single control command at each tributary station to set a value for the number of sync
characters (NUM SYNC) that are to precede non-abutted transmit messages. The procedure
for determining the best value for this parameter for tributary stations is the same as used for
control stations.

4.3.3 Protocol Operation
At this point, the DMPll has been configured for operation in the network and is ready for protocol
operation. The steps needed to start protocol operation are:
1.

Place assigned tributaries in the ISTRT state by issuing one Control command having the
request key request ISTRT state for each tributary address.

If the DMPll is a station in a point-to-point network, one control command must be issued
having the request ISTRT state request key with a tributary address of one in BSEL3.

The DMPII confirms that the protocol is functional at each tributary by issuing a control
response having the protocol event code for DDCMP RUN state entered, one for each control command issued.

When the protocol is running at the control station and at all tributary stations, data traffic on the
network can start.
4.4 CRITERIA FOR DETERMINING USER-DEFINED PARAMETERS
User-defined 'fSS and GSS parameters are in two categories:
1.

Polling parameters that provide for user program control over the dynamic activity of the
polling algorithm end.

Communications line parameters that provide the user program with the ability to control
data traffic over the physical communications line.

Polling parameters are described in detail in Section 5.2.3. Communications line parameters include:
I.
2.
3.
4.
5.
6.

Selection interval timer,
Number of sync characters,
Babbling tributary timer,
Maximum transmitted message count,
Streaming tributary timer, and
XDT preset.

The values for the selection interval timer and the number of sync characters are related as are the
values for the babbling tributary timer and the maximum transmitted message count. As a result, these
related parameters are described in Section 4.4.1. Setting the Selection Interval Timer and in Section
4.4.2, Setting the Babbling Tributary Timer.

4-9

4.4.1

Setting the Selection Interval Timer

The function performed by the selection interval timer at a DMPII depends on the mode selected for
that DMPII. In full-duplex, point-to-point networks, this timer is used as a reply timer for the purpose
of message accountability. This timer operates as a selection interval timer when the mode for the
DMPII is one of the following:
I.
2.
3.

A full-duplex control station,
A half-duplex control station, or
A half-duplex point-to-point station.

In this capacity, it performs the link management function as well as providing for message accountability.
Link management is the process of controlling the transmission and reception of data over networks
where there are two or more transmitter/receiver devices connected to the same physical communications link. This applies to half- and full-duplex, multipoint networks as well as to half-duplex, pointto-point links. On half-duplex links, only one transmitter can be active at any time and on full-duplex
links, only one tributary transmitter can be active at a time.
A station on such links can transmit when it has been selected or given the link. Link ownership is
passed through use of the select flag in the DDCMP message header. Detecting a select flag in a received message allows the receiving station to transmit after message reception is completed. Sending
the select flag also means that this station stops transmitting after the current message is sent.
A selection timer detects the loss of a select flag by timing the interval needed to receive the longest
message from a station. A timer is started when a station is selected and reset when a valid message is
received at that station. When the timer interval is exceeded at the sending station (indicating that a
message was not received during the period of the timer) it is assumed that messages with the select
flag set were either transmitted or received in error.
At this point, the station that sent the messages with the select flag set owns the communications line
and continues transmitting as if it had received a valid select return.
The values assigned to select interval timers at stations in half-duplex point-to-point networks should be
different at both stations to prevent a possible deadlock condition where both stations become synchronized in timing out. For both multipoint control stations and half-duplex, point-to-point stations, the
criteria for determining the value for a select timer include such elements as:
I.
2.
3.
4.
5.

Maximum message length,
Number of synch characters,
Line speed,
Line turnaround time, and
Message processing delays.

As indicated in Table 4-2, the GSS parameter number of sync characters has two defaults - one for
low-speed operation (10) and one for high-speed operation (15). The functional speed range of a
DMPII is specified by the line unit high speed/low speed switch. This switch is placed in the correct
position when the DMPII is installed (see Section 2.6). In the low speed position, the DMPII can
operate at line speeds below 250K b/s. In the high speed position, the device line speed can range from
250K b/s up to 1M b/s. It is recommended that the default for the correct line speed range be taken for
this parameter.
Some recommended values for a selection interval timer are detailed in Table 4-3.

4-10

The formula used to find the values listed in Table 4-3 is:
8 bits per byte X 348 bytes per message + RTSjCTS delay =
data rate (bits
per second)

Timer (in seconds)

NOTE
Most modems include an RTS JCTS delay that must
be included in the computation of the value for the
selection interval timer. When operating with an external (EIA) modem, the typical delay used is 150
milliseconds. The delay used when operating with
the integral modem is 100 microseconds. In networks where traffic involves messages of 40 bytes or
less, two RTSjCTS delays must be used. The added
delay is necessary because a complete message of
this length can fit into the line unit transmit silo
causing the selection interval timer to be started before any characters are transmitted.
The values listed in Table 4-3 represent absolute minimums. In most events specific applications need
added delay time to prevent a timeout during reception of a valid message. Requirements for added
delay time can be caused by processing delays that occur when receiving from a non-DMP11 device or
by line delays found with satellite stations. When determining this value, the value chosen for the selection interval timer should represent the time that the system can reasonably expend waiting for a response from another station.
When used as a reply timer, the selection interval timer sets the maximum waiting period between sending a message and receiving an acknowledgement (ACK) before taking error recovery actions. This
timeout is necessary to recover from power outages and to prevent the protocol from becoming deadlocked.
The same criteria used to determine a value for a selection interval timer in multipoint networks are
also used to determine a value when this timer is used as a reply timer. As shown in Table 4-3, the
default value for both uses is 3 seconds.
Table 4-3
Bits Per
Second

Recommended Selection Interval Timer Values
Computed Timer Value
for a 256-Byte Message

4800

580 ms + 150 ms = 730 ms

---. 800 ms

9600

290 ms + 150 ms = 440 ms

---. 500 ms

56K

49.7 ms + 0.1 ms = 49.8 ms

---. 50 ms

250K

11.1 ms + 0.1 ms = 11.2 ms

---. 12 ms

The figures shown in Table 4-3 include the following overhead elements:
256 bytes of data.
28 bytes of header, sync, and pad characters.
64 bytes of extra overhead to allow for the emptying of the transmit silo.
Total 348 bytes
4-11

4.4.2 Setting the Babbling Tributary Timer
This user parameter is compatible with half-duplex and full-duplex, multipoint network control stations.
A babbling tributary is a tributary that continues to transmit valid DDCMP messages after an implementation dependent timeout, thereby taking control of the communications line and denying equal
access to other tributaries in the network. This situation is controlled by the babbling tributary timer
which monitors the total time a tributary continuously transmits without giving up the communications
line. When this period exceeds the timeout period of the babbling tributary timer, the user program is
informed through a control response having the code for a babbling tributary and the identification of
the offending tributary. When a babbling tributary is detected, the control station takes no action beyond this indication.
A main consideration in determining a value for the babbling tributary timer is the total time interval
that a given tributary will need to end a selection interval. Determining the value for this timer is similar to that for the selection interval timer in that the same range of elements are used as criteria for
computing the value. The main difference in the two decisions is that the total number of message bytes
should be used in babbling tributary timer parameter computations rather than the number of bytes in
the longest message.
A value for the maximum transmitted message count parameter must also be considered with the parameter for the babbling tributary timer. The user-defined parameter to set this counter places a limit
on the number of messages that a tributary can transmit during the period of the selection interval
counter. This is done by forcing the select flag (terminating the selection interval) when the count of
messages sent by a tributary equals the setting of the maximum transmitted message count. This count
relieves the user program from having to limit the number of messages queued for transmission in order
to avoid a babbling tributary condition.
In any event, the period set for the timeout of a babbling tributary timer should be long enough to be
certain that timer expiration indicates a definite error condition. In addition, the parameter assigned to
the maximum transmitted message count should also be considered when setting the period of the
babbling tributary counter.
4.4.3 Setting the Streaming Tributary Timer
A streaming tributary is a tributary station on a multipoint line (or point-to-point station) that continues
to assert the carrier signal on the link after it has given up the link. In normal operation, the link is
returned to the control station when it receives a select flag or when the period of the selection interval
timer is exceeded. A timeout of the streaming tributary timer indicates a potential jamming of the link
by a defective tributary station, a defective point-to-point station, or a malfunctioning modem.
The streaming tributary timer is started when the link is returned to the control station by the remote
station and stopped when the carrier is dropped by that station. When a streaming tributary is detected,
when the streaming tributary timer expires, the user program is informed in the same manner as a
babbling tributary. No further action is taken by the control station DMPII.
Determining a value for the streaming tributary timer needs consideration of such elements as modem
delays and settling time of the communications line. As with determining periods for the selection interval timer and the babbling tributary timer, the period specified for this timer should be long enough to
prevent an early running out of the timer. For most network applications the default of one second will
be enough.
4.4.4 Criteria for Determining a Value for XDT Preset
This user-defined parameter provides for a fixed delay between the transmission of each data or maintenance message from the control station to a tributary, causing data and maintenance message traffic
delays in one direction only. It has no effect on the polling rate for a tributary when the tributary is
polled using DDCMP control messages. DDCMP control messages are not subject to the XDT preset
delay, and tributaries usually respond immediately to a poll.

4-12

This delay is valuable in states where the remote station is a character-interrupt device (such as a
OUP1l) needing added processing time between receive messages or where the remote station has limited receive buffer resources.
NOTE
When selecting a value for this parameter, the user
should consider that this delay timer is incremented
by the value of the user-defined parameter DELTA
T, not by the system clock (a one millisecond timer).
The default value for this parameter is zero.
4.5 ERROR COUNTER ACCESS
As described in Section 5.3, the OMP1l has a large complement of error counters designed to isolate a
wide range of error conditions. The TSS for each assigned tributary has seven error counters with three
statistical counters that provide background information for error analysis. In addition, the GSS at each
station has four error counters that tabulate global errors. The three TSS statistical counters are 16 bits
long, and the threshold error counters are three bits long. The rest of the TSS/GSS error counters are
eight bits long.
4.5.1 Reading the Counters
Both TSS and GSS counters are accessed through the correct control command. The contents of the
counter or counters returned through an information response (see Sections 3.3.3 and 3.4.3). Through
the control command, the user program has the option of either reading or both reading and clearing
the counters. When doing error analysis, it is recommended that a user program read and clear these
counters to be certain of a zero-base for sampling of the counters. If copies of the counters are maintained in main memory, it is also recommended that counters be read and cleared.
NOTE
The three-bit threshold error counters are automatically reset when the maximum count is reached. Access to these counters is limited to reading only (see
Section 5.3.1.2).
The OMP11 error and statistical counter structure is designed to be a complete unit. As an example of
this structure, consider the data errors inbound counter and the data messages received statistical
counter. The data errors inbound counter tabulates the errors related to the validity of message reception such as block check errors. The data messages received counter records the total number of messages received. A ratio of message reception errors to total number of messages received can be taken
from these two counters.
4.5.2 Counter Skew
When performing error analysis, there is a potential of skew between counts because of read time delays and the requirement that counters be read one at a time. The chance of skew between counts is a
function of line speed: the higher the line speed, the greater the probability of a skew condition. An
example of this potential skewing is the possible difference between the number of selection intervals
and the number of selection timeouts. A skew can result from added selection timeouts occurring while
the counters are being read.
If there is a requirement that error/statistical counters be read without the potential for skew, this can
be done by halting the protocol at the tributary. With the protocol halted, the contents of the error /statistical counters in a TSS and in the GSS are frozen at the counts recorded when the protocol
was halted. The counters can then be read without the problem of skew because of read time delays.

4-13

4.6 ERROR RECOVERY PROCEDURES
Within a DMPll-implemented network, there are three basic levels of error recovery including the user
program:
1.

Procedural violations where the user program is only informed.

Recovery from errors that ask the user program to initiate protocol shutdown.

Fatal errors resulting in system shutdown with a minimum of notice to the user program.

As shown in Table 3-8, procedural error codes from 100 to 140 are reported to the user program with no
recovery needed. The last two procedural errors (codes 300 and 302) include error recovery levels two
and three. All network errors need recovery through protocol shutdown. However, after receiving the
control response, queue overflow can result in a network shutdown (see Table 3-7).
4.6.1 Recovery from Network Errors
In order to recover from network errors, the protocol must be halted at the tributary or station recording the error. Two similar but separate procedures are recommended for recovery from threshold errors
and babbling and streaming tributary errors. These recovery procedures are described in the sections
that follow.
4.6.1.1 Recovery from Threshold Errors - DMPII threshold errors are detailed in Section 3.4.2.3. and
5.3.1.2. The recommended recovery procedure that should be started by the user program at the station
recording the errors is listed below:
I.

Halt the protocol (see Section 3.3.3.1).

Read the error counters to determine the type and cause of the threshold error condition. If
the error results from a lack of receive buffers, correct the condition. If the transmit or selection threshold is exceeded, check the functional condition of the remote station.

When the conditions causing the errors are cleared, restart the protocol (see Section 4.3.3).

4.6.1.2 Recovery from Babbling and Streaming Tributary Errors - These errors result when the userdefined parameter (or default) specifying the period of the correct timer is exceeded. Therefore, a timeout can result from an actual error condition or from the period of the timer being too short for the type
of message activity on the line (see Sections 4.4.2 and 4.4.3). A recommended recovery procedure to be
used when finding these conditions is:
1.

Halt the protocol (see Section 3.3.3.1).

Check the value of the timer parameters and increase if the value is not correct.

Restart the protocol (see Section 4.3.3).

I f this error condition continues, reconfigure the station as specified by Section 4.3.1.

When the cause of the timeout is at the refl?otestation, action must be taken at the remote
station to determine and correct the fault. The local station is at fault only if the values of the
timer parameters are not correct.

4.6.2 Recovery from Procedural Errors
The two procedural errors that need a recovery procedure are:
1.

Non-existent memory error, and

Buffer too small error.

The recovery procedure for each of these errors is detailed in the sections that follow.
4-14

4.6.2.1 Recovery from a Non-Existent Memory Error - Non-existent memory errors occur when the
DMPII tries to access an allocated receive or transmit buffer having an invalid UNIBUS address.
When this error is detected, the D MP 11 sends a control response to the user program with the invalid
address (see Section 3.4.2.4). It is up to the user program to determine whether the non-existent address
is in a transmit or receive buffer.
NOTE
Depending on microcode processing situations, the
non-existent memory address returned to the user
program could have been incremented to the next sequential location.
The recommended recovery procedure for this error is as follows:
1.

Halt the protocol for the tributary or station that records this error in order to start the return
of all outstanding buffers (see Section 3.3.3.1).

If the error is in a buffer from the common pool, the user program should issue the global
halt command to start the return of all outstanding receive buffers from the common pool.

Restart the protocol and reallocate buffers as necessary. If this error occurs again, a possible
main CPU or DMPII malfunction may be present.
NOTE
If the network line speed is more than 56K bls, the
requests for retransmission generated by a non-existent memory address can result in the overflow of
the DMP11 response queue causing a fatal system
error (see Section 4.6.2.3).

4.6.2.2 Recovery from a Receive Buffer too Small Error - When the DMPII receives a message, it
first checks to see if a buffer from the common buffer pool linked list is available. If one is available, it
uses that buffer. If the common buffer pool is empty, has a quota of zero, or is not enabled, the private
buffer linked list is checked. If a private buffer is not available, the receiving station NAKs the incoming message. The steps taken by the DMPII microcode in this process are listed below:
1.

Is the message number in sequence? Yes, continue; no, ignore message.

Is the common buffer pool enabled? Yes, continue; no, go to step 6.

Does the common buffer pool quota = O? Yes, go to step 6; no, continue.

Is a common pool buffer available? Yes, continue; no, go to step 6.

Is the common pool buffer too small? Yes, go to step 8; no, use this buffer.

Is a private buffer available? Yes, continue; no send NAK - buffer temporarily not available.

Is private buffer too small? Yes, send NAK - buffer too small; no, use this buffer.

Is private buffer available? Yes, go to step 7; no, send NAK - buffer too small.

4-15

NOTE
The DMPll does not scan the common pool or private linked list structures looking for a buffer of acceptable size. Instead, it uses the next available buffer from the list.
Buffer too small errors apply only to receive buffers. The procedure for recovery from this error depends on whether the allocated buffer is from the common pool or is a private buffer. The recovery
procedures are detailed below:
A.

Common pool buffer too small:
I.

Assign a private buffer of acceptable size to the receiving tributary through a buffer address/character count command (see Section 3.3.4).

Both private and common pool buffers too small:
I.

Halt the protocol for the affected tributary in order to start the return of all outstanding private buffers.

Restart the protocol.

Assign a private buffer of acceptable size to the receiving tributary through a buffer address/character count command (see Section 3.3.4).

Private buffer too small and common pool not enabled
I.

If buffers from the common pool are available to other tributaries and are of acceptable size,
enable common pool buffers for this tributary (see Section 3.3.3).

If the common buffer pool is not in use for other tributaries, follow recovery procedure B.

4.6.2.3 Recovery from a Queue Overflow Error - This error is always fatal to the DMPll that recorded the error because it forces automatic shutdown of the device. The basic cause of this error is that the
link blocks from the free linked list are not available (see Section 5.4.1.1). Usually, this error results
when the internal response queue overflows because the DMPII generates responses faster than the
user program can recover responses from the queue. This error can also occur if a large number of
receive buffers are allocated. An example of the cause of the response queue running over is the occurrence of continuous running over non-existent memory errors in high . .speed networks (see Section
4.6.2.1).
When this error occurs, the DMPll lists the most current entry in the response queue to the user program. The user program then has three seconds after being interrupted to read the response. If it is read
during this three-second interval, the next response is listed, and the user program again has three seconds to read the next response. As long as the user program reads each response within three seconds,
the process can continue until the internal response queue is empty. These responses can then be studied to determine the cause of the queue overflow.
After the last response is listed, or the three-second response period ends, the DMPII shuts itself down.
At this point, to return the DMPII to functional status, the startup procedure must be started from the
beginning (see Section 4.3).
4.7 BOOTING A REMOTE STATION
DMPl1-implemented networks provide the user program at the multipoint control station or point-to-

4-16

point station with the ability to boot the main CPU at a remote station that has been shutdown because
of a power outage or software malfunction. There are three ways this boot function can be performed:
1.

Remote load detect: the control station starts the primary maintenance operation protocol
(MOP) boot procedure for a remote station. See Section 1.7 for the name, version, and order
number of the MOP specification.

Power on boot: the first poll received after power-up at the remote station causes the DMPll
at that station to ask the control station to start the primary MOP boot procedure.

Invoke primary MOP: the user program at the remote station causes, the DMPll to ask that
the control station start the primary MOP boot procedure.

NOTE
These procedures apply only to tributary stations in
PDP-II-based multipoint networks and to one node
in a PDP-II-based point-to-point link. VAX-II systems cannot be booted using these DMPII procedures.
Primary MOP boot procedures need the DMPll line unit switch-configured in the way specified in this
section. The steps taking place at the remote station and over the communications line leading to each
of the three primary MOP boot functions are shown in the sections that follow.
4.7.1 Steps Leading to a Remote Load Detect Boot
The steps taking place at the remote DMPll and the host PDP-II in response to an Enter MOP Mode
message from the control station are:
1.

The DMPll NPRs a tight loop routine into main memory.

The DMPll transfers control to the routine through the ac low vector. This routine removes
the CPU from the active state to prevent any intervention during the NPR process.

The DMPll then sends a primary MOP Request program message to the control station.
The control station responds in turn with a primary MOP Memory load with transfer address
message with the boot or related program to be loaded into the remote station main memory.

The DMPll NPRs that program into main memory and starts executing the program.

At this point, the remote station is operating in the way planned by the down-line loaded
program.

The steps occurring over the communications line during a remote load detect boot are:
1.

The control station sends an enter MOP mode message to a remote tributary station.

The remote station identifies the address and password in the message and then removes the
main CPU from the active state.

The remote station then responds with a primary MOP request program message.

The control station responds to this message with a primary MOP memory load with transfer
address message having the boot or related program to be loaded into the host PDP-II at the
remote station.

4-17

4.7.2 Steps Leading to a Power On Boot
When power is restored after a shutdown at a remote station, the DMPll performs the same steps
performed during a remote load detect boot. However, the first two steps performed over the communications line are omitted and the tributary station responds to the first poll from the control station
with a MOP mode request program message, the tributary station then continues the same sequence
used in the remote load detect boot procedure.
4.7.3 Steps Leading to an Invoke Primary MOP Boot
This boot operation starts when a user at a tributary or point-to-point station sets the boot and master
clear bits in the DMPll initialization register (see Section 3.3.1.1). The steps taken by the DMPII are
the same as with a power-on boot.
4.7.4 Line Unit Switch Settings for the Boot Functions
At remote stations in networks supporting the primary MOP boot functions, the line unit switches must
be configured in a specific way to correctly perform the boot functions (see Section 2.6).

NOTE
The switch setting procedures described below apply
only to tributary stations in a multipoint network
and to one node in a point-to-point network.
The four-bit unit number of each DMPII must be set in the line unit switches. This unit number allows
the boot program, once loaded into the host PDP-II, to identify the specific DMPII that is booting.
The line unit number is placed in switch pack E 121 with switch number 1 the LSB and switch number
4 the MSB.
As described in Section 3.3.2, the operating mode of a DMPll can be created through the mode definition command or by specific settings of the line unit switches.

NOTE
When primary MOP booting is supported in a network, the operating mode of each tributary station
must be set in the line unit switches instead of
through the mode definition command.
The operating mode of a DMPl1 is specified by setting the mode enable switch, switch number 8 of
switch pack E121, to one (OFF) and by setting switch numbers 5,6, and 7 of switch pack E121 to the
needed operating mode. The settings for these line unit switches are listed in Table 4-4.
4.7.4.1 Switch Settings for the Power On Boot Function - To enable the power on boot function at a
remote tributary station, switch number 10 of the line unit switch pack EI21 (power-on-boot) must be
set to zero (ON). In addition, the tributary address of this station must be set into line unit switch pack
E134. For the tributary address, switch number 1 is the LSB and switch number 8 is the MSB. If the
mode is point-to-point, the address switches can be ignored.
4.7.4.2 Switch Settings for the Invoke Primary MOP Boot Function - Line unit switch settings for the
invoke primary MOP boot function are exactly the same as those for the power-on boot function. To
perform this boot function, the user program at the remote station must set the boot and master clear
bits in the DMPll initialization register (see Section 3.3.1.1).
4.7.4.3 Switch Settings for the Remote Load Detect Boot Function - To enable the remote load detect
boot function at a remote tributary station, switch number 9 of line unit switch pack EI21 (remote load
detect) must be set to one (OFF). For the remote load detect boot function, the switch-assigned tributary address, which is included in the enter MOP mode message, also serves as the password. When using
the remote load detect boot function in point-to-point networks, the line unit tributary address switches,
for protection purposes, can be set to a unique value because the address of a point-to-point node is
always one.
4-18

Table 4-4

Mode Switch Settings

Mode
Switches

Line
Characteristics

Network
Configuration

DMCll
Line Compatibility

7
5
6
ON ON ON

Half-Duplex

Point-to-Point

Yes

OFF ON ON

Full-Duplex

Point-to-Point

Yes

ON OFF ON

Half-Duplex

Point-to-Point

OFF OFF ON

Full-Duplex

Point-to-point

ON ON OFF

Half-Duplex

Multipoint Control
Station

NjA

OFF ON OFF

Full-Duplex

Multipoint Control
Station

NjA

ON OFF OFF

Half-Duplex

Multipoint Tributary
Station

NjA

OFF OFF OFF

Full-Duplex

Multipoint Tributary
Station

NjA

4-19

CHAPTER 5
ASPECTS OF DMPll
MICROCODE OPERATION

5.1

INTRODUCTION

There are a number of aspects of the DMPII microcode operation that interface with a user program
and must be carefully considered in the design and operation of such a program. These include:
1.

Operation and use of the dynamic polling algorithm,

Error and statistical information recorded by the DMPII microcode for user analysis, and

Structure of the DMPII internal data base.

Each of these is detailed in the sections that follow.

5.2

DMPll DYNAMIC POLLING ALGORITHM

In multipoint networks implemented by two or more DMPII intelligent communications line controllers, polling frequency and priority for each assigned tributary are determined automatically by the
microcode polling algorithm and applied dynamically. The polling algorithm uses the following information to determine these elements:
1.
2.
3.

A tributary's poll response history,
A tributary's user-defined parameters, and
A tributary's protocol state.

The multipoint network control station polls all assigned tributaries that are not in the DDCMP halt
state. The protocol state of every assigned tributary is maintained in the TSS by the control station.
When a tributary is available for polling, all transmit messages for that tributary are sent as the poll, up
to the limit set by the maximum transmitted message count. If no transmit messages are available for a
tributary, the DMPII automatically transmits the correct DDCMP control message.
The DMPII polling algorithm determines which tributary is to be polled next, based on each tributary's polling urgency level. The DMPII polling algorithm uses the user-defined TSS and GSS polling
parameters for categorizing tributaries into polling levels and determining the rate at which polling
urgency is increased within each polling level. A tributary's polling level is based on its response history. This classification device, combined with a regular incrementing of polling urgency, results in the
most active tributaries being polled most often.

5.2.1

Polling States

As indicated in Section 3.3.3, the control station maintains the poll response history for a tributary by
continuously monitoring its polling activity and assigning it to a state correct to its response rate. These
states are:
1.

Active - the polling algorithm maintains a tributary as active when it responds to polls with
data messages.

5-1

Inactive - the polling state of an active tributary changes to inactive when it responds to a
consecutive number of polls with non-data DDCMP messages. The count of consecutive nondata messages received from a tributary in the active state is defined by the user-defined
parameter NDM-INACT (number of no-data messages needed to go inactive).

Unresponsive - a tributary, either active or inactive, is placed in the unresponsive state when
it fails to respond in any way to a consecutive number of polls. (Each poll results in a selection timeout.) ·The count of consecutive polls without responses is defined by the user-defined
parameter TO-UNRESP (number of timeouts to go unresponsive).

Dead - an unresponsive tributary continues to be unresponsive to consecutive polls until the
number of selection interval timeouts defined by the user-defined parameter TO-DEAD
(number of timeouts to go dead) is exceeded. Different from tributaries in the other polling
states, dead tributaries are always polled on a round-robin basis with the period between polls
determined by the user-defined global parameter DEAD-T (dead timer).

When specifying the parameters controlling polling levels, the user has the option of accepting the
defaults (see Table 4-3 and 4-4) or selecting specific values in place of these defaults. As described in
Section 3.3.3.1, the user program can set the polling state of a tributary to any state at any time by
issuing a latch polling state control command. The polling state set by a control command stays in effect, regardless of tributary performance, until polling control is given back to the polling algorithm by
the user program. This is done by issuing an unlatch polling state control command.
Figure 5-1 shows a state diagram describing the changes between polling states. The actual changes
depend on the particular polling parameters (see Section 4.3.2).

NOTE
A tributary in the inactive, unresponsive, or dead
polling state is immediately returned to the active
state when it responds to a poll with a valid data
message or when the user program allocates a transmit buffer to that tributary.
5.2.2 Computation of Urgency
Computation of a tributary's urgency is based on three user-defined polling parameters:
1.

Q - the starting value of polling urgency (U) for a tributary,

R - the rate by which urgency (U) is increased for a tributary, and

DELTA T - the polling algorithm global update interval

Figure 5-2 graphs the relation between these three parameters. This graph shows the effect of Q = 0, R
= 6, and DELTA T = 50 ms for a tributary. Figure 5-2 shows that the value of the urgency (U) axis
ranges from zero to 255. In addition, three base lines are shown having values for U of 127, 128, and
255. The values of these base lines have the following meaning:
1.

Zero to 127 - A tributary with an urgency in this range is never polled until its value of U
reaches 128.

128 - This is the threshold level at which a tributary becomes acceptable for polling.

129 to 254 - The tributary is available for polling when in this range. Within a group of
tributaries the tributary to be polled next is the tributary with the highest value of U.

5-2

255 - This value of U represents the highest polling urgency. A group of tributaries all having the maximum value of U would be polled on a round-robin basis.

During each DELTA T time period, the control station polling algorithm updates the urgency of each
functional tributary by adding the value of R for the correct polling state (dead) to the urgency value of
each tributary. This updating sequence is performed on the TSS data base in the order in which tributaries were created at the control station. When the polling algorithm determines that the next poll is to
be sent, it selects the tributary to be polled by scanning the TSS data base (in the order the tributaries
were assigned starting at the TSS following the last tributary polled. In this process, the tributary with
the highest value of U from among active, inactive, and unresponsive tributaries is selected as the next
tributary to be polled.
If an urgency of 255 is detected during the process of scanning the TSS data base for the tributary with
the highest value of U, the scan process halts and that tributary is immediately selected for polling.
Once the selected tributary is polled, its urgency goes back to the assigned value of Q for its polling
level.

As previously stated, dead tributaries are polled at a rate determined by the user-defined parameter
DEAD T. One dead tributary is polled at each running out of the DEAD T timer, and the scan of dead
tributaries is resumed from the last dead tributary polled.

Figure 5-1

State Diagram of Polling State Transitions

5-3

TIME TO
MAXIMUM URGENCY

r-------------------------------~-----------------------------255

------------------------- -----

ELIGIBLE FOR
POLLING

5"
u>-

128
::J 127
a:::

MINIMUM
POLLING INTERVAL

~-------------~--------------~

NOT POLLED

Il..

~ t =50ms
MK·1962

Figure 5-2

Relationship Between Polling Parameters Q, R, and Delta T

5.2.3 Criteria for Determining Polling Parameters
Although there are no absolute rules for determining specific values for the polling parameters, there
are general rules that provide a basis for getting acceptable values for these parameters. The user-defined polling parameters include the following:
1.

DELTA T - the GSS parameter for specifying the rate of urgency updating.

Q and R - the TSS parameters used in the computation of polling urgency.

POLL DELAY - the GSS parameter that specifies the delay to be used between sequential
polls.

DEAD T - the GSS parameter that specifies polling rate for dead tributaries.

Details for determining the values of these parameters are described in the sections that follow.

5-4

5.2.3.1 Criteria for Determining DELTA T - The 50 ms default value for DELTA T provides for
minimum polling intervals which range from 50 ms (Q = 0, R = the threshold value 128) to 6.4 seconds (Q = 0, r = 1). The default value of 50 ms is the smallest acceptable value for DELTA T and
represents the actual time needed for the polling algorithm to update the urgencies of 32 tributaries.

NOTE
The mlmmum polling interval defines the time
needed for a tributary to reach the urgency threshold
value of 128 and is the absolute minimum time period needed to become acceptable for polling. The
maximum polling interval for a tributary cannot be
determined because it is a function of such variables
as line speed, message traffic, the number of tributaries in a network, and the polling states of those
tributaries. In addition, these variables prevent the
determining of the point in time at which any tributary in a network will be polled.
For almost all multipoint network applications, the default value of 50 ms for DELTA T is not enough.
However, in specific instances (for example, a network formed by low traffic devices), a higher value of
DELTA T may be recommended.

5.2.3.2 Criteria for Determining Values of Q and R - For a given value of DELTA T, the minimum
polling interval is a function of the user-defined parameters Q and R. For example, if a minimum polling interval of 0.8 second is needed for a tributary (assuming a DELTA T of 50 ms), the parameters Q
= 0 and R = 8 fills this requirement (see Figure 5-3). As shown in Figure 5-3, the time to reach maximum polling urgency for these values of Q and R is 1.6 seconds. Figure 5-3 also shows that for R = 4,
the value of Q = 64 causes the minimum polling interval to stay at 0.8 second, but needs a correspondingly longer time (2.4 seconds) to reach maximum polling urgency.

NOTE
Reaching maximum polling urgency does not guarantee that a tributary will be polled. Instead, it represents maximum acceptance for polling.
When determining values of Q and R for tributaries in a multipoint network, a minimum polling interval and a time to maximum urgency must be determined for each tributary in each of the polling states
(active, inactive, and unresponsive). In each state, these values are a function of the total amount of
data traffic expected from a tributary while in a specific polling state. The primary goal when determining the values of Q and R for each tributary should be to make sure that the tributary having the
most regular data message response rate during past polling activity will be the tributary most often
polled.
Default values for the parameters Q and R are based on line speeds in the one megabyte range. The
characteristics of the defaults are very short minimum polling intervals (0 to 0.3 second) and times to
maximum urgency (50 ms to 0.65 second).
Figure 5-4 graphs the relationship between the default values of Q and R for each of the three polling
states. When all tributaries in a network have the default value for Q and R and all tributaries are in
the active polling state, polling is round-robin.

5.2.3.3 Determining a Value for Poll Delay - The user-defined global parameter, poll delay sets a
fixed delay between control station polls, providing a device for message traffic without changing the
values of Q and R for individual tributaries. During this delay, transmission from the control station to
5-5

255

127~----------~~----------~r------------.-------------------------------

(!)

::J
...J

Q..

50ms

0.5

0.8

1.0

2.0

1.6
DELTA

2.4

3.0

4.0

50ms
MK·1975

Figure 5-3

Relationship Between Polling Parameters Q, R, and the Minimum Polling Interval

tributary stations is halted for the interval defined by the poll delay timer. This interval starts when the
tributary polled deselects itself.
The ability to control message traffic through a single parameter is valuable in multipoint networks
where DMPlls are configured together with slower character interrupt communication devices, such
as the DUPl1. The value selected for poll delay in these instances is a function of character handling
rates of the non-DMPll devices.

In remote mUltipoint networks where the distance between the control station and each differs tributary
station differs significantly, the resulting differences in communication line settling time can increase
transmission and reception errors. In such instances, the settling time for the most distance tributary
station should be used for determining a value for this parameter.
For DMPll-implemented, high-speed, local networks, this parameter is not necessary. In such instances, the default value for POLL DELAY, which is zero, should be accepted.
5.2.3.4 Determining a Value for DEAD T - This global parameter (which is the period of the dead
tributary timer) sets the rate at which dead tributaries are polled. Dead tributaries are polled on a
round-robin basis, with one tributary polled at each termination of the dead tributary timer.

5-6

255r------------.r---------------------------------------------~r_---

ACTIVE 0=255,

R=O

ELIGIBLE FOR
POLLING

s::;:u

(!)

~ 128
(!)

::J
..J

oa..

NOT POLLED

O~---r----r_--_r----r_--_r----~--_r--~~--~----r_--_r--~~~~

50ms

0.3 sec

0.65 sec

DELTA T =50ms

Figure 5-4

Relationship Between the Default Values for Q and R for the Three Polling Levels

Polling dead tributaries can include a significant impact on network line use. In general, the shorter the
period of this timer, the larger the impact on line use. Also, for a given value of DEAD T, the impact on
line use will drop as system line speed increases. In determining a value for this parameter, the primary
goal should be to limit the impact of dead tributary polling on network line use.
Although the best rate of dead tributary polling will differ from application to application, the value
determined for DEAD T should be based on the period of the selection interval timer for the specific
application. For example, if the period of the dead tributary timer is equal to that of the selection interval timer, only dead tributaries are polled. For most system applications, it is recommended that the
period of the dead tributary timer be three to 10 times larger that of the selection interval timer, depending on system line speed. The default value for DEAD T is 10 seconds. This is approximately three
times larger than the default value for the selection interval timer.

5-7

5.3 ERROR RECORDING
Error recording by the DMPll is based on the use of a standard set of DDCMP counters built so that
specific subsets of these counters apply directly to the cause of specific error conditions. The complete
set of counters include counters specific to tributary error conditions, counters for recording global errors, and counters that provide a statistical base for conducting error analysis.
In a multipoint network, errors are recorded by these counters at the control station for all assigned
tributaries in the network. Identical sets of counters are maintained at tributary stations, one set for
each tributary assigned at a station. As a result, user programs at any DMPll in a network can detect a
malfunction and determine total error rates by accessing the correct TSS error counters.
When a station detects an error in a received message, it updates the counter and transmits a DDCMP
negative acknowledge message (NAK) to the transmitting station to inform it of the error. A NAK has
tributary address and reason code fields that identify the source and the reason for the NAK.
In general, NAKs are sent by data receivers to data senders when the data message transmitted by the
sender was received in error. By recording both NAKs sent and received, each station in a network is
capable of compiling error statistics on its tributaries. For more detail on NAKs, NAK reason codes,
and D DCMP message types, see Appendix A.
5.3.1 Error Counter Structures
There are three types of error counters maintained by the DMPll:
1.

Data link error counters - These counters record errors about the physical communications
line on a control/tributary station pair basis.

Threshold error counters - These counters record the presence of continuous errors that are
reported when a specific threshold has been exceeded. These counters also apply to a control/tributary station pair.

Station error counters - These global counters record errors specific to the physical communications line.

Statistical information is recorded for each tributary for use in studying error counts. Three counters
provide this information:
1.
2.
3.

Data messages transmitted counter,
Data messages received counter, and
Selection interval counter.

Figure 5-5 shows a map of both error and statistical counters as they appear in a TSS structure. Figure
5-5 shows that certain counters report a classification of errors of a single type of error and that flags
for these counters record the occurrence of the specific errors within the class. Each of these counters
can be considered a group counter that records each occurrence of the errors defined by the flags. The
function of the flags is to provide an indication of the type of errors recorded by the group counter.
Through the use of sampling methods (for example, reading and clearing a group counter and flags at a
specific frequency) it is possible to solve errors recorded by the group counter to a specific error type
within the class.
Reading or reading and clearing a counter is done through the read and read and clear TSSjGSS control commands (see Section 3.3.3.1). These commands access the TSSjGSS locations shown in Figures
5-5 and 5-6 on word boundaries only. The content of the TSSjGSS location is returned to the user
program through an information response (see Section 3.4.3).

5-8

TRIBUTARY STATUS SLOT (TSS)
ADDRESS (OCTAL)

RESERVED
RECEIVE THRESHOLD ERRORS

TRANSMIT THRESHOLD ERRORS
SELECTION THRESHOLD ERRORS

7
I--

DATA MESSAGES TRANSMITTED

DATA MESSAGES RECEIVED

SELECTION INTERVALS

DATA ERRORS OUTBOUND
RESERVED

OREP

LBTS

LBTU

RBTS

RBTU

IRTS

NRTS

REMOTE BUFFER ERRORS

SELECTION TIMEOUTS
RESERVED

OHBCC

LOCAL BUFFER ERRORS

RESERVED

IIREP IIDBCC IIHBCC

RESERVED

ODBCC

DATA ERRORS INBOUND
RESERVED

I I

LOCAL REPLY TIMEOUTS
REMOTE REPLY TIMEOUTS

MK-1960

Figure 5-5

TSS Map of Error and Statistical Counters

5-9

GLOBAL STATUS SLOT (GSS)
ADDRESS (OCTAL)

REMOTE STATION ERRORS

I
16

RSTR

RSEL

LUNDR

RMHFE

ROVRN

LOCAL STATION ERRORS

LOVR

LMHFE

LOVRN

GLOBAL HEADER BLOCK CHECK ERRORS

MAINT. DATA BLOCK CHECK ERRORS
MK-1959

Figure 5-6

GSS Memory Map

NOTE
Except for the clearing action that results from a
read and clear control command, TSS and GSS
counters are cleared when the user program master
clears a DMPll. Also, when a tributary is created,
the TSS counters for that tributary are cleared.
However, placing a tributary in the DDCMP HALT
or ISTRT states has no effect on these counters.
5.3.1.1 Data Link Error Counters - Counters in this class record physical communications line errors
using message traffic between a control and tributary station pair (or two point-to-point stations). Data
link error counters are eight-bit counters which, if not cleared, latch at the maximum count of 255. All
counters in this class can be read or read and cleared by the user program by means of the control
command and information response. The errors recorded by the seven counters in this class are:
1.
2.
3.
4.
5.
6.
7.

Data errors outbound,
Data errors inbound,
Local buffer errors,
Remote buffer errors,
Selection timeout errors,
Local reply timeout errors, and
Remote reply timeout errors.

The counters that record these errors are described in detail under the headings that follow.
Data Errors Outbound Counter
This group counter records communications line errors for message traffic outbound from this station.
The three flag bits for this counter indicate the types of outbound errors being recorded.
OHBCC -

Outbound Header Block-Check Error. This flag is set when a NAK specifying a data or
control message header block-check error is received.

OBDCC -

Outbound Data Field Block-Check Error. This flag is set when a NAK specifying a data
field block-check error is received.

5-10

OREP

Outbound REP Response. This flag is set when a NAK is received in response to a
DDCMP REP message sent from this station (see Appendix A).

Data Errors Inbound Counter
This group counter records communications line errors for message traffic inbound to this station. The
three flags for this counter indicate the types of inbound errors being recorded.
IHBCC

Inbound Header Block-Check Error. This flag is set, while in the DDCMP run state,
when a message with a header block-check error is received at this station. When this
error occurs, the receiving station transmits a NAK to the sending station. If an error of
this type occurs when this station is in a protocol state other than run, the error is recorded as a global header block-check error (see Section 5.3.1.3). Point-to-point stations
in the run state also record this error. Multipoint control stations record this error for
the selected tributary, regardless of the address field in the receive message. Multipoint
tributary stations record this error only if the address field matches their station address.

IDBCC

Inbound Data Field Block-Check Error. This flag is set when a data message with a
data field block-check error is received at this station. The receiving station transmits a
NAK to the sending station each time this error occurs.

IREP

Inbound REP Response. This flag is set when a NAK is sent in response to a REP message received by this station (see Appendix A).

Local Buffer Errors
This group counter records each failure of the user program at this station to allocate receive buffers of
acceptable size or quantity to store data messages from the remote station. The two flags for this
counter report the type of buffer allocation error recorded.
LBTU

Local Buffer Temporarily Unavailable. This flag is set when a buffer is not available for
a message received by this station. This condition results in a NAK being transmitted
by this station to the sending station.

LBTS

Local Buffer Too Small. This flag is set when a message is received by this station and
the available receive buffer is too small to store the message. This condition results in a
NAK being transmitted by this station to the sending station.

Remote Buffer Errors
This group counter records each failure of the user program at the remote station to allocate buffers of
acceptable size or quantity to store data messages sent by this station. The two flags for this counter
report the type of buffer allocation error recorded.
RBTU

Remote Buffer Temporarily Unavailable. This is set when a NAK is received indicating
that a buffer is not available at the remote station for a data message sent by this station.

RBTS

Remote Buffer Too Small. This flag is set when a NAK is received indicating that the
buffer at the remote station is too small for a data message sent by this station.

Selection Timeout Counter
At multipoint control stations and half-duplex point-to-point stations, this group counter records each
selection timeout incurred by the tributary (see Appendix A). It is not necessary for multipoint tributary stations to maintain selection timeout counters. In addition, the two stations in a full-duplex, point-topoint network are continuously selected, ending the need for this counter. The two flags associated with
this counter report the type of selection timeout error recorded.

5-11

NRTS

No Reply To Select. This flag is set to record the running out of the selection timer
without the receipt of a valid message or header, or without the detection of an attempted transmission.
NOTE
The DMPll detects attempted transmissions by
monitoring the modem signal carrier detect or
through receiving a DDCMP synchronization sequence.

IRTS

Incomplete Reply To Select. This flag is set when a selection interval timeout occurs,
preceded by receiving a valid message header not having the select bit set, or the detection of an attempted transmission from the remote station.
NOTE
At full-duplex, point-to-point stations, the selection
interval timer is used as a reply timer. Therefore, at
these stations this counter is not used.

Local Reply Timeouts
This counter records the sending of a DDCMP REP message to a remote station following the timeout
of the reply interval at this station. A REP message is sent when data messages transmitted to the
remote station are not acknowledged (ACKed).
Remote Reply Timeouts
This counter records the transmission of ACKs sent in response to a REP message received. NAKs sent
in response to REP messages received are recorded as data errors inbound. A remote station sends REP
messages to this station to indicate that it did not receive ACKs for data messages sent. This station
received the messages, but the sending station did not receive the ACK sent in response to the data or
REP messages.
5.3.1.2 Threshold Error Counters - Threshold error counters provide a user program with information
on consecutive or continuous errors occurring on the physical communications line or at remote stations. Figure 5-5 shows a map of these counters as they exist in a TSS structure. The three-bit threshold
counters record the following error conditions:

A message is transmitted but not acknowledged (transmit threshold counter).

A message is received incorrectly and NAKed (receive threshold counter).

A selection interval timer runs out (selection interval threshold counter).

When a threshold counter collects a consecutive count of seven, a control response is sent to the user
program with information that the specific threshold error and the counter are cleared. This action
provides the user program with information on error conditions that interrupt data transmission.
At multipoint network control stations, all three counters are maintained in the TSS structure for each
assigned tributary. At tributary stations, only the transmit and receive threshold counters are maintained. In point-to-point networks, one set of threshold counters is maintained at each station. The functions of the three error counters in this class are described under the headings that follow.
Transmit Threshold Counter
Transmit threshold errors recorded by this counter differ, based on the protocol state of the tributary or
station. At a control or tributary station, the detection of any consecutive combination of the following
conditions increments this counter:

5-12

This station is in the ISTRT state and it sends a STRT message.

This station is in the ASTRT state and it sends a STACK message.

While in the run state, a NAK is received at this station with a reason code other than 3 (see
Appendix A).

While in the run state this station sends a REP message.

This counter is cleared when this station or tributary enters the ASTRT, ISTRT, or run state. When in
the run state, this counter clears before reaching a count of seven for the following reasons:
1.

On successful transmission of a message.

On receipt of an acknowledgement when no messages are left.

Receive Threshold Counter
In the RUN state, the receive threshold counter records the sequential occurrence of receive message
errors for its tributary up to a maximum count of seven. This counter is incremented when a NAK is
sent for anyone of the following reasons:
1.

A message header block-check error.

A message data field block-check error.

A NAK response to a REP message.

A buffer is temporarily not available.

A receive message overrun.

A message is too long for the available buffer.

A message header format error.

NOTE
These receive error conditions agree with DDCMP
NAK reason codes.
There are three instances that cause this counter to be cleared before reaching a count of seven:
1.

The ISTRT, ASTRT, or run state is entered

A valid control message is received

A valid data message is received in the RUN state

Selection Threshold Counter
This counter records each consecutive end of the selection interval timer. In multipoint networks, this
counter is implemented only at the control station, one for each assigned tributary. For half-duplex,
point-to-point networks, one counter is implemented at each station. This counter is cleared when a
valid message is received having the select bit set.

5.3.1.3 Station Error Counters - Station error counters are 8-bit global counters that record the following types of general error conditions:

5-13

A hardware or software fault at a remote station.

A hardware or software fault at the local station.

Data errors occurring on the physical communication line that are not detected by a header
block-check.

The names of the counters that record these error conditions are:
I.
2.
3.
4.

Remote station error counter,
Local station error counter,
Global header block-check error counter, and
Maintenance message data field block-check error counter.

Figure 5-6 is a memory map showing the location of these counters in the GSS. All station counters are
eight-bit counters which if not cleared, latch at the maximum value of 255. Figure 5-6 shows that the
local and remote station error counters are group counters configured with sets of error flags. The group
counters and flags record errors in the same way as data link error counters and can be use in a similar
way for error analysis (see Section 5.3.1).
Station error counters are found in the global status slot (GSS) at each DMPII. The user program has
the option of either reading or reading and clearing these counters using the control commands read
GSS or read and clear GSS with a tributary address of zero. Accessing GSS locations is done on twobyte boundaries.
Remote Station Error Counter
This group counter records errors caused by a fault at a remote station or by data errors not seen on the
communications line inbound to this station. The specific errors recorded and the functions of the error
flags are described below.
ROVRN -

Remote Receive Overrun. This flag is set when a NAK is received indicating that a
receive overrun occurred at the remote station.

RMHFE -

Remote Message Header Format Error. This flag is set when this station detects a header format error in a message from a remote station. The remote station is informed of
this error by a NAK sent from this station.

RSEL

Remote Selection Address Error. This flag is set only at multipoint control stations. It
records receiving a message containing an address that does not match the address of
the tributary selected.

RSTR

Remote Streaming Tributary. This flag is set only at multipoint control stations and
half-duplex, point-to-point-stations. It records the occurrence of one of the following
events:
1.

The communications line not released by the remote station after the babbling
tributary timer ran out (see Section 4.3.2).

The communications line not released by the remote station at the end of the
selection interval (see Section 4.3.2).
When this flag is set, a control response having the correct error condition code is
passed to the user program (see Section 3.4.2).

5-14

Local Station Errors
This group counter records global errors caused by a fault at this station or by a data error not seen on
the communications line outbound from this station. The specific errors recorded and the functions of
the error flags are described below.
LOVRN -

Local Receive Overrun NAK Sent. This flag is set when a receiver overrun occurs at
this station while in the DDCMP run state. A NAK is sent to inform the remote station
of the error.

LOVR

Local Receive Overrun, NAK Not Sent. This flag is set when a receive overrun occurs
at this station and is not NAKed. Failure to NAK a receive overrun at the local station
can be caused by:
1.

A receive overrun at this station while it is receiving a message header or,

A station that is not in the DDCMP run state

LUNDR -

Local Transmit Underrun. This flag is set when a transmit underrun occurs at this station.

LMHFE -

Local Message Header Format Error. This flag is set when a NAK is received at this
station indicating a message header format error.

Global Header Block-Check Error Counter
This counter records the occurrence of header block-check errors at all stations not in the DDCMP run
state. This counter records header block-check errors for maintenance messages and for messages that
are received with a station address that does not match the address of this station. If these types of
errors occur when the station is in the DDCMP RUN state, they are recorded as inbound header blockcheck errors.
Global Maintenance Data Field Block-Check Error Counter
This counter records data field block-check errors occurring in maintenance messages and provides the
mechanism for detecting maintenance message data fields, which are not valid, during operations in the
DDCMP maintenance state.
5.3.1.4 Statistical Counters - These Counters provide statistical information for evaluating the error
counts recorded by the DMP11. They are maintained in the TSS for each assigned tributary as well as
in the single TSS at each point-to-point station. All statistical counters are 16 bits wide and, if not reset,
will latch at the maximum value of 65,535. The statistical counters are:
Data Messages Transmitted Counter
This counter records data messages transmitted by this station following protocol startup. The count
can be used as a statistical base when evaluating the number of data errors outbound, local reply timeouts, and remote buffer errors. This counter records the total number of messages transmitted since
startup while in the run state. This count does not include attempts to retransmit messages.
Data Messages Received Counter
This counter records the total number of data and maintenance messages received by this station following protocol startup while operating in the DDCMP run and maintenance states. The count can be
used as a statistical base when evaluating the number of data errors inbound, remote reply timeouts,
and local buffer errors.
Selection Interval Counter
At multipoint control and point-to-point stations, this counter records the number of messages transmitted with the select bit set. This count is maintained in all states of network operation and includes all

5-15

valid DDCMP messages transmitted with the select bit set. The counter can be used as a statistical
base to determine the ratio of selection timeouts to the total number of selections made. Also, for a
multipoint control station, this counter records each poll made.
5.4 STRUCTURE OF THE DMPll INTERNAL DATA BASE
The DMPII internal data base provides the device for controlling the assignment and completion of
transmit and receive buffers, queuing of DMPII responses, and assignment of TSS structures to assigned tributaries for the storage and maintenance of tributary and global status information. A map of
this data base, which is maintained in the 4K-byte data memory, is shown in Figure 5-7.
This internal data base is implemented by three basic structures:
I.
2.
3.

Linked lists,
Slot mapping Table, and
TSS and GSS structures.

Each of these data base structures are described in the sections that follow.
5.4.1 Linked Lists
A linked list is an open-ended data list made up of fixed length blocks linked by pointers. Figure 5-8
shows the standard format for DMPII linked list structures. Each of these blocks, called link blocks,
have seven bytes of data and a one byte pointer to the next link block in the list. The pointer is in the
last block in the list and is a terminator.
The DMPII linked list structure is made up of three types of linked lists. These are:
1.

The fn!e linked list - The free linked list is a list of empty link blocks used by the DMPII
microcode to form the types of linked lists left.

The response linked list - Responses for the user program are queued on this list for posting.

Buffer linked lists - These lists have the accessing information for each receive and transmit
buffer allocated, based on one link block per allocated buffer.

Each type of linked list is described in detail in the sections that follow.
5.4.1.1 The Free Linked List - The free linked list is the source from which all other linked lists get
link blocks. The addition of a link block to the receive, transmit, or the common pool buffer linked list is
started by the user program through a buffer address/character count command (see Section 3.3.4).
Also, the addition of a link block to the response linked list is started by the DMPll when a response is
posted to the user program (see Section 3.4).
Access to a linked list is controlled by two list pointers: one pointing to the start of the list and one
pointing to the end of the list. When a link block is removed from the free linked list, the start of list
pointer is changed to point to the next available link block in the free linked list. When a link block is
completed by one of the functional linked lists, it is added to the end of the free linked list. The start
and end of the list pointers for the free linked list are maintained in the station GSS.
The 1720 bytes allocated to the free linked list translate into a total of 215 link blocks available for use
by the functional linked lists. As a result, the free linked list functions as a finite resource for the functional linked lists. When the last link block is removed from the free linked list, the start of list pointer
is set to the terminator value of octal 377 in order to indicate that there are no more link blocks available. In this instance, the next request by the response linked list for a link block generates the fatal

5-16

error queue overflow. Given a fatal error condition, the buffer allocation plan for a user program should
be designed to be certain that enough link blocks are always available for use by the response linked list
maintained at the station controlled by that program (see Section 3.4.2.1).

5.4.1.2 The Response Linked List - This linked list functions as a queue of buffer completion, control,
and information responses to be posted to the user program. When preparing a link block to conduct a
control or information response, the DMPII clears all bit position that are not used in the link block to
zero. However, link blocks restored to the free linked list stay the same. The start and end of the list
pointers for the response linked list are maintained in the GSS for the station.

5.4.1.3 Buffer Linked Lists - A buffer linked list is provided for each type of message buffer allocated
by a user program. These are:
1.
2.
3.

Common pool receive buffers,
Private receive buffers, and
Transmit buffers.

Each link block in a buffer linked list provides the location and size of a buffer in main memory.
STARTING
ADDRESS
(OCTAL)

OOOOP-------------------------------------~

LINKED
LISTS

215 ENTRIES
8 BYTES PER ENTRY

3270~------------------------------------~

GLOBAL STATUS SLOT

72 BYTES

3400~------------------------------------~

SLOT MAPPING TABLE (SMT)
256 BYTES
4000~------------------------------------~

TRIBUTARY
STATUS
SLOTS
(TSS)

32 ENTRIES
64 BYTES PER ENTRY

7777~

____________________________________~
MK-2389

Figure 5-7

Map of DMPII Data Memory

5-17

LINK BLOCK
POINTER A

START OF LIST
POINTER

DATA

LINK BLOCK
POINTER B

..-

DATA

LINK BLOCK
POINTER C

DATA

.0
• E

377

TERMINATOR

DATA

MK·1961

Figure 5-8 DMPII Linked List Structure Format

5-18

Common Pool Buffer Linked List
Controlling receive messages through a common pool of buffers provides an efficient method of controlling buffer allocation. The common pool buffer linked list provides a queue of receive buffers available to all assigned tributaries according to the quota assigned to each tributary.
Common pool buffers are assigned through the buffer address/character count command and enabled
with tributary quota assignments through the control command (see Sections 3.3.4 and 3.3.3). The
start-of-list and end-of-list pointers for this linked list are maintained in the station GSS.
Receive Buffer Linked List
This linked list operates as a queue of private receive buffers with a separate list maintained for each
tributary assigned on a multipoint network. For point-to-point stations using private receive buffers, one
list is maintained at each station. The start-of-list and end-of-list pointers for each receive buffer linked
list are maintained in the tributary's (or station's) TSS structure.
Transmit Buffer Linked Lists
This list functions as a queue of transmit buffers, with one list maintained for each tributary assigned
on a multipoint network. For point-to-point stations, one list is maintained at each station. The start-oflist and end-of-list pointers for each transmit buffer linked list are maintained in the tributary's (or
station's) TSS structure.
A unique feature of the link block for this list is that it has a message number field. The number in this
field is the DDCMP message sequence number from the header of the message transmitted out of the
buffer described by the specific link block. The DMPII microcode uses this field to find the buffer for
a message that has been NAKed (and therefore must be sent again).
5.4.2 Slot Mapping Table
Under DDCMP, the 8-bit message header address field permits no more than 255 unique tributary
addresses in a multipoint network. However, the DMPII limits the number of assigned tributaries to
32. In order to implement DDCMP, a tributary in a DMPll-implemented multipoint network can have
a TSS address in the range of one to 255 (zero is reserved to address the aSS).
TSS addresses are assigned at both control and tributary stations through the slot mapping table
(SMT). As shown in Figure 5-7, this table occupies 256 (octal 400) locations in DMPII data memory,
with one location for each of the 255 potential tributary addresses.
The function of the SMT is to map an eight-bit tributary address into one of the 32 available TSS
structures. When a tributary is deleted, (see Section 3.3.3.1), its TSS and SMT entry is available for
assi~nment again. Also, when 32 tributaries are assigned and a try is made to create a 33rd, a procedural error is posted to the user program (see Section 3.4.2.4).
5.4.3 TSS and GSS Structures
Figure 5-7 shows that the TSS and ass structures occupy separate sections of the DMPII data memory. The ass is a single 72-byte structure and the 32 TSS structures are made up of 64 bytes each.
Together, the TSS and ass structures occupy a little over half of the DMPll data memory. These
structures are described in the sections that follow.
5.4.3.1 The Global Status Slot - A detailed map of the ass is shown in Appendix F. This map shows
that 64 of the 72 bytes in this slot have global control and status information with the- other eight bytes
used as a storage area for subroutine vectors.
The ass is used to maintain control and status information specific to the operation of the DMPll
microcode, to record event counts and error conditions that are global and to store global parameters.
Most of the ass is reserved for microcode control and status information.

5-19

Access to the GSS is done on two-byte boundaries. With the exception of an eight-byte storage area, a
user program can read any GSS location at will through the control command, with the contents of the
addressed location returned to the user program through an information response (see Sections 3.3.3.1
and 3.3.4.1). A user program can read and clear only the GSS station error counters (see Section
5.3.1.3) through the control command and information response. Through the control command, a user
program can write the five global parameters (see Table 4-4) into GSS octal locations 33 through 37.
5.4.3.2 Tributary Status Slots - At a multipoint control station, a separate TSS is maintained for each
assigned tributary in the network. At a tributary station, a separate TSS is maintained for each tributary assigned at that station. In a point-to-point network, one TSS is maintained at each station. When a
tributary is deleted, its TSS is available for assignment again. However, when the protocol is halted at a
tributary, or placed in the maintenance, ISTRT or ASTRT protocol states, the TSS stays complete.
Through use of the control command, all TSS locations can be read by the user program and specific
locations can be written into or read and cleared. When a location is to be read or read and cleared, the
data read is posted to the user program through an information response (see Sections 3.3.3.1 and
3.3.4.1 ).
A detailed map of a TSS is presented in Appendix G. This map shows that a TSS has four general types
of tributary-specific information:
1.
2.
3.
4.

Protocol and tributary status,
Error and statistical counters,
Message exchange variables, and
Polling parameters.

A description of each of these types is given in the sections that follow. In addition, the information
types that apply to a user program are referenced to that section of the manual where the topic is explained in detail.

Protocol and Tributary Status
This type includes information on the tributaries protocol state, its status with relation to the communications line, its protocol status, and its polling status. Tributary polling status is maintained only at a
multipoint network control station. Because this information is only related to and used by the DMPII
microcode, it can be read by a user program.

Error and Statistical Counters
These counters provide the user program with a wide range of error counts and a set of statistical counts
that permit analysis of the relation of specific error counts. These counters can be read and cleared by
the user program. The function of each of these counters are described in detail in Section 5.3.

Message Exchange Variables
This type includes a range of variables maintained and used by the DMPII microcode to control the
sequential exchange of message data. Each TSS also maintains a group of timers that can be set to a
specific timeout value by the user program. This group of timers directly affect message traffic conducted over the communications line. These timers are:
1.
2.
3.
4.

Transmit delay timer,
Selection interval timer,
Maximum transmit message count, and
Babbling tributary timer.

The link management functions performed by these timers is detailed in Section 4.3.2. Note that a
timeout value for the selection interval timer is stored in each tributary's TSS. The actual timer is
maintained in the GSS.

5-20

Polling Parameters
These parameters are user-defined values used by the polling algorithm in conducting dynamic polling
activity in mUltipoint networks. The functions performed by the DMPll polling algorithm and the criteria for determining the values for each parameter are described in detail in Section 5.2.

5-21

CHAPTER 6
SERVICE

6.1 SCOPE
This chapter provides information for servicing the DMPII. It includes the maintenance philosophy,
maintenance functions, troubleshooting methods in a multipoint environment, preventive maintenance,
and corrective maintenance. The section on troubleshooting in a mUltipoint environment includes:
•

The general approach to multipoint troubleshooting.

•

Some common problems with different multipoint network configurations.

•

The use of error counters and other information for isolating problems to a specific part of
the physical link.

The corrective maintenance section has short descriptions of the diagnostics associated with the
DMPII for VAX-II and PDP-II systems.

6.2 MAINTENANCE PHILOSOPHY
The field replaceable unit (FRU) for the DMPII is either a defective module (M8207-Y A microprocessor or M8203 line unit) or cable. Training of field service personnel is applied to functional and
application troubleshooting, using diagnostics for fault isolation to the FRU. Also, spare parts for module repair are not stored in the field. Finally, typical applications of the DMPII do not permit long
troubleshooting sessions and component troubleshooting/repair needs, at least, a I6-channel logic
analyzer.
CAUTION
When inserting or removing the M8207-YA microprocessor module, be sure not to move the priority
plug or control read-only memories (CROMs).
Always make sure that the CROMs are set firmly and in the correct socket. Otherwise, erratic operation of the DMPII may result.

6.3 TROUBLESHOOTING METHODS FOR MULTIPOINT
Some complex multipoint network configurations have a potential for wasting valuable time when trying to isolate a problem. For this reason, troubleshooting techniques for multipoint networks differ from
those for point-to-point networks.
The following sections describe these troubleshooting techniques for:
•
•

Approach
Error counters

6-1

6.3.1 Approach
Before trying any corrective measures, it is important to get some basic information about the network
configuration and the type of problem. This information can be found by asking the user and by referring to
the topography diagram for the network. The topography diagram is generated at installation time and is
maintained by the field service representative. The flowchart (Figure 6-1) shows a typical approach to
troubleshooting from the time a service call is placed until corrective maintenance starts. This procedure
should be followed to help isolate a failing tributary before a technician is sent to a site.
6.3.2 Error Counters
In multipoint networks, many tributaries tie to the same transmission line. Because of this it is more
difficult to determine which link, if any, is causing errors. To aid in troubleshooting, the DMPll uses
error counters. Every OMPll in the network (the control station and each tributary) uses error
counters to record errors. This allows user programs at any DMPll in the network to determine error
rates and detect a malfunctioning link.
The main way in which errors are indicated to the DMPll is by DOCMP negative acknowledge messages (NAKs). Each NAK has an address field and a reason code that identifies the source and reason
for the NAK. In general, NAKs are sent by the station receiving a message to the station which sent the
message when an error is detected in the received message. By recording NAKs sent and NAKs received, each point or tributary in the network can compile statistics on the condition of the link created
between the two stations. DDCMP has been designed so that even if one of the stations on the link
cannot record errors, the other station may be used to record errors for all communications in both
directions on the link.
There are three main categories of error counters used by the DMPll: data link counters, station
counters, and threshold counters. Data link counters and threshold counters are maintained for each
tributary/ control station pair on a physical link. These counters are found in the tributary status slots of the
data memory (Figures 6- 2 and 6-3). Station counters are maintained for the complete physical link and are
found in the global status slot of the data memory (Figure 6-4).
The information found by checking errOr counters may help in identifying a problem area. For detailed
information on error counters, refer to Chapter 3.

6.3.2.1. Data Link Counters - Data link counters are of two types: cumulative and background. The
cumulative counters are 8-bit registers which latch at 25510. The background counters are 16-bit registers which latch at 6553510. The cumulative data link counters record and total occurrences of an error
and group them into the following classes:
•
•
•
•
•
•
•

Data errors outbound,
Data errors inbound,
Local reply timeouts,
Remote reply timeouts,
Local buffer errors,
Remote buffer errors, and
Selection timeouts.

Background data link counters are used to provide a statistical base for the cumulative error counters
and therefore record:
•
•
•

The number of data messages transmitted,
The number of data messages received, and
The number of selection intervals.

6-2

CALL RECEIVED
FROM CONTROL
STATION

......_ ...... YES

GATHER THE FOLLOWING
INFORMATION FROM THE
FAILING TRIBUTARY

CALL MODEM
SERVICE REP.

RELOAD TRIB SOFTWARE
AND INITIALIZE POLLING
FROM CENTRAL STATION

RX PROBLEM
IN MODEM OR
LINE. CALL SERVICE

GO TO FAILING
STATION

MK·2398

Figure 6-1

Typical Isolation Troubleshooting Flowchart (Sheet 1 of 5)

6-3

RELOAD SOFTWARE AND
RESTART AT CONTROL
AND TRIB STATIONS

CALL MODEM
SERVICE REP.

END
NO

GO TO CONTROL
STATION

CALL MODEM
SERVICE REP.

TX PROBLEM IN
MODEM OR LINE.
CALL SERVICE

MK-2399

Figure 6-1

Typical Isolation Troubleshooting Flowchart (Sheet 2 of 5)

6-4

RX PROBLEM IN
MODEM OR LINE
CALL SERVICE

YES

CHECK FOR STREAMING
TRIB. CALL SITES TO
OBSERVE RTS ON MODEM

MODEM OR LINE
PROBLEM CALL
SERVICE

GO TO SITE THAT
CAUSED CONDITION AND
RESTORE ALL OTHERS

MK·2400

Figure 6-1

Typical Isolation Troubleshooting Flowchart (Sheet 3 of 5)

6-5

GO TO FAILING
STATION

YES

TX PROBLEM IN
MODEM OR LINE
CALL SERVICE
MK·2401

Figure 6-1

Typical Isolation Troubleshooting Flowchart (Sheet 4 of 5)

6-6

RELOAD SOFTWARE AND
RESTART CONTROL

BYPASS SPLITTER OR
USE DIFFERENT PORT

CALL MODEM
SERVICE REP.

HAVE SPLITTER OR
MODEM REPLACED

INITIATE CORRECTIVE
MAINTENANCE

CALL
SUPPORT

MK·2402

Figure 6-1

Typical Isolation Troubleshooting Flowchart (Sheet 5 of 5)
6-7

STARTING
ADDRESS
(OCTAL)

OOOOr---------------------------------------~

LINKED
LISTS

215 ENTRIES
8 BYTES PER ENTRY

3270~--------------------------------------~

GLOBAL STATUS SLOT

72 BYTES

3400~--------------------------------------~

SLOT MAPPING TABLE (SMT)
256 BYTES
4000~--------------------------------------~

TRIBUTARY
STATUS
SLOTS
(TSS)

32 ENTRIES
64 BYTES PER ENTRY

7777~

______________________________________~
MK-2389

Figure 6-2

Data Memory

6-8

TRIBUTARY STATUS SLOT (TSS)
ADDRESS (OCTAL)

RESERVED
RECEIVE THRESHOLD ERRORS

TRANSMIT THRESHOLD ERRORS
SELECTION THRESHOLD ERRORS

7
I---

DATA MESSAGES TRANSMITTED

DATA MESSAGES RECEIVED

SELECTION INTERVALS

DATA ERRORS OUTBOUND
RESERVED

OREP

IIREP

SELECTION TIMEOUTS
RESERVED

OHBCC

!DBCC

H BCC

LBTS

LBTU

RBTU

NRTS

REMOTE BUFFER ERRORS
RESERVED

LOCAL BUFFER ERRORS
RESERVED

ODBCC

DATA ERRORS INBOUND
RESERVED

I I

I
I

RBTS

IRTS

LOCAL REPLY TIMEOUTS
REMOTE REPLY TIMEOUTS

MK-1960

Figure 6-3

Tributary Error Counters

6-9

GLOBAL STATUS SLOT (GSS)
ADDRESS (OCTAL)

REMOTE STATION ERRORS

I
16

RSEL

LUNDR

RMHFE

ROVRN

LOCAL STATION ERRORS

I
17

RSTR

LOVR

LMHFE

LOVRN

GLOBAL HEADER BLOCK CHECK ERRORS
MAINT. DATA BLOCK CHECK ERRORS

MK·1959

Figure 6-4

Station Error Counters

A point-to-point station maintains a single set of data link counters. Multipoint stations (control and
tributary) maintain a separate set of data link counters for each created tributary. Data link counters
are cleared by:
•
•
•

A master clear of the DMPII,
A control-in command to establish the tributary, or
A user issued control-in command to RDjCLR error counters.

Data Errors Outbound - This eight-bit group C0unter records NAKs received for data errors occurring
on the communications channel outbound from this station. There are three types of outbound errors for
which this counter records NAKs received: header blockcheck (OHBCC), data field blockcheck
(OBDCC), and reply response (OREP). Three separate flag bits indicate which type of outbound error
is being counted.

OHBCC (outbound header blockcheck) is set when a NAK with a reason code of one is received
for a header block check error for either data or control messages.
OBDCC (outbound data field blockcheck) is set when a NAK with a reason code of two is received for a data field block check error.
OREP (outbound reply response) is set when a NAK with a reason code of three is received for a
reply response.
Data Errors Inbound - This eight-bit group counter records occurrences which usually result from data
errors on the communications channel inbound to this station. Three separate bits indicate specific error
types connected to this counter.

IHBCC (inbound header blockcheck) is set when messages having header blockcheck errors are
received. When this error occurs, point-to-point stations and multipoint control stations send a
NAK with a reason code of one. A multipoint control station records this error for the selected
tributary regardless of the address field in the received message. A multipoint tributary records
this error only if the address field matches its station address.
IDBCC (inbound data field blockcheck) is set when NAKs with a reason code of two are to be
sent for data field blockcheck errors.

6-10

IREP (inbound reply response) is set when NAKs with a reason code of three are to be sent for
reply response.
Local Reply Timeouts - This eight-bit counter records occurrences which result from:
•

The loss of communication between two stations while the one recording this error has data to
transmi t, or

•

The choice of an unacceptable value for the reply timer.

Specifically, this error counter records the sending of a REP message.
Remote Reply Timeouts - This eight-bit counter records occurrences which result from:
•

The loss of communications between two stations while the remote station has data to transmit, or

•

The choice of an unacceptable value for the remote station reply timer.

Specifically, this counter records ACKs sent in response to a REP. The remote station sent a REP
because it received no acknowledgement for messages it previously sent. The local station received
those messages, but the remote station never received the acknowledgement.
Local Buffer Errors - This eight-bit counter records the fact that the user program at this station failed
to correctly allocate receive buffers at this station to data messages from the remote station. Two separate bits indicate the specific errors for this counter.
LBTU (local buffer temporarily unavailable) is set when a buffer is temporarily unavailable. This
condition indicates that a NAK with a reason code of eight is to be sent.
LBTS (local buffer too small) is set when a local buffer is too small for the incoming message.
This condition indicates that a NAK with a reason code of 16 is to be sent.
Remote Buffer Errors - This eight-bit counter records the fact that the user program at the remote
station failed to correctly allocate receive buffers to data messages from the user program at this station. Two separate bits indicate the specific errors for this counter.
RBTU (remote receive buffer temporarily unavailable) is set when a NAK with a reason code of
eight is received.
RBTS (remote receive buffer too small) is set when a NAK with a reason code of 16 is received.
Selection Timeouts - This eight bit counter records the occurrences which result from:
•

Loss of communications with a remote station,

•

Data errors on the communication channel to or from the remote station, and

•

The choice of an unacceptable value for this stations select timer.

This counter is used only by point-to-point, half-duplex, or multipoint control stations. Two separate bits
indicate the specific errors for this counter.
N RTS (no reply to select) is used to record selection intervals in which no transmission is received

6-11

from the tributary, and no transmit try is detected. Specifically, it records the expiration of the
select timer without the receipt of a valid control message or header, or the detection of a tried
transmission.
I RTS (incomplete reply to select) is used by point-to-point, half-duplex, and multipoint control
stations to record selection intervals which were not correctly terminated. Specifically, it records
the expiration of the select timer preceded by the receipt of a valid control message, receipt of a
valid header, or detection of a transmission. A tried transmission is indicated by:
1.
2.
3.

A carrier signal,
A DDCMP synchronization sequence, and
An SOH, ENQ, or DLE.

Data Messages Transmitted - This 16-bit counter records messages transmitted by this station, and
latches at a count of 65535. It can be used as a statistical base when evaluating data errors outbound,
local reply timeouts, and remote buffer errors.
Data Messages Received - This 16-bit counter records messages received by this station, and latches at
a count of 65535. It can be used as a statistical base when evaluating data errors inbound, remote reply
timeouts, and local buffer errors.
Selection Intervals - This 16-bit counter records the number of times this station selects the other station. It also latches at a count of 65535. Specifically, it records the number of messages transmitted
with the select flag on. It is only used by point-to-point, half-duplex, and multipoint control stations. It
can be used as a statistical base when evaluating the number of selection timeouts.

6.3.2.2

Station Counters - Station counters are eight-bit counters which latch at 255 and record unusual occurrences. These occurrences may be the result of :
•

A hardware or software fault at this station

•

A hardware or software fault at a remote station

•

A data error on the communication channel not seen by the header blockcheck field.

A single set of these counters is used for all tributaries on a multipoint link.
There are four types of station counters:
1.
2.
3.
4.

Remote station errors,
Local station errors,
Global header blockcheck errors, and
Maintenance data field blockcheck errors.

Station Counters are cleared by:
•
•

A master clear of the DMP11 or
A user issued control-in command to RD jCLR the counter.

Remote Station Errors - This eight-bit counter records occurrences caused by a fault in a remote station or by a data error on the channel inbound to this station which is not found. Four separate bits
which is not seen indicate the specific errors for this error counter.
ROVRN (remote receive overrun) is set when a NAK with a reason code of nine is received for a
receive overrun.

6-12

RMHFE (remote message header format errors) is set when a message header has a format error.
This condition indicates that a NAK with a reason code of 17 is to be sent.
RSEL (remote selection address error) is set when a multipoint control station receives a message
having an address field which does not match the address of the selected tributary. RSEL is a flag
used only by multipoint control stations.
RSTR (remote streaming tributaries) is set by either one of two events: 1) an implementation,
which depnds on the maximum transmission interval, is exceeded without letting go of the channel
(babbling tributary), or 2) the channel is not let go following the end of a selection interval.

Local Station Errors - This eight-bit counter records occurrences caused by a fault in this station or by
a data error on the channel outbound from this station which is not found. Four separate bits indicate
the specific errors for this error counter.
LOVRN (local receive overrun, NAK sent) is set for local station receive overruns. This condition
indicates a NAK with a reason code of nine is to be sent.
LOVR (local receive overrun, NAK not sent) is set by a receive overrun when a NAK is not sent.
For a multipoint tributary, this occurs if an overrun occurs while receiving a header. For other
stations, this occurs when the station is not in the DDCMP run state.
LUNDR (local transmit underruns) is set when a transmit underrun occurs.
LMHFE (local message header format error) is set when a NAK with a reason code of 17 is received to indicate a message header format error.

Global Header Blockcheck Errors - This eight-bit counter records the occurrences of header blockcheck errors that are not recorded on a per tributary basis. Specifically, it counts header blockcheck
errors for maintenance messages and for messages to tributaries where the address field does not match
the station address.

Maintenance Data Field Block Check Errors - This eight-bit counter records the occurrences of data
field blockcheck errors for maintenance messages.
6.3.2.3 Threshold Error Counters - Threshold error counters are used to determine if a continuous
fault is present. A continuous fault is one which occurs seven times. When a threshold counter reaches
its maximum value (7), the user program is informed by a control response.
In the DDCMP run state, threshold counters are cleared when the user is informed. In this way, the
user is always informed of a continuous fault. In the DDCMP ISTRT and ASTRT states, threshold
counters are not cleared when the user is informed. In this way, the user is not continuously informed of
a remote station which is not operating.
A point-to-point station maintains a single set of threshold counters. A multipoint control station maintains a separate set for each tributary. A multipoint tributary maintains a single set unless it supports
multiple tributary addresses, in which event it maintains a single set for each tributary address.
There are three types of threshold error counters: transmit, receive, and selection.

Transmit Threshold Errors - This three bit counter is incremented (if it is less than 7) in the following
ways.
1.

The DMP11 is in the ISTRT state when a STRT message is sent.

6-13

The DMPII is in the ASTRT state when a STACK message is sent

The DMPII is in the running state and a NAK with a reason code other than three (REP
response) is received, or when sending a REP message.

The Transmit Threshold Error counter is cleared:
•

On entering the ISTRT, ASTRT, or run states.

•

While in the run state and one of the following occurs:
I.

A transmit threshold error is reported.

A NAK, ACK, or data message is received acknowledging a new message.

A NAK, ACK, or data message is received when no messages are outstanding.

Receive Threshold Errors - This three bit counter is incremented (if less than 7) when a NAK with one
of the following reason codes is sent.
Reason Code

Description

1
2

Header blockcheck error.
Data field blockcheck error.
REP response.
Buffer temporarily not available.
Receive overrun.
Message header format error.

3
8
9
16

This counter is cleared when:
•

Entering the ISTRT, ASTRT, or run states,

•

A control message with a correct header blockcheck is received without a header format error,

•

A data message with correct header and data field blockchecks is received without a header
format error, and

•

In the run state, a receive threshold error is reported.

Selection Threshold Errors - This three-bit counter is only used by multipoint control stations and halfduplex, point-to-point stations. It is incremented (if less than 7) when a selection timeout occurs.
It is cleared when a message is received which has the select bit set, or while in the running state and a
selection threshold error is reported.

6.3.3 Error Counter Analysis
The software operating system records all the error counters but does not try to study or take any specific action as a result of any specific errors. The system manager or another operator instructs the
software to recover the counters. The counters are studied by the operator or system manager and then
the software is instructed to perform a specific function relative to the counter indications.

6-14

For information on recovering the counters, refer to the system specific DECnet system managers guide
or the network manager. Also, refer to Section 6.6.5 for a description of OCLT, which allows access to
the error counters.
The following example shows how the counters can be used in diagnosing a system failure. Refer to
Figure 6-5.
Assume that a full-duplex, multipoint network is made up of seven tributaries and a control station, as
shown in Figure 6-5. The type of electrical interface (EIA RS-121jRS-449, V.35, integral modem), is
not important in this example. The system manager at the control station sees that data transfers over
the network appear to be slow. Some standard file transfers are taking a longer time than usual to complete. (Note that in some events this can be caused by a sudden increase in traffic on the network.) The
problem appears to be intermittent because no threshold errors have occurred and data is being transferred between all tributaries. The system manager examines the error counters at the control station
for tributary 7. They are as follows:
ERROR COUNTERS FOR TRIBUTARY SEVEN (AT THE CONTROL STATION)
DATA ERRORS OUTBOUND = 0
DATA ERRORS INBOUND = 255 (IREP, IDBCC, IHBCC)
LOCAL REPLY TIMEOUTS = 40
REMOTE REPLY TIMEOUTS = 0
LOCAL BUFFER ERRORS = 0
REMOTE BUFFER ERRORS = 0
SELECTION TIMEOUTS = 50 (IRTS)
DATA MESSAGES TRANSMITTED = 420
DATA MESSAGES RECEIVED = 310
SELECTION INTERVALS = 212
The DATA ERRORS INBOUND counter (which is latched) indicates that NAKS have been sent:
•
•
•

For header BCC errors (IHBCC),
For data BCC errors (IDBCC), and
In response to receiving a REP (IREP).

Local reply timeouts show that REP messages have been sent to the tributary because acknowledgements for messages have not been received. Selection timeouts have also occurred where the tributary
received a poll and transmitted (raised carrier) but did not deselect itself (incomplete reply to select IRTS). From the information provided so far, there appears to be a problem on the link between the
tributary's transmitter and the control station's receiver. This is enforced again by the fact that there
are no data errors outbound, indicating that no NAKs have been received from the tributary. The problem can be in the control stations' receiver hardware (M8203 line unit, modem, etc.), the cable to the
tributary, or the tributary's transmitter hardware (M8203 line unit, modem, etc.). The next logical step
is to examine the error counters at the control station for the other tributaries. On doing this, the system
manager sees that tributary six has the same type and relative number of errors as tributary seven. This
indicates that the problem is in the control station's receiver or the cable from tributaries six and seven.
(It can be that both tributaries six and seven have tranmsitter problems, but this is not probable.) Examination of the error counters for the tributaries that are left shows very few errors. This rules out the
control station hardware.
NOTE
If failures occur at the control station they are indicated by the fact that in most events all stations
are affected.

6-15

CONTROL

r-----------,
I

TSS 1

TSS 4

TSS 2
TSS 5
I INTERNAL DMP11
.....- - -.....- - -....... I NETWORK DATA
I .....T_S_S_3_~.....T_S_S_6_....... I BASE
I TSS 7
I
I

RCV

LOCATION OF
CABLE FAULT
(FOR EXAMPLE
A HIGH RESISTANCE
CONNECTION)

XMT

0'\
I

.......
0'\

XMT

I RCV

TRIB
1

XMT

I RCV

TRIB
2

XMT

I RCV

TRIB

XMT

J RCV

TRIB
4

XMT

I RCV

TRIB

XMT

I RCV

TRIB
6

XMT

I RCV

TRIB
7
MK-2396

Figure 6-5

Full-Duplex Seven Tributory Multipoint Network

The problem appears to be an intermittent cable fault on the control station's receive line (tributaries
transmit line) between tributaries five and six. (Note that this can be a modem related problem if tributaries six and seven share the same modem by using a modem splitter.) More information about the
problem can be found by examining the error counters at the tributary end of the link.
ERROR COUNTERS FOR TRIBUTARY SEVEN (AT THE TRIBUTARY STATION)
DATA ERRORS OUTBOUND = 255 (IREP, IDBCC, IHBCC)
DATA ERRORS INBOUND = 0
LOCAL REPLY TIMEOUTS = 23
REMOTE REPLY TI MEOUTS = 40
LOCAL BUFFER ERRORS = 0
REMOTE BUFFER ERRORS = 0
SELECTION TIMEOUTS = NjA
DATA MESSAGES TRANSMITTED = 310
DATA MESSAGES RECEIVED = 420
SELECTION TIMEOUTS = NjA
Note that these error counters provide the same type of information as the control station error
counters. The errors all appear to be in one direction; that is, outbound. Note also that the number of
data messages transmitted and received match those at the control station. Because data transfer is
occurring, the chance of a hard error in the hardware is low.
NOTE
Hard errors at the control or tributary stations usually are discovered during the start-up sequence.
Examination of the remainder of the tributary's error counters shows that tributary six has errors similar to tributary seven and the other tributaries have very few errors. The slow running of the network is
because of the:
•

Timeouts that occur from time-to-time when polling tributaries six and seven and

•

A number of retransmissions that occur in order to get the data transferred.

These timeouts and retransmissions of data are the result of a cable fault between tributary five and six,
as indicated in Figure 6-5~
NOTES
1. If failures occur at the control station, they are
indicated by the fact that in most events all stations are affected.
2. Hard errors at the control or tributary stations
usually are discovered during the start-up sequence.
6.4 MAINTENANCE FUNCTIONS/MAINTENANCE MODES
The maintenance functions are available to the DMPII via the maintenance control and status register
(CSR) (BSEL 1).
6.4.1 Microprocessor Control Command Register
This register is in the high byte of address 76XXXO. A description of the CSR byte is provided in the
system programming chapter (Chapter 3). The byte format and bit descriptions are in the following
sections.

6-17

BSEL 1 has all maintenance functions, including master clear, and is not for typical user communications between the user program and the microprocessor. These functions override all other control
functions. All bits are read/write. Only master clear is functional if the BSEL 1 lock out switch is set
(refer to Chapter 2). Table 6-1 describes the bit functions of BSEL1. More details are available in
Section 3.3.1.1.

6.4.2 Maintenance Modes
The DMPll microprocessor can be tested by two basic modes:
•
•

Maintenance and
System test (free-running).

The OMPl1 line unit can be tested by three basic modes:
•
•
•

Single step internal maintenance,
System test internal maintenance, and
External maintenance.

6.4.2.1 Maintenance Mode - The maintenance mode can be invoked using selected bits of BSEL 1.
These can be used to halt the microprocessor (clear bit 15), step the microprocessor (set bit 8), examine
the current CRaM location (assert bit 10 and examine SEL6), and override the current CRaM instruction with a different instruction and execute that new instruction (load SEL6 with the new instruction and assert bits 8 and 9). Refer to the M8207 Maintenance Manual for more information.
NOTE
I. Be sure that BSELliock out BST switch I, E85
is ON to allow access to the maintenance bits in
BSELI.
2. With the BST switch OFF, it is possible to master clear the microprocessor by setting bit 14 of
BSELI.
6.4.2.2 System Test - System test mode tests the performance of the microprocessor and line unit
while running at full speed and using the control ROM.
NOTE
Run inhibit switch 7, E28 must be ON to use this
mode and confirm typical operating conditions.
6.4.2.3 Single Step Internal Maintenance Mode - The user program selects this mode by setting L U
loop bit 11 and clearing run bit 15 of SELO. This allows for checking most of the line unit without
disconnecting the M8203 from the modem or from the triaxial cable. Line unit signal 08 LPBKL is set
to keep the transmitter output active, looping the output back at TTL levels to become the receiver
input. Line unit request to send (RTS) and data terminal ready (OTR) signals are held cleared. The
clocking source is the 016 step LU signal from the microprocessor, which becomes Dl step LU at the
line unit. The user program generates the clock signal which toggles step line unit bit 12 of SELO.
6.4.2.4 System Test Internal Maintenance Mode - The user program selects this mode by setting line
unit loop bit 11 and run bit 15 of SELO. This mode allows the program to perform an off-line system
test by free-running the OMPll and checking the line unit without disconnecting the M8203 from the
modem or from the triaxial cable. The transmitter output is looped back at the TTL level to become the
receiver input. The clock source is the OMPll maintenance clock, which is 48K b/s.

6-18

Table 6-1

Bit

Name

Step JLP
(Step Microprocessor)

BSEL 1 Bit Descriptions

Description

When set, moves the microprocessor through one instruction cycle
which is usually made up of three, 60 ns pulses. The run bit should be
cleared before executing this control function.

ROM In

When set, routes the contents of SEL6 as the next microinstruction to
be executed by the microprocessor when step JLP or run is set.

ROM Out

When set, modifies the source paths for SEL6 to have the contents of
the addressed CROM. If ROM out and step JLP are set, then the contents of the next CROMaddress will output to SEL6. The PC points to
the address of the CROM.

LU Loop
(Line Unit Loop)

When set, connects the line unit's serial line OUT, back to its serial line
IN. This loop back is done at the TTL level before level conversion.
When the LU loop bit is set and the run bit (bit 7) is cleared, the step
LU clock is the only one available for shifting data in or out.
When the LU loop bit is set and the run bit is set, data is clocked at
24K bls by the maintenance clock.
If the LU loop bit is cleared and the run bit set, the loopback test connector is needed in order to perform loopback testing.

LV Loop

Run

Clock Source

Mode

Set

Clear

Step LU (bit
12 via program)

Single
step
internal
maintenance

Set

Maintenance
clock @
24K bls

System
test
internal
maintenance

Clear

Set

Maintenance
clock
determined
by rate
select SWs

External
maintenance

6-19

Table 6-1
Bit

BSEL 1 Bit Descriptions (Cont)

Description

Name
I

NOTE
The DMPII must be set up in full-duplex mode to
run in any loopback maintenance mode. For external
loopback mode, cable test connectors "3250 or
H3251 or module test connectors "3254 or "3255
are needed.
4

Step LU
(Step Line Receiver)

When set, this bit causes the DMPII at a remote multipoint station to
request the control station to invoke the primary MOP boot procedure
after that station has been shut down or as a result of a software failure.
In point-to-point networks, the DMPII having the boot bit set requests
the other station to start the procedure.

Boot

MCLR
(Master Clear)

With the run bit cleared and bit 4 set, the transmitter shifts; when bit 4
is cleared, the receiver shifts. This control function is used with LU
loop bit 11 to assume transmit and receive clocks for line unit maintenance in single step maintenance mode.

When bit 6 is set, MCLR initializes both the microprocessor and the
line unit. This bit is self clearing. The microprocessor clock is enabled
and the run bit is set, placing the DMPll in the initialized state. The
microprocessors program counter is also temporarily cleared by
MCLR.
Run controls the microprocessor clock; bit 7 is set by BUS initialization
or master clear, which enables the microprocessor clock. Run can be
cleared for maintenance by the run inhibit switch (E28). See D 16 of
the M8207-YA Print Set. The BST switch is provided to prevent run
from being cleared by a runaway microcode program when the microprocessor malfunctions. Detailed discussion of the run inhibit switch is
in Chapter 2.

Run

6.4.2.5 External Maintenance Mode - The user program selects the external maintenance mode placing the DMPII in the usual running mode (bit 11 clear and bit 15 set). The cable is terminated with a
test connector based on the use of DMPII option. The termination may be made at the BC55* panel or
at the modem end of the external cable (See Figure 2-6).
The DMPII options are configured as follows:

DMPII-AA (for RS-232-C; RS-423-A)
•

The modem must be disconnected and the test connector must be attached to the cable.
This may be done at the Be55C panel or at the modem end of the external cable.

RS-232-C
RS-423-A

Test Connector

Interface Cable

Modem Cable

H325

BC55C-I0

BC05D-25

H3251

BC55C-I0

BC55D-33

6-20

•

Data rate switches (switches 8, 9, and 10 of E39) select the clock rate. This clock signal
is looped back in the test connector to simulate modem transmit and receive clocks. The
data rate for this application must not exceed 56K b/s.

•

Modem control signals are tested for correct level conversion and cable paths. These
signals are looped back in the test connector as shown in the signal flow of Figure 2-5E.

DMPII-AB (CCITT V.35/DDS)
•

The modem must be disconnected and the H3250 test connector must be attached to
the BC05Z-25 cable.

•

Data rate switches (8, 9, and 10 of E39) select the clock rate. This clock signal is looped
back in H3250 to assume modem transmit and receive clocks.

•

Modem control signals are tested for correct level conversion and cable paths. These
signals are looped in the H3250 as shown in the signal flow of Figure 2-5B.

DMPII-AC (for integral modem local use)
•

The local link connections of the BC55A connector panel are disconnected at the local
panel and the FOX switch on the BC55A connector panel is switched to half-duplex to
do the external loopback.
CAUTION
If the DMPll is connected to another running
DMPll, disconnect the cable at the BC55A connector panel during diagnostic execution.

•

Oata rate switches (8, 9, and 10 of E39) select the clock rate. This data is looped back
through the BC55A connector panel to test, transmit, and receive data. The data rate
for this application must not be less than 56K b/s.

OMPII-AE (for RS-422-A interface)
•

The modem is disconnected and an H3251 test connector is attached to the BC550-33
cable or the BC55B panel.

•

Oata rate switches (8, 9, and 10 of E39) select the clock rate. This signal is looped back
through the H3251 to assume modem transmit and receive clocks.

•

Modem control signals are tested for correct level conversion and cable paths. These
signals are looped in the H3251 as shown in the signal flow of Figure 2-50.

6.4.3 Maintenance (LED) Indicators
Six light-emitting diodes (LEOs) are installed on the M8203 line unit to permit a visual check of certain conditions. Five of these apply specifically to modem conditions. The other LEO (015) relates to
functional conditions of the OMPII. Table 6-2 provides a description of each LEO while Figure 6-6
identifies the physical locations.
6.5 PREVENTIVE MAINTENANCE (PM)
There is no specific OMPII PM schedule. A general check of voltages and connections should be done
when system PM is performed. After checking OMPII modules or cables, a complete checkout of the
device, by running all diagnostics and, if possible, the interproce'ssor test, is needed. Special attention
must be given for the following reasons:
6-21

•

The DMPII can have seating problems and

•

CRaM (control ROM) chips installed in sockets are easily moved during removal and replacement of the M8207-YA module or adjacent modules. The CRaM chips may accidentally come in contact with the etch side of the adjacent module.

6.6 CORRECTIVE MAINTENANCE ON A PDP-II PROCESSOR
Because the FRU is either a module or cable, all corrective diagnosis should be directed towards isolating the failing FRU. DMPII diagnostics are designed to help in the isolation process and should be run
starting with the M8207 microprocessor test and continuing to the DEC/XII program. The correct
sequence of diagnostics is as follows:
Diagnostic

Description

CZDMP*

M8207 Microprocessor Static Diagnostic I

CZDMQ*

M8207 Microprocessor Static Diagnostic 2

CZDMR*

M8203 Microprocessor Static Diagnostic I

CZDMS*

M8203 Microprocessor Static Diagnostic 2

CZDMT*

D MP II Microprocessor Functional Diagnostic

CZCLM*

DMPll DCLT program

CXDMD*

DMPII DEC/XII Master Module

CXDME*

DMPl1 DEC/XII Slave Module
Table 6-2

Maintenance Indicators

Designation

Name

Description

DIO

Signal Quality

A signal from the modem that indicates the presence or absence
of the carrier. When this signal is ON it indicates the presence of
the carrier and OFF indicates an absence of the carrier.

Dl1

Carrier

Indicates that the carrier is present at the receiver.

DI2

Receiver Data

When ON, indicates that a steady stream of Is are being received.

D13

Transmit Data

When ON, indicates that a steady stream of Is are being transmitted.

D14

Request to Send

When ON, indicates that the USYRT is ready to start transmitting as soon as clear to send is detected.

D15

Diagnostics

When ON, indicates microdiagnostics are running. When the microdiagnostics complete successfully, this indicator goes off.

*Indicates the revision level

6-22

, - - - - - - SIGNAL QUALITY (010)
, - - - - - - MICROOIAGNOSTIC (015)

IFr§
~
r--

CARRIER (011)
REQUEST TO SEND (014)
RECEIVER DATA (012)

I . - TRANSMIT DATA (013)

OTIIO

MK-2188

Figure 6-6

M8203 Maintenance LED Locations

6-23

6.6.1 CZDMP* /CZDMQ*
These diagnostics test the M8207-Y A microprocessor in two parts. Through dialogue with the operator
and by using the diagnostic supervisor (DS), the program allows for the adjustment of device parameters, such as the UN I BUS address, vector address, and processor type.
This program is compatible with the stand-alone diagnostic supervisor and does not exceed 16K of memory. Refer to Appendix C for details on the diagnostic supervisor.
The total time needed to run M8207 static tests is approximately from 30 seconds to 2 minutes per pass,
depending on the CPU type.
CZDMP* and CZDMQ* are compatible with XXDP+ , ACT/SLIDE, and APT. XXDP+ and
ACT/SLIDE may be run in dump or chain modes. APT can be run in program or script modes.
Memory management is not used in this program and, if installed, it is disabled. If parity memory is
installed, memory parity traps are disabled.
A summary of the tests performed are listed in Tables 6-3 and 6-4. All tests support the DMCII,
KMC II-A, DMR II, and DMPII. However, some tests listed in Table 6-4 are not executed on certain
devices. Refer to Table 6-4 for these exceptions. For more detail, refer to the diagnostic listings.

6.6.2 CZDMR* /CZDMS*
These diagnostics perform static tests of all M8203 logic. These include line unit register addressing,
USYRT addressing, static bit interaction and read/write logic tests, basic transmitter/receiver sequencing and data buffering, and static operations in character and bit-stuffing modes. In addition,
data messages are sent on the line unit at TTL level, or through an external test connector with a specific modem interface selected.
Static logic tests provide troubleshooting capabilities such as tight scope loops, switch options, and the
ability to lock on intermittent errors. Added tests provide fault isolation to allow for replacement of the
smallest field replaceable unit.
These programs conform to the stand-alone version of the diagnostic supervisor and are compatible with
ACT, APT, XXDP+, and SLIDE. Through dialogue with the operator, the programs permit modifications of device parameters such as the UNIBUS address, vector addresses, and device priority. The
operator can specify specific tests to be run and a number of looping, running, and reporting modes.
Refer to Appendix C for details on the diagnostic supervisor.
Device errors are reported as they occur. The report includes the test number and error description,
good and bad test data, and allowable device register contents.
A summary of the tests performed are listed in Tables 6-5 and 6-6. For more detail, refer to the diagnostic listings.

6.6.3 CZDMT*
The CZDMT* diagnostic performs testing on the DMPII option in a functional way to verify its correct operation under microcode controlled use of the DDCMP. This includes a ROM CRC/CCITT
check, microdiagnostic, command use, and error generation.
This functional test provides troubleshooting capabilities such as tight scope loops, switch options, and
the ability to lock on intermittant errors. Also, this program conforms to the stand-alone version of the
diagnostic supervisor and is compatible with APT, ACT, XXDPT, and SLIDE. Refer to Appendix C
for details on the diagnostic supervisor.
6-24

Table 6-3

CZDMP* Diagnostic Summary

Test Number

Description

1
2
3
4-8
9
10
11
12
13
14-27
28
29
30
31
32,33
34,35
36-38
39,40
41,42
43
44
45-76

Verifies that referencing UNIBUS device registers does not cause a T /0 trap
Verifies that run can be cleared
U,N I BUS register word, dual addressing test
Control status register write/read tests
Port 4 register write/read test
Port 6 register write/read test
UN I BUS register byte, dual addressing test
Maintenance instruction register test
M icrop!"ocessor test
Microprocessor I BUS/I BUS* register write/read tests
Microprocessor IBUS/IBUS* dual address test
Microprocessor BR register test
Scratchpad test
Scratchpad dual addressing test
I nterrupt tests
Priority interrupt tests
NPR tests
Test of extended address (EA) bits 16 and 17
N PR non-existent memory test
NPR test
ALU C-bit test
ALU tests

Table 6-4

Test
Number

(M8200)
DMCl1

(M8204)
KMCI1-A

DZDMQ* Diagnostic Summary

(M8207)
DMRll
DMPll

2
3,4
5
6,7
8

S
S

9
10
11
12-24
25

S
S
S
S
S

26
27,28
29
30

S
S
S
S

S
S
S

Description
Verifies that referencing
UN I BUS device registers does
not cause a timeout.
BR right shift test
lOP CRAM write/read test
lOP CRAM dual addressing test
lOP main memory test
lOP main memory dual addressing
test
lOP MAR test (4K main memory)
lOP (CRAM) ODT bits test
CRAM tests of jump(i)
CRAM tests of jump(i)
4K Main memory page dual
address test
Jump field, page test
Jump test
Z bit test
C bit test

6-25

Table 6-4 DZDMQ* Diagnostic Summary (Cont)
Test
Number
31
32
33
34
35
36
37
38
39
40
41
42
43

(M8200)
DMCll

(M8204)
KMCII-A

(M8207)
DMRll
DMPll

S
S

Description
Program clock bit test
Force power fail test
Microprocessor noise test
NODST instruction test
Extended CRAM test (M8206
only)
Microcode test (M8206 only)
Negative address test
Byte addressing test
PC register test
Branch field H test
Scratchpad 0 (SPO) selection
test
MOY INST H signal test
Master clear test

*S = Test Skipped

Table 6-5

CZDMR * Diagnostic Summary

Test Number

Description

I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

Microprocessor CSR addressing test (SEL 0)
INBUS/OUTBUS register 14 initialization test
INBUS/OUTBUS register 14 read/write bit test
Register 14 master clear test
Register 14 UNIBUS reset (lNIT) test
Line unit false selection test
I nbus register master clear test
Register 10-17 addressing test
Register II read/write bit test
Register 12 read/write bit test
Register 13 read/write bit test
Register 17 read/write bit test
Maintenance clock bit test
Extended register master clear test
Extended register addressing test
Registers 15, 16/AX2-I5, AX2-16 read/write bit test
AXO-15, AXO-16 read/write bit test
AX 1-15, AX 1-16 read/write bit test
AX3-15, AX3-16 read/write bit test
Register 17 AX2-16 read/write, master clear test
Transmitter buffer data test
Transmitter buffer sequencing test
TX MSG timing test, character mode with CRC
TX MSG timing test, bit mode with CRC
TX MSG timing test, character mode with no CRC
TX underrun set and clear test, character mode
TX character length timing test, character mode with CRC
TX character length timing test, bit mode with CRC

6-26

Table 6-5

DZDMR* Diagnostic Summary (Cont)

Test Number

Description

29
30
31
32
33
34
35
36
37
38
39

TXDA T A bit test, character mode with no CRC
USYRT RCV MSG test, character mode with CRC
USYRT RCV MSG test, bit mode with CRC
USYRT RCV MSG test, character mode with no CRC
USYRT RCV MSG test, bit mode with no CRC
Silo-disabled transmitter load test
Silo-disabled MSG test, bit mode with no CRC
RCV buffer test, character mode with CRC
RCV character length timing test, character mode with no CRC
RCV character length timing test, bit mode with no CRC
TX underrun error, idle marking character mode with no CRC
MSG termination with go ahead (GA) characters, bit mode with no CRC
Idle sync test, character mode
Strip sync test

40
41
42

Table 6-6

CZDMS* Diagnostic Summary

Test Number

Description

Bit stuffing test
RCV overrun error, set and clear test
Abort sequence test
Abort and idle flags test
TX underrun error, idle abort characters, bit mode
RCV disable test
Assembled bit count test
Secondary station address bit test
All parties address bit (RDALL) test
Insert error bit (lERR) test, character mode with no CRC
Switch pack printout and test
Register AX3-15 printout
CRC generation test
CRC error detection test
VCR parity generation test
VCR error detection test
Integral modem interface test, character mode with CRC
V.35 modem interface test, character mode with CRC
RS-232-C and RS-423-A modem interface test, character mode with CRC
RS-422-A modem interface test, character mode with CRC
Half-duplex bit (HALF DUPX) test
Half-duplex RCV disabled test, with silos disabled
Interaction of modem control bits
Data test, bit mode with no error detection
Data test, character mode with no error detection
Data test, bit mode with CRC-CCITT-l error detection
Data test, bit mode with CRC-CCITT-O error detection
Data test, character mode with CRC-16 error detection
Data test, character mode with odd VRC error detection
Data test, character mode with even VRC errol! detection
Contiguous ones in secondary station address mode, bit mode
DDCMP MSG test, character mode

2
3
4
5
6

7
8
9
10

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

6-27

Through dialogue with the operator, the program permits modification of device parameters such as the
UNIBUS address, vector addresses, and device priority. The operator can specify certain tests to be
run and a number of looping, running, and reporting modes.
A summary of the tests performed are listed in Table 6-7. For more detail, refer to the diagnostics
listings.
6.6.4 DMPll Microdiagnostic Error Reporting
The microdiagnostics for the DMPII are made up of two microprograms: internal diagnostics and interface diagnostics.
Internal diagnostics test registers and data paths that are internal to the microprocessor. These diagnostic routines run automatically on a master clear and must complete successfully before regular interaction with the PDP-II/VAX-I 1 can take place.
The user program is informed of the results by way of the CSRs. Table 6-8 is a summary of the possible
results.
Interface diagnostics test registers and data paths used for transfer of information between the PDP11 /V AX-II and the microprocessor. These routines run on request by the user program and exercise
the datapaths in the multiport RAM from the microprocessor side.
The interface microdiagnostics are a set of diagnostics that run with external diagnostics. The interface
microdiagnostics are started by a command from the external diagnostics to the DMPII microde.
These microdiagnostics can only be run in this way and will neither detect nor report errors. It is the
responsibility of the external diagnostics to monitor and interpret test results.
The microdiagnostics perform six tests:
1.

Window register test # I - This test writes all ones to the window registers (BSELO and
BSEL2-BSEL 7). The external diagnostics must read and verify that the registers were written correctly.

Window register test #2 - This test writes all zeros to the window registers. The external
diagnostics must read and verify the registers.

Interrupt test #1 - This test causes the micro-CPU to interrupt to vector xxO.

Interrupt test #2 - This test causes the micro-CPU to interrupt vector xx4.

NPR data transfer test - This test performs one NPR in and one NPR out. The external
diagnostics provide the input and output addresses and request a number of these transfers
before the test is exited.

"Hog bit" test - This test, when requested by the external diagnostics, does 256 NPRs in
with the "not last transfer" bit set. This takes the UNIBUS for a minimum of 100 microseconds. After the NPRs are complete, an interrupt to vector xxO is made.
The external diagnostic must set up a count register and vector xxO before starting the "Hog
bit" test. When started, the test enters a count loop. If the test is successful the final count is
less than five.
NOTE
Refer to the diagnostic listings for detailed operating
procedures of CZDMP*, CZDMQ*, CZDMR *,
CZDMS*, and CZDMT*.

6-28

Table 6-7

CZDMT* Diagnostic Summary

Test Number

Description

Address Test - Verifies all CSR addresses.
ROM Verification Test - Verifies that all ROMs are in place and that the contents are correct.
Not used on DMPII.
Initialization Test - Does master clear and waits for completion of microdiagnostics.
Interface Diagnostic - Runs additional microdiagnostic code to verify interrupt
and NPR logic.
RDI Remains Set Test - Verifies that RDI sets and remains set.
ROO Sets Test - Verifies that ROO sets when a control-in command of "read
modem" is issued.
Procedure Error 100 Test - Verifies that a procedure error of 100 (non-mode
definition command after a master clear) can be generated.
Procedure Error 104 Test - Verifies that a procedure error of 104 (illegal mode
change) can be generated.
Mode Change of Duplex Portion Test - Verifies that no procedure error occurs
when changing full-duplex to half-duplex.
Maximum Tributaries Test - Verifies that maximum tributaries can be assigned and that a procedure error of 114 results when one more than maximum
tributaries is tried. Verifies that a procedure error of 116 results when a try is
made to assign an already assigned tributary.
Read/Write Tributary Status Slot Test - Verifies that each TSS can be written
and read correctly.
Proced.ure Error 132 Test - Verifies that a procedure Error of 132 (try to write
into a reserved area of the TSS) is produced.
Read/Clear Command Test - Verifies that the read/clear command operates
correctly.
Global Status Slots Test - Verifies the GSS for read and write operations.
Halt Tributary Command Tests - Verifies the halt tributary command.
Kill Tributary Command Test - Verifies proper operation of the kill tribitary
command.
Procedure Error 102 Test - Verifies that a procedure error of 102 (illegal type
code used in an input command) can be produced.
Procedure Error 110 Test - Verifies that a procedure error of 110 can be produced. (Try to perform a non-global command for tributary address of 0.)
Procedure Error 120 Test - Verifies that a procedure error of 120 (illegal
request key on control-in).
Procedure Error 134 Test - Verifies that a procedure error of 134 (try to use
reserved bit is BSEL 7 on control-in).
Latch/Unlatch Poll Check - Verifies that the latch and unlatch poll commands
operate correctly.
Short Message Sending Test - Verifies that receive buffers returned and data is
correct for short messages.
Procedure Error 122 Test - Verifies that a procedure error of 122 (try to assign
a buffer for a not assigned tributary) can be produced.
Procedure Error 124 Test - Verifies that a procedure error of 124 (try to assign
a buffer for a halted tributary) can be produced.
Procedure Error 126 Test - Verifies that a procedure error of 126 (try to assign
a buffer with a byte count of 0) can be produced.

2-7
8-9
10

11
12
13
14

15
16

19-20
21

22
23
24

25
26

28
29
30
31
32
33

6-29

Table 6-7

CZDMT* Diagnostic Summary (Cont)

Test Number

Description

Procedure Error 130 Test - Verifies that a procedure error of 130 (try to assign
or transmit buffer for tributary 0) can be produced.
Point-to-Point DDCMP Transmit/Receive Test - Verifies that 256 bytes can be
transmitted and received correctly using point-to-point full-duplex.
Multipoint DDCMP Transmit/Receive Test - Verifies that 255 bytes can be
transmitted and received correctly using multipoint full-duplex.
Modem Register Read/Write Test - Verifies that the modem registers be written and read correctly.
Memory Extension Bits Test - Verifies that transfers can be made to user memory if it exists. Transfers are made using bit 16, bit
I 7, and bits 16 and I 7.
Point-to-Point Transmit/Receive Starting at Odd Byte - Verifies that transfers
can be made starting at an odd byte boundary using DDCMP point-to-point.
One Byte Odd to Even Transmit/Receive Test - Verifies that transfers can be
made from an odd transmit buffer to an even receiver buffer using multipoint.
Polling State Tests - Verifies that polling state degrades from active to inactive,
to potentially dead to dead.

35
36
37

38-40

42
43

Table 6-8

Internal Diagnostic Results

BSEL 6

BSEL 4

Indication

143

Random

The internal register test failed. The microcode is in a loop.

135

100

The data RAM test failed. The microcode is in a loop.

125

100

The ALU test failed. The microcode is in a loop.

264

100

The line unit test failed. The microcode is in a loop.

305

100

The line unit test failed. The microcode will interact with the
PDP-II/VAX-Il. This code usually indicates that a line unit is
not connected to the microprocessor.

305

The microdiagnostics completed with no errors.

6.6.5 Data Communications Link Test CZCLM* (DCLT)
DCLT is a communications equipment maintenance tool designed to verify DMPII to DMPII communication links. The DCLT program provides the coverage necessary to isolate the following faults:
•
Communications interface program functionality,
•
Communication modem,
•
Communication cable routing and installation, and
•
Physical link/network.
DCL T programs allow testing between modes with different hardware interfaces implementing the
same or compatible protocol. The DCLT program can be exercised under normal maintenance loopback tests:
6-30

•

Internal TTL loopback,

•

Hardware loopbacks:
Module test connectors or,
Cable test connectors,

•

Manual controlled local modem analog and digital loopback functions (full-duplex mode)

•

Programmable controlled local modem analog loopback (full-duplex mode),

•

Programmable controlled remote modem digital loopback (full-duplex mode).

The main goal of DLCT is to test the communications link and therefore assumes that the CPUs,
clocks, and DMPII s at each end of the link have previously been tested and found to be in the correct
working order.
Before studying any data, the user must have a complete understanding of the protocol formats for the
system under test.
DCLT may be used to access DMPII error counters or other information by using the print command.
The print command invokes a DCLT level called report within which the following commands are
available.
Command

Description

HELP or?

Prints help information for RPT>

TSS NNN/SW

Shows tributary status slot information where NNN is the decimal tributary and ADDR and SW are one of the following switches

ERROR

Indicates only error slots are to be printed

FULL

Indicates all tributary status slots are to be printed

OFFSET = NN

Indicates the tributary status slot whose offset is NN is to be printed

GSS/SW

Prints the global status information. Switches are the same as for TSS

LOG

Dumps the event log

EXIT

Exits back to the command level that you entered from. [DCLT> or
DP>]

DCLT is XXDP+ or APT compatible and runs under control of the diagnostic supervisor (DS). It
needs 24K of memory. For more information on DCLT, refet to CZCLM* document.
6.6.6 DEC/XII DMPII Modules
There are two DEC/X 11 modules for the DMPl1: DMD* and DME*. Together these two modules can
operate:
•

Up to 16 DMPII devices in point-to-point links.
6-31

•

A single DMPII configured as a multipoint control station communicating with up to 32
tributaries.

•

Up to 16 devices configured as multipoint tributaries on the same PDP-II UNIBUS.

These modules transmit, receive, and check 32 data messages of 1024 bytes each on a given physical
link. By default, this involves a single PDP-II system with one or more devices operated in internal or
external loopback mode. By operator selection of non-default modes, however, actual point-to-point or
multipoint operation is possible.
6.6.6.1 DMD* - DMD* is the master module. It canoperate up to 16 DMPII devices in looped back,
point-to-point modes or a single device in mUltipoint control mode. DMD* can be self-sufficient or it
can communicate with slave modules on the same or another processor.
A separate DMD* module is needed for each group of looped back DMPII devices, each control station, or each group of point-to-point devices.
The actual operating mode for each DMD* module is selected by software switch registers for that
module. The DMD* module uses switch registers SRI-SR3 as follows:
SR I has three values: 0, I, or 2
When SRI = 0:
•

All selected DMPlls run in point-to-point, full-duplex mode, with internal or external loopback on all devices.

•

SR2 has the following meaning:
If SR2 = 0, internal loopback is provided by the program. This is done using TTL-level loopback on the line unit. SR2=0 is the default mode of operation.

If SR2 = I, external loopback is provided by H3254 or H3255 test connnectors on each device. In this case the modem interface must be selected for testing by using SR3.
V.35 is selected when SR3 =0 (Default)
Integral modem is selected when SR3 = 10
V.35 is selected when SR3 = 20
RS-449/RS-232-C/RS-423-A is selected when SR3 = 100
RS-422-A is selected when SR3 = 200
If SR2 = 2, cable loopback is provided by the H3250 or H3251 test connectors.

When SRI = I:
•

All selected DMPII s run in point-to-point, full- or half-duplex, without loopback.

•

The DMD* module communicates with DME* (slave) modules on the same or other PDP-II
systems.

•

Other software switch registers are not used.

When SRI =2:
•

Only one DMPII is selected

6-32

•

The selected DMPII runs in multipoint control, full- or half-duplex mode, without loopback.

•

The DMD* module communicates with DME* (slave) modules on the same or other PDP-II
systems.

•

SR2 = the total number of tributaries on this multipoint link. The range is from I to 408.

•

SR3 = the starting tributary address. The program uses this starting address to compute the
other addresses. The address range is from 1 to 3778 and they may wraparound (377 to I) if
necessary.

6.6.6.2 DME* - DME* is the slave module. It can operate up to 16 DMPII devices in point-to-point
slave or multipoint tributary modes. A separate OME* module is needed for each group of point-topoint slaves or multipoint tributaries on a system.
As with the DMD* module, the actual operating mode for each DME* module is selected by software
switch registers for that module. The DME * module uses software switch registers SR I-SR3 as follows.
SR 1 has two values:

° 1
and

When SRI = 0:
•

All selected DMPII devices run in point-to-point slave, full- or half-duplex mode, without
loopback.

•

The DME* module communicates with the DMD* (master) modules on the same or other
PDP-II systems.

•

SR2, SR3, and SR4 are not used.

When SRI = 1:
•

All selected DMPII devices run in multipoint tributary, full- or half-duplex mode, without
loopback.

•

The DME* module communicates with a DMD* (master) module on the same or another
POP-II system.

•

SR2 = the total number of tributaries on the multipoint link on this CPU. The range is from
1 to 408.

•

SR3 = the starting tributary address. The program uses this starting address to compute the
other addresses. The address range is from 1 to 3778 and they may wraparound (377 to I) if
necessary.
NOTE
If the DMPII DEC/XII modules are configured to
run in link mode, it is recommended that the exerciser be started in run lock mode. If this is not done, the
exerciser may hang.

6.6.7 Soft Error Reports Under DEC/XII
Soft errors indicate errors which occurr causing a message retransmission. The DMD* module requests
data errors inbound and outbound for each pass. If any errors are present, they are reported as soft

6-33

errors. The soft error report may be used in the isolation of certain DMPll failures from UNIBUS
loading or data late problems.
The DMP has no data late bit or capabilities for detecting the fact that it did not get bus mastership in
time to service the synchronous line. The DMPll sees such a condition as an error in the synchronous
data stream (a BCC error, transmitter underrun, or receiver overrun) and DDCMP causes the message
to be retransmitted. This causes error counters in DMPll RAM memory to increase. A process of elimination must be used to determine whether soft errors (BCC) are caused by bus latency or failing
DMPII hardware.
Usually, the DMPII should show no errors when running in a local loopback mode. This is usually a
noise-free circuit. Therefore, any soft error reports should be examined and the cause isolated.
If soft errors are reported while running a DMPll on a completely loaded system (other devices being
exercised), they may be because of bus latency. This may be verified by running only the
DMD*DEC/X II module with only one DMPII enabled. If the soft errors discontinue, a latency condition is indicated.
If soft errors continue while running only the DMD*DEC/X II module, the DMPII device diagnostics
should be run. The problem can be a faulty DMPII or cable.

SRI and SR2 (bit 0) may be used in the isolation process. If SRI =0 and SR2= I, DEC/XII does not
set line unit loopback but it uses an external turnaround. By running with SRI =0 and SR2=0, a TTL
loopback is performed. This proves that the ending cable/turnaround connector is not faulty.
TTL loopback eliminates the level converters and the integral modem. The bit rate selected is 48K b/s
using the maintenance clock.
6.7 CORRECTIVE MAINTENANCE ON A VAX-ll/780
Because the FRU is either a module or cable, all corrective diagnosis should be directed to isolating the
faulty FRU. DMPII diagnostics are designed to help in the isolation process and should be run starting
with the basic microprocessor test. The correct diagnostics sequence is as follows:
Diagnostic

Description

EVDXA
REV 3.0 and Above

COMM Microprocessor Repair Level
Diagnostics (Level 3)

EVDMA
REV *.*

M8203 Line Unit Repair Level
Diagnostics

EVDMB
REV *.*

VAX DMPII Functional Diagnostic

6.7.1 EVDXA *.* COMM Microprocessor Repair Level Diagnostics
This diagnostic performs tests on the M8207-Y A microprocessor. It includes device initialization, register and RAM addressing, read/write testing, interrupt generation and priority, NPR operation and addressing, ALU functions and the microprocessor instruction set, and INBUS/INBUS* and OUTBUS/OUTBUS* testing. This program performs many of the tests in maintenance mode by moving the
microprocessor through several instruction sequences. This program will be run at VAX Level 3, which
is a stand-alone repair level.
6.7.2 EVDMA REV *.* M8203 Line Unit Repair Level Diagnostics
This diagnostic performs register and USYRT addressing tests, static bit-interaction and read/write

6-34

logic tests, basic transmitter and receiver sequencing tests, and static operation in character and bitstuffing mode tests. This program performs many of the tests in Internal Loopback Mode using the
USYRT maintenance bit and the line unit loopback features. In externalloopback mode it uses a turnaround connector. This program is implemented as a separate VAX diagnostic, which runs at Level 3.
6.7.3 EVDMB* VAX-llj780 DMPll Functional Diagnostic (Level 2)
The DMPII functional diagnostic verifies the operation of a VAXjDMPll in both point-to-point and
multipoint modes.
The point-to-point configuration includes two DMPlls located on the same VAX host system or on two
separate VAX systems. EVDMB supports up to 32 VAXjDMPII based tributaries in multipoint mode.
There are eight sections in the program which include:
Maintenance Wrap - This section uses the maintenance loopback feature for testing full-duplex,
point-to-point configurations. Data is looped back by the DMPII. The clock is provided. This is
the default section because it needs no manual intervention.
Cable Wrap - This section effects data loopback using the H3250, H3251, H3254, H3255, or
identical turnaround connector. The clock is provided by the DMPII.
Modem Wrap - This section allows data to be looped back at the modem. Depending on the specific modem in question, it supports local and remote loopback in either manual or programmable
mode. The modem supplies the clock. This section supports the standard interfaces: RS-232-C,
RS-423-A, RS-422-A, RS-449, and V.35.
Point-To-Point Transmitter - This section is the starter of the system to system test. It supports
system to system data communications between two VAX systems. It is used with the point-topoint receiver.
Point-To-Point Receiver - This section supports system to system data communications between
two VAX systems. It is used with the point-to-point transmitter.
Multipoint - This section supports a data communications link between multiple VAXjDMPll
systems in a multipoint configuration.
Functional Test - This section exercises the basic functionality of the DMPII within the limits set
by the VMSjDMPll device driver. The tests that make up this section are a selected subset of the
CZDMT* DMPII functional tests.
Manufacturing Test - This section is made up of those tests which support the manufacturing
process.
6.7.4 EVDMD* VAX DMPll DCLT Program
DCLT is a communications equipment maintenance tool for isolating failures to either the interface,
the telephone communications line, or the modem. It exercises DMPll to DMPII links.
DCLT runs under the control of the VAX diagnostic supervisor and needs 256K bytes of memory.

6-35

APPENDIX A
DDCMP IN A NUTSHELL

A..t DDCMP
The Digital Data Communications Message Protocol (DDCMP) provides a data link control procedure
that ensures a reliable data communication path between communication devices connected by data
links. DDCMP has been designed to operate over full- and half-duplex synchronous and asynchronous
channels in both point-to-point and multipoint modes. It can be used in a variety of applications such as
distributed computer networks, host front-end processors, remote terminal concentrators, and remote
job entry-exit systems.
A.t.t Controlling Data Transfers
The DDCMP message format is shown in Figure A-I. Three control characters are provided in
DDCMP to differentiate between the three possible types of messages:
SOH - data message follows
ENQ - control message follows
DLE - maintenance message follows
Note that the use of a fixed-length header and message size declaration obviates the requirement for
extensive message and header delimiter codes.

SOH

DATA
CRC-2
COUNT FLAG RESPONSE SEQUENCE ADDRESS CRC-1
(ANY NUMBER OF 8-BIT
16 BITS
16 BITS
8 BITS
8 BITS
14 BITS 2 BITS 8 BITS
CHARACTERS UP TO 214)

MK-2248

Figure A-I

DDCMP Data Message Format

A.t.2 Error Checking and Recovery
DDCMP uses a 16-bit cycle redundency check (CRC-16) for detecting transmission errors. When an
error occurs, DDCMP sends a separate negative acknowledge (NAK) message. DDCMP does not require an acknowledgement message for all data messages. The number in the response field of a normal
header or in either the special NAK or acknowledge (ACK) control message specifies the sequence
number of the last good message received. For exanlple, if messages 4, 5, and 6 have been received
since the last time an acknowledgement was sent and message 6 is bad, the NAK message specifies
number 5 which says "messages 4 and 5 are good and 6 is bad." When DDCMP operates in full-duplex
mode, the line does not have to be turned around; the NAK is simply added to the sequence of messages
for the transmitter.

A-I

When a sequence error occurs in DDCMP, the receiving station does not respond to the message. The
transmitting station detects, from the response field of the messages it receives (or via timeout), that
the receiving station is still looking for a certain message and sends it again. For example, if the next
message the receiver expects to receive is 5, but receives 6 instead, the receiver will not change the
response field (which contains a 4) of its data messages. The receiver will say, "I accept all messages up
through message 4 and I am still looking for message 5."

A.l.3 Character Coding
OOCMP uses ASCII control characters for SYN, SOH, ENQ and OLE. The remainder of the message, including the header, is transparent.
A.l.4 Data Transparency
OOCMP defines transparency by use of a count field in the header. The header is of a fixed length. The
count in the header determines the length of the transparent information field, which can be from 1 to
16,383 bytes long. To validate the header and count field, it is followed by a CRC-16 field; all header
characters are included in the CRC calculation. Once validated, the count is used to receive the data
and to locate the second CRC-16, which is calculated on the data field. Therefore, character stuffing is
avoided.
A.1.5 Data Channel Utilization
OOCMP uses either full-duplex or half-duplex circuits at optimum efficiency. In the full-duplex mode,
ODCMP operates as two independent one-way channels, each containing its own data stream. The only
dependency is the acknowledgements, which must be sent in the data stream in the opposite direction.
Separate ACK messages are unnecessary, reducing the control overhead. Acknowledgements are simply placed in the response field of the next data message for the opposite direction. If several data
messages are received correctly before the terminal is able to send a message, all of them can be acknowledged by one response. Only when a transmission error occurs, or when traffic in the opposite
direction is light (no data message to send), is it necessary to send a special NAK or ACK message.
In summary, ODCMP data channel utilization features include:
1.

The ability to run on full-duplex or half-duplex data channel facilities,

Low control character overhead,

No character stuffing,

No separate ACKs when traffic is heavy; this saves on extra SYN characters and intermessage gaps,

Multiple acknowledgements (up to 255) with one ACK, and

The ability to support point-to-point and multipoint lines.

A.2 PROTOCOL DESCRIPTION
DDCMP is a very general protocol; it can be used on synchronous or asynchronous, half-duplex or fullduplex, serial or parallel, and point-to-point or multipoint systems. Most applications involving protocols
are half-duplex or full-duplex transmission in a serial synchronous mode; that operating environment
will therefore be emphasized in this description.
The header is the most important part of the message because it contains the message sequence numbering information and the character count, the two most important features of DDCMP. Because of
the importance of the header information, it merits its own CRC blockcheck, indicated in Figure A-2 as

A-2

CRC-I. Messages that contain data, rather than just control information, have a second section which
contains any number of 8-bit characters (up to a maximum of 16,383) and a second CRC (indicated in
Figure A-2 as CRC-2).
Before the message format is discussed in greater detail, the message sequencing system should be explained because most of the header information is directly or indirectly related to the sequencing operation.
In DDCMP, any pair of stations that exchange messages with each other number those messages sequentially starting with message number one. Each successive data message is numbered using the next
number in sequence, to modulo 256. Thus, a long sequence of messages would be numbered 1, 2,
3, ... 254, 255, 0, 1, ... The numbering applies to each direction separately. For example, station A might
be sending its messages 6, 7, and 8 to station B, while station B is sending its messages 5, 6, and 7 to
station A. Thus, in a multipoint configuration where a control station is engaged in two-way communication with 10 tributary stations, there are 20 different message number sequences involved - one
sequence for messages from each of the 10 tributaries to the control station and one sequence for messages from the control station to each of the 10 tributaries.
Whenever a station transmits a message to another station, it assigns its next sequential message number to that message and places that number in the sequence field of the message header. In addition to
maintaining a counter for the sequentially numbered messages which it sends, the station also maintains
a counter of the message numbers received from the other station. It updates that counter whenever a
message is received with a message number exactly one higher than the previously received message
number. The contents of the received message counter are included in the response field of the message
being sent, to indicate to the other station the highest sequenced message that has been received.
When a station receives a message containing an error, that station sends a negative acknowledge
(NAK) message back to the transmitting station. DDCMP does not require an acknowledgement for
each message, as the number in the response field of a normal header (or in either the special NAK or
positive acknowledgement message ACK), specifies the sequence number of the last good message received.
When a station receives a message that is out of sequence, it does not respond to that message. The
transmitting station detects this from the response field of the messages which it receives; if the reply
interval expires before the transmitting station receives an acknowledgement, the transmitting station
sends a reply (REP) message. The REP message contains the sequence number of the most recent
unacknowledged message sent to the remote station. If the receiving station has correctly received the
message referred to in the REP message (as well as the messages preceding it), it replies to the REP by
sending an ACK. If it has not received the message referred to in sequence, it sends a NAK containing
the number of the last message that it did receive correctly. The transmitting station then retransmits
all data messages after the message specified in the NAK.
The numbering system for DDCMP messages permits up to 255 unacknowledged messages outstanding, a useful feature when working on high-delay circuits, such as those using satellites.
However, the DMPII limits the maximum number of unacknowledged messages outstanding to be 127.

A.3 MESSAGE FORMAT
With the above background, it is now time to explore the various DDCMP message formats in full
detail, as shown in Figure A-2. The first character of the message is the class of message indicator,
represented in ASCII with even parity. There are three classes of messages; data, control, and maintenance. These are indicated by class of message indicators SOH, ENQ, and DLE. The next two characters of the message are broken into a 14-bit field and a 2-bit field. The 14-bit field is used in data and
maintenance messages to indicate the number of characters that will follow the header CRC field and

A-3

NOTE 1

COUNT FLAG
RESPONSE SEQUENCE ADDRESS
CLASS
14 BITS 2 BITS
8 BITS
8 BITS
8 BITS

xxxxxxxx
SOH - DATA MESSAGES

10000001
00000101
ENQ {ACKNOWLEDGEMENT
NEGATIVE ACKNOWLEDGE 00000101

REASONS:

ENQ
DLE

XXXXXXXXXXXXXX
CHARACTER COUNT
00000001000000
00000010---------

BCC HEADER ERROR
BCC DATA ERROR
REP RESPONSE
BUFFER UNAVAILABLE
RECEIVER OVERRUN
MESSAGE TOO LONG
HEADER FORMAT ERROR

00000101
00000101
START ACKNOWLEDGEMENT 00000101
10010000
MAINTENANCE MESSAGE

REPLY MESSAGE
START MESSAGE

XX
QS
QS
QS

CRC 1
16 BITS

INFORMATION
ANY NUMBER CRC 2
OF 8-BIT
16 BITS
CHARACTERS

XXXXXXXX

XXXXXXXX
RESP #
RESP #
RESP #

MESSAGE#
00000000
00000000

XXXXXXXX
ADDRESS
ADDRESS
ADDRESS

00000000
00000000
00000000
00000000

LSTMESS#
00000000
00000000
00000000

ADDRESS
ADDRESS
ADDRESS
ADDRESS

000001
000010
000011
001000
001001
010000
010001

00000011000000
00000110000000
00000111000000
CHARACTER COUNT

QS
11
11
'1

NOTES:
1. ONLY THE DATA MESSAGE AND THE MAINTENANCE MESSAGE HAVE CHARACTER COUNTS, SO
ONLY THESE MESSAGES HAVE THE INFORMATION AND CRC2 FIELDS SHOWN IN THE MESSAGE
FORMAT DIAGRAM ABOVE.
2. "RESP #" REFERS TO RESPONSE NUMBER. THIS IS THE NUMBER OF THE LAST MESSAGE
RECEIVED CORRECTLY. WHEN USED IN A NEGATIVE ACKNOWLEDGE MESSAGE, IT IS ASSUMED
THAT THE NEXT HIGHER NUMBERED MESSAGE WAS NOT RECEIVED, WAS RECEIVED WITH
ERRORS, OR WAS UNACCEPTED FOR SOME OTHER REASON. SEE "REASONS."
3. "MESSAGE#" IS THE SEQUENTIALLY ASSIGNED NUMBER OF THIS MESSAGE. NUMBERS ARE
ASSIGNED BY THE TRANSMITTING STATION MODULO 256; I.E., MESSAGE 000 FOLLOWS 255.
4. "LSTMESS#" IS THE NUMBER OF THE LAST MESSAGE TRANSMITTED BY THE STATION. SEE THE
TEXT DISCUSSION OF REP MESSAGES.
5. " ADDRESS" IS THE ADDRESS OF THE TRIBUTARY STATION IN MULTIPOINT SYSTEMS AND IS
USED IN MESSAGES BOTH TO AND FROM THE TRIBUTARY. IN POINT TO POINT OPERATION, A
STATION SENDS THE ADDRESS "'" BUT IGNORES THE ADDRESS FIELD ON RECEPTION.
6. "Q" AND "S" REFER TO THE QUICK SYNC FLAG BIT AND THE SELECT BIT. SEe TEXT.
MK-2249

Figure A-2

DDCMP Message Format in Detail

A-4

form the information part of the message. In control messages, the first 8 bits of the 14-bit field are
used to designate what type of control message it is; the last 6 bits are generally filled with zeros. The
exception is in NAK messages where the last six bits are used to specify the reason for the NAK. The 2bit field contains the quick sync and select flags.
The quick sync flag is used to inform the receiving station that the message will be followed by sync
characters; the receiver may wish to set its associated synchronous receiver hardware into sync search
mode and sync strip mode. This will re-establish synchronization and syncs will be discarded until the
first character of the next message arrives. The purpose of this is to permit the receiving station to
engage any hardware sync-stripping logic it might have and prevent it from filling its buffers with sync
characters. The select flag is used to control link management in half-duplex or multipoint configurations, where transmitters need to get turned on and off.
Link management is the process of controlling the transmission and reception of data on links where
there may be two or more transmitters and/or receivers actively connected to the same signal channels.
This will be true of half-duplex, point-to-point links, as well as full- and half-duplex multipoint links. On
half-duplex links, only one transmitter may be active at a time; on full-duplex links, only one slave
transmitter may be active on the link at a time.
A station on such a link may transmit when it has been selected or granted ownership of the link. This
ownership is passed by use of the select flag existing in all messages. A select flag set in a received
message allows the addressed station to transmit after completing reception of the message. The select
flag also means that the transmitter will cease transmitting after the message is sent.
The response field contains the number of the last message correctly received. This field is used in data
messages and in the positive and negative acknowledge types of control message. Its function should be
evident from the preceding discussion of sequence control.
The sequence field is used in data messages and in the REP type of control message. In a data message,
it contains the sequence number of the message as assigned by the transmitting station. In a REP message, it is used as part of the question, "Have you received all messages up through message number
(specify) correctly?"
The address field is used to identify the tributary station in multipoint networks and is used in messages
both to and from the tributary. In point-to-point operation, each station uses an address of 1.
In addition to the positive and negative acknowledgement and REP types of control message, there are
also start and start acknowledge control messages. These are used to place the station which receives
them in a known state. In particular, they initialize the message counters, timers, and other counters.
The start acknowledge message indicates that this has been accomplished.
Figure A-2 also shows the maintenance message. This is typically a bootstrap message containing load
programs in the information field. A complete treatment of maintenance messages and start up procedures is beyond the scope of this book.

NOTE
Refer to the DDCMP Specification Order AAD599A-TC for a complete detailed description of
DDCMP.

A-5

APPENDIX B
FLOATING DEVICE
ADDRESSES AND VECTORS

B.I FLOATING DEVICE ADDRESSES
UNIBUS addresses starting at 760010 and continuing through 763776 are designated as floating device addresses (see Figure B-l). These are used as register addresses for communications (and other)
devices interfacing with the PDP-II and VAX-II Systems.
NOTE
Some devices are not supported by VAX-II, however, the same scheme applies; that is, gaps are provided as appropriate. The convention for assigning
these addresses is as follows:
A gap of 108 must be left between the last address of one device type and the first address of the next
device type. The first address of the next device type must start on a modulo 108 boundary. The gap of
108 must also be left for devices that are not installed but are skipped over in the priority ranking list.
Multiple devices of the same type must be assigned contiguous addresses. Reassignment of device types
already in the system may be required to make room foradditional ones.

B.2 FLOATING VECTOR ADDRESSES
Vector addresses, starting at 300 and proceeding upward to 777, are designated as floating vectors.
These are used for communications (and other) devices that interface with the PDP-II and VAX-II.
NOTE
Some devices are not supported by VAX-II, however, the same scheme applies. Vector size is determined by the device type.
There are no gaps in floating vectors unless required by physical hardware restrictions (in data communications devices, the receive vector must be on a zero boundary and the transmit vector must be on a
48 boundary).
Multiple devices of the same type are assigned vectors sequentially.

B-1

777 777
DIGITAL EQUIPMENT
CORPORATION

2K
WORDS

(FIXED ADDRESSES)

DR11-C
~

1K
WORDS

770 000
767 777

t
USER ADDRESSES

764 000
763 777

t
1K
WORDS

·FLOATING ADDRESSES
DIGITAL EQUIP CORP (DIAGNOSTICS)

760 010
760 006
760 000
757 777

001 000
000 777
80

VECTORS

FLOATING VECTORS

000 300
000 277
TRAP & INTERRUPT
VECTORS

48
VECTORS

000 000
MK-2190

Figure B-1

UNIBUS Address Map

B-2

Floating CSR Address Devices

Rank

Option

Decimal
Size

Octal
Modulus

I
2
3
4
5
6
7
8
9
10
II
12
13
14
15
16
17
18
19
20
21
22
23
24

DJII
DHII
DQII
DUll
DUPII
LKIIA
DMCII/DMRII
DZII * and DZVII/DZ32
KMCII
LPPII
VMV21
VMV31
DWR70
RLllandRLVll
LPAII-K
KWII-C
Reserved
RXll
DRII-W
DRII-B
DMPII-AD
DPVII
ISB II
DMVII-AD

4
8
4
4
4
4
4
4
4
4
4
8
4
4
8
4
4
4
4
4
4
4
4
8

10
20t
10
10
10
10
10
10
10
10
10
20t
10
10 (extra only)
20 (extra only)
10
10
10 (extra only)
10
10 (After second)
10
10
10
20

* DZ I I-E and DZ II-F are dual DZ II s and are treated by the algorithm as two DZ II s.
tStarting CSR ADDRESS must be an even multiple of 20 (octal).

B-3

Floating Interrupt Vector Devices
Decimal
Octal
Size
Modulus

Rank

Option

I
2
2
2
3
4
5
6
7
8
9
10
II
12
13
14
14
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47

DCll
KLl1 (extra)
DLII-A(extra)
DLII-B (extra)
DPII
DMII-A
DNII
DMII-BB
DH II modem control
DRII-A
DRII-C
PA611 (reader & Punch)
LPDll
DTll
DXII
D L II-C
DLI1-D
DLII-E
DJ II
DH II
GT40
L PS I I
DQII
KWII-W
DUll
DUPII
DV and modem control
LKII-A
DWUN
DMCII/DMRII
DZII/DZ32
KMCII
LPPII
VMV21
VMV31
VTVOI
DWR70
RLII/RLVII
RX02
TS II
LPAII-K
IPII/IP300
KWII-C
RX II
DRII-W
DRII-B
DMDPII-AD
DPVII
ML II
ISBII
DMVII-AD

4
10
4
10*
4
10*
4
10
410
4
10
2
4
2
4
2
4
4
10*
4
10*
8
10*
4
10
4
10*
4
10*
4
10*
4
10*
4
10*
4
10*
4
lOt
8
10
12
30*
4
lOt
4
10
4
10*
4
10*
6
10
4
10
4
10
4
10*
4
10*
4
10
4
10
4
10
4
10
4
10
4
10*
2
4
2
4
2
4 (after the first)
4
10
2
4
4
10
2
4 (after the first)
.2
4
2
4 (after the first)
4
10
4
10
2
4 (MASSBUS device)
4
10
4
10

*The vector for the device of this type must always be on a 108 boundary.

t These devices can have either a M7820 or M7821 interrupt control module. However,
it should always be on a 10 8 boundary.

B-4

B.3

EXAMPLES OF DEVICE AND VECTOR ADDRESS ASSIGNMENT

Example I

VAX-II

The first device requiring address assignment in this example is a DMCII (number 7 in the device
address assignment sequence; number 26 in the vector address assignment sequence).
The only devices used are:
I DMCII
I DZ32
I DMPll
Device
(Option)

Device
Address

Vector
Address

Comment
Gap left for DJ II
Gap left for DH II
Gap left for DQ II
Gap left for DUll
Gap left for DUPII
Gap left for LKIIA

760010
760020
760030
760040
760050
760060
DMCII

760070
760100

300

First DMCII
Gap left between DMCII and next
device

DX32

760110
760120

310

First DZ32
Gap left between DZ32 and next
device

DMPII

760130
760140
760150
760160
760170
760200
760210
760220
760230
760240
760250
760260

Gap left for KMCII
Gap left for LPPII
Gap left for VMV21
Gap left for VMV31
Gap left for DWR 70
Gap left for RL II
Gap left for LPAII-K
Gap left for KWII-C
Gap left for reserved device
Gap left for RX II
Gap left for DRII-W
Gap left for DRII-B

760270
760300

FIRST DMPII
Gap left after last device, in
this case the DMPII, to indicate
that none follow.

B-5

Example 2

PDP-II

The first device requiring address assignment in this example is a DMRII (Number 7 in the device
address assignment sequence; Number 26 in the vector address assignment sequence.)
The only devices used are:
1 DMRII
2 DZlls
2 DMPlls

Device
(Option)

Device
Address

Vector
Address

Comment

Gap left for DJ 11
Gap left for DHII
Gap left for DQ 11
Gap left for DUll
Gap left for DUPII
Gap left for LK 11 A

760010
760020
760030
760040
760050
760060
DMRll

760070
760100

300

First DMRII
Gap left between DMRII and next
device

DZII
DZII

760110
760120
760130

310
320

First DZII
Second DZII
Gap left between DZ 11 and next
device
Gap left for KMCII
Gap left for LPPll
Gap left for VMV21
Gap left for VMV31
Gap left for DWR 70
Gap left for RL 11
Gap left for LPAII-K
Gap left for KWII-C
Gap left for reserved device
Gap left for RX 11
Gap left for DRII-W
Gap left for DRII-B

760140
760150
760160
760170
760200
760210
760220
760230
760240
760250
760260
760270
DMPII

760300
760310
760320

330
340

First DMPII
Gap left after last device, in this
case the DMPll, to indicate that
none follow.

B-6

APPENDIX C
LINE UNIT AND MODEM CONTROL
REGISTER FORMATS

C.I LINE UNIT AND MODEM CONTROL REGISTER FORMATS
The modem signals made available by the M8203 line unit can be examined or modified by the user
program if needed. This supplies the flexibility needed to meet the different modem interface requirements of different countries. The bits for these signals are in registers 13 and 17. The user program can
read registers 13 and 17 and write into register 13 through control commands.
To read registers 13 and 17, the user program sets the code read modem control in the request key and
the contents of registers 13 and 17 are returned to the user by an information response. The contents of
register 13 appear in BSEL4 and the contents of register 17 appear in BSEL5. To write into register 13,
the user program issues a control command having the request key write modem control and the new
value of register 13 in BSEL4.
Another line unit register that can be written into by the user is the extended register 3-15. The main
purpose of writing into this register is to select the correct interface when test connectors are used. The
user writes into this register with the write modem test register request key.
NOTE
Extended register 3-145 is used for diagnostic purposes only
Read Register 13

Bit Name

Description

o CARRIER

Received line signal detector, usually referred to as carrier detect,
indicates that there is an audio tone being received from the remote modem. Usually, in full-duplex applications, it is on when the
communications line is intact and the remote modem has the signal
request to send asserted (the modem is transmitting). In half-duplex applications, carrier detect is on whenever the carrier (audio
tone) is on the line and request to send at the local modem is off.
This signal is also acceptable to the DMPll integral modem.

STANDBY

This signal, when asserted, shows that the indicator local modem is
in condition to operate with the typical or standby communications
facilities. An on condition indicates that the local modem is conditioned to operate in its standby mode with the usual facilities replaced by the predetermined standby facilities. The off condition
indicates that the local modem is conditioned to operate in its usual
mode. This signal applies only to modems that support this feature.

C-l

CLEAR TO SEND

This signal is generated by the local modem to indicate whether or
not it is ready to transmit data. This signal has a slightly different
meaning with differnt modems. With some modems it indicates
that the carrier is being received from the remote modem and,
therefore, is an indication that an acceptable communication channel exists. In others, it is only a delayed version of request to send.
In this case clear to send is "probably" clear to send.

MODEM READY

This signal shows that the modem is ready to operate. The on condition indicates that the local modem is connected to the communications line and is ready to swap control signals with the DMPII.
The off condition indicates that the local modem is not ready to
operate. This signal, where implemented, is used by the DMPII to
detect either a power off condition or a cable-related modem malfunction.

HALF-DUPLEX

This signal, when asserted, shows that the M8203 line unit is in the
half-duplex mode. This means that the DMPII is connected to a
communications line designed for transmission in either direction,
but not in both directions at the same time. When cleared, it implies full-duplex operation, which is two-way independent transmission in both directions.

REQUEST TO SEND

This signal controls the data channel transmit function of the local
modem, and on a half-duplex channel, to control the direction of
data transmission of the local modem. On a full-duplex channel,
the on condition maintains the modem in the transmit mode and
the off condition maintains the modem in the nontransmit mode.
On a half-duplex channel, the on condition maintains the modem in
the transmit mode and prevents the receive mode. The off condition maintains the modem in the receive mode. A transition from
off to on instructs the modem to enter the transmit mode. The
modem responds by taking such action as may be necessary and
shows completion of such actions by asserting clear to send, thereby indicating to the DMPII that data may be transferred across
the communications channel. A transition from on to off instructs
the modem to complete the transmission of all data that was previously transferred to the modem and then assume a non-transmit
or receive mode, whichever is correct. The modem responds to this
instruction by turning off the clear to send signal when it is again
prepared to respond to an on condition of request to send.

DATA TERMINAL READY This signal controls the switching of the local modem to and from
the communications line. When asserted, this signal serves to inform the local modem that the DMPII is ready to operate. This
signal also prepares the modem for connection to the communication line and maintains this connection as long as it is on. When
turned off, this signal causes the local modem to disconnect after
all data previously transferred to the modem has been transmitted.
This signal can be used by the local modem to detect a power off
condition in the DMPII or a cable-related modem malfunction.

RING

This signal indicates whether an incoming call signal is being received by the local modem. When on, this signal indicates that an

C-2

incoming call (ringing) signal is being received by the local
modem. The on state of ring must appear approximately at the
same time as the on segment of the ringing cycle (during rings) on
the communications line. The off condition must be maintained
during the off segment of the ringing cycle (between rings) and at
all other times that ringing is not being received. This signal is not
affected by the state of data terminal ready.
Read Register 17
Bit Name

Description

MODE

This bit indicates the functional mode of the line unit. A one indicates character oriented protocol operation and a zero indicates
bit oriented protocol operation. The M8203 line unit initializes this
bit to one.

ECS

This is the maintenance clock (24K Hz).

TEST MODE

This signal indicates whether or not the local modem is in a test
condition. (This signal applies only to modems that support this
feature.) When in the on condition, this signal indicates to the
DMPII that the local modem has been placed in a test condition.
The on state of this signal is in response to the on condition of the
signal maintenance mode 1 or maintenance mode 2 and indicates
that the test condition is assigned. The on condition can also be in
response to either local or remote execution by means of any
modem test condition. Activation of a telecommunications network
test condition (for example, facility loopback) that is known to the
modem can also cause this signal to be on. In the off condition, this
signal indicates that the modem is not in the test mode and is available for typical operation.

IN COMPOSITE READY

This bit indicates that the receive silos are ready to receive another
character (M8203 maintenance function only).

OUT COMPOSITE READY This bit indicates that data is present at the bottom of the transmit
silo ready for transfer to the USYRT (M8203 maintenance function only).

TX DATA

This bit indicates that a data bit is present on the output of the USYRT serial data stream (M8203 maintenance function only).

SIGNAL QUALITY

This signal indicates whether or not there is a chance of an error in
the receive data. When on, this signal indicates that there is a
chance that no error has occured. An off condition indicates a high
chance of an error. The criterion for this is created by the modem
manufacturer. For example, this signal may indicate the quality of
the receive data based on the recovery of the carrier, recovery of
bit timing, or distortion of the eye pattern in the demodulator. Signal quality may also be based on the state of the equalizer for those
modems equipped with equalizers.

C-3

SIGNAL RATE

This signal is used to indicate one of the two data signaling rates of
a dual-rate synchronous modem or one of the two signal rate ranges
of an asynchronous modem. The on condition indicates the higher
data signaling rate or range of rates and the off condition indicates
the lower data signaling rate or range of rates. The rate of timing
signals, if included in the interface, can be controlled by the signal,
as correct.

Write Register 13
Bit Name

Description

o NOT USED
SELECT STANDBY

This signal is used to select normal or standby communications facilities. An off to on transition instructs the local modem to replace
the usual facilities with predetermined standby facilities. This signal is maintained in the on state when the standby facilities are
needed. An on-to-off transition of this signal instructs the local
modem to replace the standby facilities with the normal facilities.
This signal is maintained in the off state when the usual facilities
are needed for use.

MAINTENANCE MODE 2

This signal is used to control the remote loopback (remote loopback) test condition at the remote modem. When asserted, this signal causes the local modem to signal the creation of the remote
loopback test condition at the remote modem. After asserting this
signal and detecting a test mode on condition, the local modem can
operate in full-duplex mode to exercise both the local and remote
modem. In the off condition, this causes the local modem to signal
the release of the remote loopback test condition. The remote loopback test condition places the communication system out of service
to the D MP 11 connected to the modem. When the remote loopback test condition is started, the remote modem indicates an off
condition to data mode and an on condition to test mode. If the remote loopback test condition at the remote modem is started from
the local modem through manual means or by means of the remote
loop back test condition, the local modem allows data mode to respond as usual and maintain test mode in the on state.

MAINTENANCE MODE 1

This signal is used to control the local loopback (local loopback)
test condition at the local modem. In the on condition, the output of
the local modem transmitter is transferred to the input of its receiver. After creating the local loopback test condition, the local
modem asserts the test mode center. With test mode in the on
state, the local modem can operate in full-duplex to exercise the
interface. When this signal is in the off state, the modem is released from the local loopback test condition. The local loop back
test condition does not prevent an incoming call through the signal
ring.

HALF-DUPLEX

The DMPII uses this bit to place the line unit into the half-duplex
mode. The user program cannot set or clear this bit. The DMPII

C-4

can change line characteristics only through the the mode command. The M8203 line unit has a hardware interlock that prevents
simultaneous transmission and reception when in the half-duplex
mode. While the receiver is actively receiving, or carrier detect is
asserted, the data loaded into the transmit silo will not be loaded
into the USYRT for transmission and, therefore, request to send
will not be shown to the modem. Similarly, while the transmitter is
transmitting, the receiver is disabled from receiving data.
5

SELECT FREQUENCY

DATA TERMINAL READY This signal controls switching of the local modem to and from the
communications line. When asserted, this signal informs the local
modem that the DMPll is ready to operate. This signal also prepares the modem for connection to the communication line and
maintains this connection as long as it is on. When turned off, this
signal causes the local modem to disconnect after all data previously transferred to the modem has been transmitted. This signal
can be used by the local modem to detect a power off condition at
the DMPll or a cable-related modem malfunction.

POLLING

This signal is used to select the transmit and receive frequency
bands of a modem. In the on condition, the higher frequency band
is selected for transmission to the communications channel and the
lower frequency band is selected for reception from the communications channel. When off, the lower frequency band is selected
for transmission to the communications channel and the higher frequency band is selected for reception from the communications
channel.

This signal determines whether or not the local modem will quickly
respond to new data on the communications line. This signal is used
at control stations in multipoint networks where the remote
modems operate in switched carrier mode. This incoming signal to
the control station appears as a series of short message bursts transmitted by each tributary, as it responds to the poll from the control
station. In order to allow rapid accommodation to signals from
many tributaries appearing in fast order, the control station informs the local modem when a new signal is about to start by asserting polling for a short interval. For synchronous systems, clock
timing on the incoming message varies from message to message
because the remote modems are in no way synchronized to each
other. If the time interval between messages is too short, the clock
holdover after the end of one message may preclude rapid synchronization on the following message. The use of this signal allows the
control station to reset the modem receiver timing recovery circuit
enabling it to respond much faster to the line signal present after
polling has been turned off. This signal applies only to modems that
support polling.

Write Extended Register 3~ 15
Bit Name

Description

o SELM

This bit is a select interface bit. When set, the modem interface is
selected by the other bits in the register.

C-5

RESERVED

EN C32

This is reserved for CRC-32 enable.

INT

Setting this bit (when bit 0 is set) selects the integral modem, but
does not select the speed or the filter for the integral modem. The
correct cable or test connector must be use,d with the corresponding
data rate that is selected.

V.35

This bit indicates that the V.35 interface is selected.

BCC 32

This bit is reserved for the BCC match of CRC-32.

XYZ

This bit indicates that the EIA single ended interface. This includes RS-232-C and RS-423. This is the default when Bit 0 is not set.

422

This bit indicates that RS-422 is to be selected. This is a differential interface and must be switch selected in normal mode.

C.2 RS-449 VERSUS RS-232-C
The most common interface standard in use during the last few years is RS-232-C. However, when used
in modern communications systems, it has critical limitations. The most serious of these is speed and
distance.
For this reason, interface standard RS-449 was developed to replace RS-232-C. This standard maintains a degree of compatibility with RS-232-C, to make an upward transition to RS-449 easier.
The most significant difference between RS-449 and RS-232-C is the electrical characteristics of signals used between the data communications equipment (DCE) and the data terminal equipment
(DTE). The RS-232-C standard specifies only unbalanced circuits, whereas RS-449 specifies both balanced and unbalanced circuits. The specifications for these two circuit types supported by RS-449 are
in EIA Standards RS-422-A for balanced circuits and RS-423-A for unbalanced circuits. These new
standards permit increased transmission speeds and allow an increased distance between the DTE and
DCE.
The maximum transmission speeds supported by RS-422-A and RS-423-A specified circuits vary with
circuit length. The usual transmission speed limits are 20K bls for RS-423-A at 200 feet and 2M bls
for RS-422-A also at 200 feet. These usual transmission speeds can be changed by tradeoffs between
speed and distance.
Another major difference between RS-449 and RS-232-C is the specification of two new connectors to
make the leads needed to support added circuit functions and the balanced interface circuits easier.
One connector is a 37-pin cinch used to make the majority of data communications applications. The
other is a nine-pin cinch for applications needing secondary channel functions. Some of the new circuits
implemented by RS-449 support local and remote loopback testing and standby channel selection.
The transition from RS-232-C to RS-449 will not happen immediately. Therefore, applications that
need connection between RS-232-C and RS-449 interfaces must follow the limitations of RS-232-C,
which specifies a transmission speed of 20K bls at a maximum distance of 50 feet.

C-6

APPENDIX D
MODEM CONTROL

0.1 MODEM CONTROL
There are two levels of modem control available to the DMPII. The first level is provided by the
M8203 line unit, and the second by the OMPII microcode.
0.1.1 Line Unit Modem Control
The M8203 line unit provides the following modem control functions:
1.

Modem ready lockout of request to send, and

Prevention of simultaneous transmission and reception in half-duplex mode.

0.1.1.1 Modem Ready Lockout of RTS - The lockout function automatically inhibits the setting of
request to send until modem ready is asserted. This feature relieves the DMPll of having to check the
functional state of the modem each time the communications line is connected.
0.1.1.2 Half-Duplex Mode - When set, half-duplex (bit 4 of line unit register 13) specifies that the
line unit is in the half-duplex mode. In half-duplex mode, a hardware interlock prevents the line unit
from transmitting and receiving at the same time. During reception (carrier detect asserted), data
loaded into the line unit transmit silo does not move to the USYRT for transmission. Therefore, request
to send does not display to the modem. Similarly, during transmission, reception is disabled.
NOTE
This hardware lockout prevents the M8203 line unit
from being used in the half-duplex mode on a fullduplex modem with the continuous carrier option installed.
0.1.2 Modem Control Implemented by the DMPII Microcode
The main modem signal monitored by the DMPII is modem ready. Once this signal becomes true, the
OMPII reports any transition from on-to-off to the user program by issuing a control response with the
code for the system event modem disconnect. Transmission is inhibited by the line unit by interlocking
the signals modem ready and request to send.
NOTE
Modem ready is examined at one millisecond intervals, therefore, transitions smaller than one millisecond, depending on when they occur, mayor may
not be detected by the DMPl1.
Another modem signal monitored by the OMPII is ring detect. In this event, a transition from off-to-on
is reported to the user program as long as this signal stays high for at least one millisecond. The OMPll
reports this through a control response with the code for the system event modem ring detect.

D-l

When the signal carrier detect is dropped by the modem for longer than 128 milliseconds, the user
program is informed by a control response with the code for the system event modem carrier loss.
When modem ready is asserted at a point-to-point station, a special timer, referred to as the call timer,
is started. The call timer runs until a valid DDCMP header is receive\:! or the timer expires (65 seconds). When the counter expires, the DMPII clears data terminal ready and posts a control response to
the user program for the system event modem disconnect. The system then waits three seconds and reinitializes. The purpose of the call timer is to disconnect the line when an incoming call cannot communicate with the DMPII. This condition can be caused by a noisy telephone connection or a wrong number.
The flow shown by the diagrams that follow desribe the processing of EIA modem control signals by
the DMPII. Each diagram represents a serial flow for a specific modem control function. However, the
functions performed, as represented by each diagram, are performed in parallel. The readable and
writeable modem signals listed on the diagram for modem status can be read and written through the
control command using the request keys read modem test register and write mode control. The symbols
used in these diagrams are defined below.

CONDITIONS TO BE SATISFIED. IF CONDITIONS ARE NOT
MET SERIAL FLOW DOES NOT CONTINUE.

ACTIONS PERFORMED BY DMP11.

C__)

ENTRY OR EXIT SYMBOLS

MK·1973

D-2

--......,r---POWER ON

SYSTEM
INITIALIZATION

[2.4]

DEVICE
MASTER CLEAR

[2.4]

[2.5]

LEAVE DTR IN
ITS PRIOR STATE
POWER ON
BOOT ENABLED
(SWITCH)
USER ISSUES
MODE DEFN
COMMAND

MODE DEFINED
IN SWITCHES

[2.2]

BOOT
BIT SET

[2.6]

[2.2]
SET DTR
SET CALTMR
=ON

SET DTR
SET CALTMR
= ON

[2.3]
I.....---~----'

.-------;-----------------r-----------~ A r-----------~----------------~
USER REQUEST
MODEM STATUS
READ/WRITE

CALTMR

= ON

DSR = ON

[2.2]

DSR = ON
CD= ON

[2.1]
MODSTAT

SCAN FOR
REMOTE
LOAD
MESSAGE

VALID
MESSAGE
[2.3] r - - - - " " - - " - - - - o
B ...- - - - - - - - 1

RING = OFF
TO ON FOR
1 mSEC
NOTIFY USER OF
RING DETECT
(CONTROL
RESPONSE)

MESSAGE TO
BE SENT

DSR = ON TO
OFF FOR
1 mSEC

P/MOP = ON

NOTIFY USER OF
DSR DROP
DISCONNECT
(CONTROL RESPONSE)

P/MOP = OFF

TRANSMIT

NOTES:
[2.1] REMOTE LOAD DETECT IS A MAINT. DDCMP MESSAGE INITIATING A DOWN LINE LOAD.
[2.2] CALTMR - (CALL TIMER): USED TO DETERMINE IF VALID MESSAGE IS RECEIVED. "ON" INDICATES TIMER
RUNNING.
[2.3] P/MOP - PRIMARY MAINTENANCE OPERATION PROTOCOL, REQUESVlNG REMOTE LOAD.

[2.4] RUN COMPLETE MICRODIAGNOSTICS (MICROPROCESSOR AND LINE UNIT)
[2.5] RUN PARTIAL MICRODIAGNOSTICS - MICROPROCESSOR ONLY
[2.6] INVOKE PRIMARY MOP BOOT (BIT FIVE, BSEL 1)
MK·1974

D-3

FULL-DUPLEX

HALF-DUPLEX

_ _L

RECEIVER NOT
ACTIVE

NO LINE ERRORS
ENCOUNTERED

LINE ERRORS
ENCOUNTERED

CARRIER WAIT
TIMR -1 SEC

[3.1 ]

CD = OFF

CWTIMR
=0

CTS
NOTIFY USER OF
STREAMING
STATION

= ON

CTS =0

CTSTMR

TRANSMIT DATA OR
CONTROL MESSAGE

CD = OFF
HALF DUPLEX

CTS

= OFF

[3.4]

[3.3]

FULL
DUPLEX

NOTES:
[3.1] CWTIMR -- CARRIER WAIT TIMER. IT DETECTS THE CONDITION WHERE THE LINK WAS NOT RELINQUISHED IN
TIME BYTHE REMOTE END.
[3.2] CTSTMR -- CLEAR TO SEND TIMER. TIME CLEAR TO SEND GOING AWAY WHEN DROPPING RTS BECAUSE OF LINE
ERRORS. FALL-OUT COVERS CONDITION OF CONSTANT CTS MODEMS.
[3.3] IN FULL DUPLEX MODE -- DMP11 ASSERTS RTS CONSTANTLY. RTS IS DROPPED ONLY WHEN THERE ARE LINE
ERRORS.
[3.4] SOFTWARE INTERLOCK - CONDITION IS SWITCH SELECTABLE FOR CONSTANT CTS MODEMS.
MK·1963

D-4

(

TRANSMIT 2 )

I
LESS THAN 7 REQUEST
REMOTE PROGRAM LOAD
MESSAGES SENT

TIMER

7 REQUEST REMOTE PROGRAM
LOAD MESSAGES SENT

TIMER

MK-1964

D-5

(

RECEIVE)

FULL-DUPLEX

HALF-DUPLEX

TRANSMITTER NOT
ACTIVE
IN HALF DUPLEX, LINE UNIT
PROVIDES INTERLOCK TO
ENABLE RCV CLOCK ONLY IF
TRANSMITTER IS IDLE.
(RTS =OFF)

CD = ON
SYNCED ON NEW
FRAME

_ _ _~_ _ _ CD DROPPED IN THE
MIDDLE OF DATA
CD
OFF
- - - . . . , . . . - - - - STREAM

RESET CDTIMR
TO 128 mSEC

CD= ON

CDTIMR = 0

NOTIFY USER OF
CARRIER LOSS
(CONTROL RESPONSE)
END OF
RECEIVE

MK-1965

D-6

(

[6.1]

MOD STAT)

WRITE MODEM
STATUS

READ MODEM
STATUS

[6.2]

~-----------------4.~~~.~----------------~
[6.1]
READABLE MODEM SIGNALS:

[6.2]
WRITEABLE MODEM SIGNALS:

CARRIER
CLEAR TO SEND
MODEM READY (DSR)
HALF DUPLEX
REQUEST TO SEND
DATA TERMINAL READY
RING
TEST MODE

DATA TERMINAL READY
SELECT STANDBY
REMOTE MODEM LOOPBACK
LOCAL MODEM LOOPBACK
NEW (POLLING) SIGNAL
SELECT FREQUENCY

MK-1966

D-7

VALID MESSAGE
RECEIVED FOR
THIS STATION

SET CALTMR
= OFF

CALL TIMER = 0

~
NOTIFY USER
OF DISCONNECT
(CONTROL RESPONSE)

TIMER = 0

NOTE: CALTMR = ON
MK·2741

D-8

APPENDIX E
INTERFACE DIAGNOSTIC
COMMAND FLOWCHARTS

The DMPII interface diagnostic command permits a user program to test the data paths from the
DMPII through the CSRs to the main CPU. A typical use of this command is to include a routine
capable of issuing this command and evaluating diagnostic' results as part of the user program. This
routine, then, is executed as part of the system initialization process. Such a routine can also be part of
a system diagnostic package running in the main CPU.
This command performs the following tests with results available for evaluation by the main CPU diagnostic routine:
1.

Set the CSRs (except BSEL 1) to all ones.

Set the CSRs (except BSEL 1) to all zeros. The main CPU diagnostic routine has the option
of looping back to perform test 1 again.

Test input interrupt vector (XXO).

Test output interrupt vector (XX4).

N PR the content of one main CPU location to another. Block NPRs can be done by incrementing the NPR in and NPR out addresses.

To perform these tests, the startup procedure detailed in Section 4.3 should be completed up to and
including the mode definition command. At this point a control command with the run interface diagnostic request key can be issued to perform the tests listed above and detailed in the flowcharts that
follow.

NOTE
When these tests have completed successfully, the
mode, if software selected, must be re-established in
order to continue system configuration.

E-l

DMP 11

PDp·" /VAX·"

TESTA
WRITE ALL 1's
INTO THE CSR's
START OF INTERFACE
DIAGNOSTICS

ISSUE CONTROL
COMMAND TO
DMPll TO RUN
INTERFACE
DIAGNOSTICS

WAIT FOR BSEL 7
TO GO TO ALL 1'S.
IF THE DMP11 DOES
NOT RESPOND IN
.5 mSEC THE USER
PROGRAM SHOULD
ASSUME AN ERROR.

SET EACH CSR
(EXCEPT BSEL 1)
TO ALL l's IN
THE FOLLOWING
ORDER - 0, 2, 3,

4,5,6,7
NO

ERROR

NOTE THAT I N CASES
LIKE THIS THE
1 4 - - - -..... MICROCODE
WILL HANG
IN A LOOP.
WAIT FOR PDP-ll /VAX-ll
TO INDICATE THAT NEXT
TEST SHOULD BE RUN.

WRITE ZERO
INTO BSEL 0

TEST A IS COMPLETE.

'iN'F'ORM OM P11 TO
EXECUTE N EXT TEST

MK-1976

E-2

DMP11

PDP11/VAX11

DTSTB

TEST B
WRITE ALL ZERO'S
INTO THE CSR's

SET EACH CSR
(EXCEPT BSEL 1)
TO ALL ZERO'S IN
THE FOLLOWING
ORDER- 0,2,3,4,

WAIT.5 mSEC
FOR BSEL 7 TO
BE CLEARED.

5, 6, 7

AT THIS POINT THE
USER PROGRAM HAS
THE CHOICE OF GOING
BACK AND EXECUTING
TEST A OR CONTINUING
ON TO TEST C.

ERROR
LOOP ON
ERROR

WRITE 2008
INTO BSEL 0
(MICROCODE WILL HANG)

NO
IF BSEL 0 IS ALL
1's THEN GO TO
TEST C. IF BSEL 0
IS EQUAL TO 2008
THEN GO TO TEST A.

WRITE 3778
INTO BSEL 0

MK·1988

E-3

DMP11

PDP11/VAX11

DTSTC

TEST C
TEST INPUT
INTERRUPT

NOTE THAT INPUT
IS FROM THE
DMP11 's PERSPECTIVE

SET UP
INTERRUPT
VECTORS
XXO=INTCO
XX4=INTC4

(MICROCODE WILL HANG)

WRITE 3778
INTO BSEL 3

START A TIMER
TO WAIT FOR
INTERRUPT

TIMER
VALUE 2: .5 mSEC

FORCE INTERRUPT
TO XXO (RDI)

INPUT INTERRUPT
TEST PASSES
INTC4

ERROR
OUTPUT INTERRUPT
TEST FAILS
MK·1987

E-4

DMP11

PDP11/VAX11

DTSTD

TEST 0
fE'SToUTPUT
INTERRUPT

"OUTPUT' IS FROM
THE DMP11's
PERSPECTIVE

SET INTERRUPT
VECTORS:
XXO=INTDO
XX4=INTD4

(MICROCODE WILL HANG)

WRITE ZERO
INTO BSEL 3

START A TIMER
TO WAIT FOR
INTERRUPT

FORCE INTERRUPT
TO XX4 ROO

TIMER > .5 mSEC
VALUE -

NO INTERRUPT OCCURS
NO

ERROR

OUTPUT INTERRUPT
TEST PASSES
INTDO

ERROR
INPUT INTERRUPT
TEST FAILS
MK-1986

E-5

DMP11

PDP11/VAX11

DTSTE

TEST E
NPR TEST
NPR IN*ADDRESS
(KNOWN DATA)
BSEL 4 - ADDRESS BITS 0-7
BSEL 5 - ADDRESS BITS 8-15
BSEL 2 - ADDRESS BITS 16-17
(IN BIT POSITIONS 2,3)

YES

WRITE 2008
INTO BSEL 0

NPR OUT *ADDRESS
BSEL 6 - ADDRESS BITS 0-7
BSEL 7 - ADDRESS BITS 8-15
BSEL 3 - ADDRESS BITS 16-17
(IN BIT POSITIONS 2,3)

ESTABLISH A LOCATION
IN MEMORY THAT
HAS A KNOWN VALUE
AND A LOCATION THAT
IT CAN BE COPIED INTO

( FROM DMP11 PERSPECTIVE)

END

LOAD THE NPR IN
ADDRESS INTO BSEL 4,
5,2 AND NPR OUT
ADDRESS INTO BSEL 6,
7,3

END OF
INTERFACE
DIAGNOSTICS
INIT

WRITE ZERO
INTO BSEL 0
DO AN NPR IN
BASED ON ADDRESS
IN BSEL 4,5,2

ERROR

DO AN NPR OUT
BASED ON ADDRESS
IN BSEL 6,7,3

WAIT.5 SEC
FOR BSEL 0 TO
GO TO ALL 1 's
BSEL 0 -3778

MK-1985

E-6

DMP11 SYNCHRONOUS CONTROLLER TECHNICAL MANUAL
EK-DMP11-TM-001

Reader's Comments

Your comments and suggestions will help us in our continuous effort to improve the quality and usefulness of our publications.

What is your general reaction to this manual?
In your judgement is it complete. accurate. well organized.
well written. etc?
Is it easy to use? ~~~~~~~~~~~~~~~~~~~~~~~~~~~

What faults or errors have you found in the manual?~~~_~~~~~~~~~~~~~~~~_

Does this manual satisfy the need you think it was intended to satisfy?~~~~~~~~~~~~~_
Does it satisfy your needs? _~~~~~~_ _ _ Why? _ _ _~~~~~~~~~~~~~~__

Please send me the current copy of the Technical Documentation Catalog. which contains information
on the remainder of DIGITAL's technical documentation.

Name _____~_____~~~~~~~~~~~_____~~_________
Title _~~~~~~~~~~_____~~~~_~______
Company_~_~~_ _~~_~~_~_~~_

Street_~~

________________

_______________

City _~~~_~~~~_~_~~~_______
State/Country _ _ _~~_~~_~~~~~_
Department_~~~~_~~~~~~~~~~_ Zip~~~_____~~~~_ _~~~~~~~~_

Additional copies of this document are available from:

Digital Equipment Corporation
444 Whitney Street
Northboro, MA 01532
Attention: Printing and Circulation Services (N R2/M 15)
Customer Services Section

Order No.

EK-DMP11-TM-001

---------------------.Fold Here-------------------.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Do Not Tear - Fold Here and Staple. -

IIIIII
BUSINESS REPLY MAIL
FIRST CLASS

PERMIT NO. 33

MERRIMACK. NH

POSTAGE WILL BE PAID BY ADDRESSEE

Digital Equipment Corporation
Educational Services Department & Publishing
Continental Blvdy (MK1/2M26)
Merrimack, N.H. 03054

- --

No Postage
Necessary
if Mailed in the
United States

Digital Equipment Corporation. Bedford, MA 01730