Digital PDFs

EK-DMV11-TM-001

2000

210 pages

Original

48MB

view download

OCR Version

15MB

view download

Document:	DMV11 Synchronous Controller Technical Manual
Order Number:	EK-DMV11-TM
Revision:	001
Pages:	210
Original Filename:

OCR Text

EK-DMV11-TM-001

DMV11

Synchronous Controller
Technical Manual

EK-DMV11-TM-001

DMVI1
7

TM.

Synchronous Controller
Technical Manual

PREPARED BY EDUCATIONAL SERVICES

OF
DIGITAL EQUIPMENT CORPORATION

1st Edition, September 1981

(

~ Copyright © 1981 by Digital Equipment Corporation
"The matenalin this manual is. for mformatlonal purposes and is
subject to change without notice.

Dlgltal Equipment Corporatlon assumes no responsnblllty for any

o CTFOTSWhlch may appear m thlsmanua]

Printed in U.S.A.

This document was set on DIGITAL’s DECset-8000 computerized

//,-uc-\\

typesetting system.

The following are trademarks of Digital Equipment Corporation,

Maynard, Massachusetts:

DIGITAL
DEC
PDP

DECsystem-10
DECSYSTEM-20
DIBOL

DECUS
UNIBUS

EDUSYSTEM
VAX

MASSBUS
OMNIBUS
0S/8
RSTS
RSX

VMS

IAS

CONTENTS

Page

PREFACE

ok ek e

N +—
W DN —
—
N
W

INTRODUCTION
INTRODUCTION.....ccoii eINTRODUCTION TO MULTIPOINT........c........... e ——————————————————
DMV11 GENERAL DESCRIPTION .............. e ——eaeeeeeeaaeeaeeerererr————————
STANDARD APPLICATIONS ...
DMV11 SYSTEM OPERATION ..., e
COMMAND/RESPONSE STRUCTURES ...ttt
Input Commands ....... S
S
S
Output Responses ............eueeeeen ett eteeeeeeeeetiteiieeeeeeeetettraaaeeeeeeerttra—aaaaerans
PROTOCOL SUPPORT........ootiiiiiiiiiiiiiiiiiiiiiitietvieeeereenessseeennaesanensseranrerrerrranreann...
Data MESSAZES ..uuueiiiiiiiiieeeeiiiie e e eeteee e ee e e e ettt e e e e e tat e e e eatb e e eaatr e et eaaans
Control MESSAZES...ccvvvuniieiiiiiiiiee et e e e e v e e e e
MaIntenancCe MESSAZES ....ccvvvuueieiiiiieeeeiiieeeeiee
e e et e e et e e e ert e e s ran e sran e
GENERAL SPECIFICATIONS ...,
Environmental Specifications..........ccccoevviiiiiiiiiieeeeeieeeeeeeccccciiiireereeeeeeeeeeeeees
Electrical SpecifiCations.......ccouuuiiiiiiiiiiiiiiee
e
Performance SpecCifiCatiOnS .........cuvviuiuiiiiiiiiiei i
e e e e eeeeaeeen

1-1
1-1
1-1
1-3
1-3
1-4
1-4
1-6
1-6
1-6
1-6
1-6
1-8
128
1-8
1-8

CHAPTER 2

INSTALLATION

2.1

2.6.4

INTRODUCTION. ...
2-1
UNPACKING AND INSPECTION ....cooiiiiiiiiiieeiieeeeeeeeeeeeceeeeeee, e 2-1
INSTALLATION CONSIDERATIONS. .....ooriiiiieeeeeeeeevveveeer
e 2-1
PREINSTALLATION CONSIDERATIONS ... 2-2
Device Placement .......oovveiiiiiiiiieiccee
ee
2-6
System ReEQUITEIMENTS ...cvviviiiiiiiiiiiiiiiiieiiiieeieeeieeiirrereeereeeeeeeeaererrereareererrereernaee.
2-6
INSTALLATION ..o, eeeeeeeeeteeeeeeeeeteee————————————aaaaeeeerrrrrens
2-9
DMV 11 SYSTEM TESTING ...t 2-10
Functional Diagnostic Testing .........ceeeiiiiiiiiiiiiieii
e 2-10
DEC/X11 System EXerciser .......cc.ccceeuveeuneeee ettt e, 2-10
Final Cable ConneCtiOnNS..........cuiviuiiiiiiiieeeieeeeeeieieiiiiiceeie
e e e e e e e eeeeeeaeeaenenae s 2-10
DMVI11 LInK TeStING...ccciieeiiiiiiiiiee ettt
eteee e e e e e e e ereaae e e e 2-10

CHAPTER 3

COMMAND AND RESPONSE STRUCTURES

W L L L LW LWwWw

Ptk pd

e R RN

ek e

ek ek

R S

CHAPTER 1

INTRODUCTION.....c oo
3-1
COMMAND STRUCTURE ..o
3-1
Control and Status Registers................ e
——eeeeeee e ———————eeeerrat——————aarennnraan, 3-1
Input Commands OVEIVIEW .....cccoeeeeiiieiieiieeeeeeeieeeeeeeeeeeeeeeeeee
e - 3-5
Output Responses OVEIVIEW .......ccoccuiieiiiiiiiiiiiiiiiiiiieee e
3-5
DMV 11 INPUT COMMANDS ...t
e e e aeee 3-6
Microprocessor Control/Maintenance Command...........cccoecveeernviiieennnnnneenns 3-6
Mode Definition Command........cooeeeieeeiiiiiiii 3-6

2.2

2.3

2.4
2.4.1
2.4.2
2.5

2.6
2.6.1

2.6.2

—i

2.6.3

1ii

CONTENTS (Cont)

Page

3.5

Control CommMmAaNnd ..........cooveviiiiiiriiieeee
e eeeecccieeee e e eeerirreee e e s e e siiereeeeeeeesesienes 3=
Buffer Address/Character Count (BA/ CC) Command ........cccovneveiiennnen. 3-16
DMV11 OUTPUT RESPONSES ...ttt eerrerreeceeeee e ... 3-18
Buffer Disposition RESPONSE......ccevveiiiiiiiiiiiiiiicieiceninncccccciininnei 3-19
Control Response...........ccoevvvvvviiiinennnnn. s eiereeetitieseisiasieeerennsnnaeersanerrasans 3-20
Information Response.........ccccuueennnne. triiieeeesereeeiereeneirterrnnataiasaeassseaeeranerenas 3-26
TSS/GSS ACCESS ............... eeeeiieetereeeiieneaaense aaaaaaas veheeraeeeereieaeaeesrnsnrens 3-26

CHAPTER 4

PROGRAMMING TECHNIQUES

4.1

INTRODUCGTION. ..ot
s s s s e s 4-1
COMMAND/RESPONSE DISCIPLINE AND HANDSHAKING ................. 4-1
Command DISCIPINE.......oooiiiiiiiiiieee e 4-2
Retrieving Responses.........c..cce.......eeberetieseieeeerunssserna Y serraaaeaeanssterarntararesaes 43
CSR Interface Interactions ...........ccceevvvvunenennnn. ieeerieiseeseriresreeernennrrernreresrarannans 4-3
DMV11 START-UP.....ooooviiiiiiiiieeieeee, rereeeeenan e teteieieeeeeeeanaaea——————————————_, 4-3
Configuration Procedure ........cccovvveeeeeiiinerennnee. e aaaaaeaeaaaeaas 4-4
Specifying User-Defined Parameters ...........cccccceviinniiiiiiii, 4-4
Specifying TSS Parameters ................ eriieeiidenataens eeeeeeeerrrr————————aararrana— - 4-6
|
Specifying GSS Parameters.......c.cocovuuiiieeiiiniiiiiiiiiiiiiiiceee, 4-11
Protocol OPeration..........cuviiiiiiieiiiiiiiiiiiieeiee et 4-13
CRITERIA FOR DETERMINING COMMUNICATIONS
L
|
LINK PARAMETERS. ..., e erereeeereraa————————————————————— 4-13
Setting the Selection Interval Timer ......... e iieeereeneeiiiitheseteese e e seenenensS 4-14
Setting the Babbling Tributary Timer................. eeeereivevetherarraeeeareeerenuaasaeeas 4-16
Setting the Streaming Tributary Tlmer ................ eeeer
e e e e ra— e aaaaaaaans 4-16
ERROR COUNTER ACCESS ...ttt 4-17
Reading the Counters ............ccoevvevrvviiiireieeeenenennnee.et
ee e
e eearerar i aaaans 4-17
COUNLET SKEW..oiiiiiiiiiieieeeeeeceeiicee
e eierer e Cereeiressennnrrrrnnasaasessasarernnns 4-17
ERROR RECOVERY PROCEDURES ...ttt 4-17
Recovery from Network EIrors ......cc.cvvecviieiiiiiiniciecieciieccieneiicc e, 4-18
Recovery from Threshold Errors.......... erredandnnisas erveerainarsaninesearanes e 4-18
Recovery from Babbling and Streammg Tributary Errors .................... 4-18
Recovery from Procedural Errors............U UUPUUUUUPRPUPRRPPPRY - T i .
Recovery from a Nonexistent Memory Error ... 4-18
Recovery from a Receive Buffer Too Small Error..........cooeininn. 4-19
Recovery from a Queue Overflow Error.................rreereeneeeeeesirnreeesnnnens 420
BOOTING A REMOTE STATION ... et et eerr e ————————————————————————. 4-20
- Steps Leading to a Remote Load Detect BOOt..........cocoviiiiii 4-21
Steps Leading to a Power-On Boot........... R rvinninnnrenien e aaaaaeans 4-21
Steps Leading to an Invoke Primary MOP Boot........ccccovviiiiiiiinns, 4-22
DMV11 Switch Settings for the Boot Functions ...........cccccooniinnie 4-22
Switch Settings for the Power-On Boot Function........c.ccovvuniiiiennnninnn, . 4-22
Switch Settings for the Invoke Primary MOP Boot Function................ 4-22
Switch Settings for the Remote Load Detect Boot Function................ 4-23
MAINTENANCE REGISTER EMULATION........ eebirivveasserreereasaeaaeeeersrsisesesees 4-23

3.3.3

3.3.4

34
3.4.1

3.4.2
3.4.3

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
4.5

4.5.1
4.5.2

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
4.8

CONTENTS

Page

PREFACE

N N

W NI =

N —

CHAPTER 1

INTRODUCTION

INTRODUGCTION. ..ottt ettt
eressse e e s e e e e e e e e e e e eaeetaebeabebebbaaeeeeeeeeaeesINTRODUCTION TO MULTIPOINT........ eeieeeeeieeeterettittteataaaeaiaaaaaaaaaaaeeeaersrasns
DMV11 GENERAL DESCRIPTION. .............. e eereeertterretraseaaaasaaaaaaaaaassanss
STANDARD APPLICATIONS ...
DMV 11 SYSTEM OPERATION ....ooiiiiiiiiiiiiiiiiiiiiiiiieeieteeeeeeeee
eeeeeeeeeeeeeeeenes e
COMMAND/RESPONSE STRUCTURES........ etttterrtr————————————————arraaaatataaarraanes
INput ComMMEANAS ...veeiiieeiiiiieeieeeceee et
Output Responses .........ccceeeenene. ettt ettt teeeetetttee.eeeeeseseeeeeeteeteteeeeeeetererrrennaaaaaaan
PROTOCOL SUPPORT.....uiieeieieeeeeeee e
DAt M ESSAEZES .uuuueiieeeeieeeeiriiiiiiiiire
e e e eeeeereeeettneereeri
e se e se sees ses
eeaeesiaaaans
Control MESSageS.....ccvuuieeieiieriiiiene
et
eeeereei
e e r———————
Maintenance MesSages......ccoeveveeiireiinnreeeinneennnns eresreeerranrerearhaaarerraarararreasens
GENERAL SPECIFICATIONS .................. vesessessssssssssssrnresinnannnansssnsssssesarasanrasane
Environmental SpecifiCations.........ccevviiiiiiiiiiiiiiiiieiceeeeeee o
Electrical SpecifiCations........cceeiieeeiiiiiiiiiiietteeeee e
Performance SpecifiCations ........ccuvieiiiiiiiiiiniiiiiiieieeeeeeee e

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

CHAPTER 2

INSTALLATION

2.1

2.6.4

INTRODUCGTION. ..ottt
vaaaaseaassenbassesseaseeaeeeeeeeeeeee et eeereeneeeses 2-1
UNPACKING AND INSPECTION ...eeeeeeeeens e 2-1
INSTALLATION CONSIDERATIONS. ... 2-1
'PREINSTALLATION CONSIDERATIONS.......coiiiii e 2-2
Device Placement .........ccooeiiiiiiiiiiieniieiciiiiiiieee
e, rrereeerseenrennsraneeaeseasaese 2-6
System ReqUITEMENTS .....cuvviieiiieiiiiieeee et 2-6
INSTALLATION ....ccccoiiiiiiiiiiiiccrecn, P 2-9
DMV 11 SYSTEM TESTING ....ootttitiiiiieiieeiiceceeseee s 2-10
Functional Diagnostic TeStINg .........uuuuuuuuiiiiiiiiieieiereeeeeeerveereevaeeeeaes 2-10
DEC/X11 System EXErCiSer ......ccccovviiiiiiiiiiiiiiiiicieciecicc
s 2-10
Final Cable Connections............ e —————— e ———————————— 2-10
DMV 11 LinK TesSting......cieieiiiieeeeiiiaee et eceene e e e renie e 2-10

CHAPTER 3

COMMAND AND RESPONSE STRUCTURES

2.2
2.3
2.4
2.4.1
2.4.2
2.5

2.6
2.6.1

2.6.2

N
W
i
NS

W L L0 LW L) W L

2.6.3

INTRODUCGCTION......ootttititiiieiiiiieieesesseese
s e sesssss s s e
sessesssannnaes
COMMAND STRUCTURE ...ttt
e s e reeeeresseaeseeeees
Control and Status Registers................ ettt ——————eeaaaaaeeaaaeeeeeeeeereees
Input Commands OVETVIEW .........ueeeeiiriiuiiieeeiiiiiereeeiiieee
e sraree e Output Responses OVEIVIEW ........ccceiriiiiieieiiiiiiiiiiiiiicn e
DMV 11 INPUT COMMANDS ...t
Microprocessor Control /Maintenance Command...........ccccceivviiiiniiriinnnnnnn,
Mode Definition Command .............eueeeiiiiiiiiiriiieeeeeeeeerevaeeaeeeaaae

111

3-1
3-1
3-1
3-5
3-5
3-6
3-6
3-6

CONTENTS (Cont)

Page

3.3.3

Control Command ..........ooouveeeveeeiieieceiee
et eeee e
eeereeesreeesnneeenneeeene

3.3.4

3.5

Buffer Address/Character Count (BA/ CC) Command ......ccoeeeeevveeevreevvnnnnnns 3-16
DMYVI11 OUTPUT RESPONSES. ... ..ot... 3-18
Buffer Disposition ReSPOnSe.........cccvvveeiieecviiieiiecciieee
e
eeeccveeeeeeeeeenveeeee.. 3-19
Control Response.......cccceeveevieevivveevnnnnnnn. eeieaerieneereaiesettt —————————— 3-20
Information Response............cccuuueeeeeee e ieunessieesesereeeeieesereeieinenterraaaeeteeerereees 3-26
TSS/GSS ACCESS ... vveierereeeinerreaeserereereerernsennnnnnens 3-26

‘CHAPTER 4

PROGRAMMING TECHNIQUES‘

4.1

4.2.3

INTRODUCTION .................................................................................................
COMMAND/RESPONSE DISCIPLINE AND HANDSHAKING .................
Command DISCIPINE ......coeeeiiiiieiciii
i
ee
Retrieving Responses...................................,....;..‘.'...... e eeeeeerrr————aaaan
CSR Interface Interactions .................. Rettt ———————aeaeeeeereraraa————

4-1
4-1
4-2
4-3
4-3

4.3

DMVI11 START-UP....ooovriiiiieeeeeeeeeee e, suenisniabnsnsnssssssnnagassnassnssnnans

4-3

4.3.1

Configuration Procedure ....... rivviererheestieresieaerareeans ervarivesrereresseessrannarraraeseaes
Specifying User-Defined Parameters ..........uuvuvieriieiiiiiieiiieiieiiieeiiiieiiiaen
Specifying TSS Parameters ......cccceeeeeeeiiiiiiiiinnnnnnn, eeteeeeeerrrer——aeeeeeera—.
Specifying GSS Parameters.........ccceeeeeeiieeeeeennnn. et ——————————————————
Protocol Operation...........ccccccvuuiiriviiiiiieeieeeeeee
e ee e, e————————————————
CRITERIA FOR DETERMINING COMMUNICATIONS
.
LINK PARAMETERS ..., eterthrr——————————————————————
Setting the Selection Interval Timer ...........eevererrreranriiettirerthnnineenseseartnnanrarenns

4-4
4-4
4-6
4-11
4-13
’
4-13
4-14

34
3.4.1

3.4.2
3.4.3

4.2
4.2.1
4.2.2

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
4.5

4.5.1

4.5.2

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

4.8

379

Setting the Babbling Tributary Timer................. ereereeererererr————aaeeartraaaaaaaenns 4-16
Setting the Streaming Tributary Tlmer ................ e eeeeeeerettrr——aeeeeeearaaaanans 4-16
ERROR COUNTER ACCESS................. ST eveeensuensinniivitnnnnnnsesanannnnnnnnnane 4-17
Reading the COUNLETIS .....coovvuiiiiieiiiieee
e
e
eevaa e - 4-17
Counter SKeW.....coeeeeeeeiiiieierieieiiiceeeeen, eveiieens Cevererineisererneereeerresiensasaesaeereens 4-17
ERROR RECOVERY PROCEDURES ... 4-17
Recovery from Network Errors ..o, 4-18
Recovery from Threshold Errors........................... eenneeieieseeruuaeaseererannns - 4-18
Recovery from Babbling and Streammg Tributary Errors .................... 4-18
Recovery from Procedural Errors............ettt e rrerrreaeeeeeerrrratineeeeesesersnineeeeeeeenes 418
Recovery from a Nonexistent Memory Error .........vveeeiiiiiiiiiiinnnennnnnes 4-18
Recovery from a Receive Buffer Too Small Error...........oooveviiiinieennn. 4-19
Recovery from a Queue Overflow Error...........rrreeeenrrrreeeeenensnneneeeeeennns 4220
BOOTING A REMOTE STATION .........cooeeveeee. e
ee e r—————— 4-20
. Steps Leading to a Remote Load Detect Boot.........cccccce.......eereee
e e ———— 4-21
Steps Leading to a Power-On Boot .......... Siemeseeenns veribesees teeirerheiereeeseerrnannaerenens 4-21
Steps Leading to an Invoke Primary MOP Boot........ccoiiiieiiiiiiiieieniieennees 4-22
DMV11 Switch Settings for the Boot Functions ...........ccccccvnnniiiiinnnn. 4-22
Switch Settings for the Power-On Boot Function..................... reeeeeeeen 4-22
Switch Settings for the Invoke Primary MOP Boot Function................ 4-22
Switch Settings for the Remote Load Detect Boot Function................. 4-23
MAINTENANCE REGISTER EMULATION......ccccccviiiiiiis 4-23

CONTENTS (Cont)

Page

CHAPTER
5
5.1

INTRODUCTION .................................................................................................

5-1

5.2

DMV11 POLLING ALGORITHM ....................... e

5-1

5.2.1

Calculating Polling Urgency.......cccoeeeeeeeennnnnne. SR S

5.2.2

Criteria for Determining Polling Parameters Cieieeieeereenasaessanees [RSTURRURUURRUPR.

ORTRRTRPOT .

Determining Values for Q and R ...........................................

Determining a Value for Poll Delay............... eereeeeeeeeeeeeeeererereer————————

5-8

Determining a Value for DEAD T ...oooviiiiioeeeeeeeeeeeeeeee e e

5-8

- 5.2.2.2

Determining a Value for DELTA T ...oooviivioieeeeeeeeeeeeeeeeeeee e, - 5-6
5-7

o Ptk

—

B W N —

— O 00 ~d N L B W N —

ERROR COUNTERS ....... ettt
—————— ettt aaaaanan e

W N =

e VRV RV EV EVRVEVEV RV NV RV RV RV RV NV RV RV NNV RV RUNT R RV RN
NSW
n PRRARRLWWLWWLWLWWWWWLWWLWWWWWLWWLWWW b

)

ASPECTS OF DMV11 MICROCODE OPERATION

5-8
Data Link Error COUNTETS...cuum i eeeeeeeeeeeieeeeeeeeeeeeeeeeeeeeeeeeseesesesessesseasannieens
59
Data Errors Outbound ............................................................................
5-9
Data Errors Inbound...,v 5-11
-

Local Reply Tlmeouts TS
A
SMeerrereeanreeererrenteseenrrnrnressians 5-11

Remote Reply Timeouts......coooooevveviiiiiniiieiiiiiinnen, eeeeeeeerraaeeeeranaa. v =11
- Local Buffer Errors........e reitie e raas AT et et e et reneetaeennaaesneienns =12

Remote Buffer Errors...........e e i e re i aeteiaseen e ran e reatb e ran s 5-12
Selection Timeouts .......oeeevvvuneeennnn.... e
e
v erseeseannansassaernannn 5-12
Data Messages Transmitted.................. hebvientrassserternarteseerrasienienertrrnssenss 5-12
Data Messages Received.................... iiesveirinnsreeresneansesncennanronnas e 5-12
SeleCtioN INtErValS . .o i
e 5-13
Station Error COUNTETS couvveiin
e
e e e e eee e enns rerraeaees reereraranns 5-13
Remote Station Errors...ccceeeeeeeeeeeeeeeeeeeeeeeeeeeeeeennn, i e veeveeeenraeriarairaiis 5-13
L0Cal Stat1ON BrTors ..o
e ee e ee e e e e e e eees 5-13
Global Header Block-Check Errors ....o.ovimeiiiiiieeiiee e 5-14
Maintenance Data Field BIock—Check Errors........................................ 5-14
THhreShOld ErTOr COUNTETS ..coeveeeeeeeeeeeeeeeeeee
e e e e eeeeeeeesieesaseeeeeresaseeeeeeeaaaasesesaes 5-14
Transmit Threshold Errors....cceeeeveeeiieiiivnnennnn... i eeseuersasanreaatsnaannrrannes 5-14
Receive Threshold Errors................ere et taees e
——aeaaa e e et ——————— . 5-15
|
Selection Threshold Errors.................e ieiesevnteserernnneseernanseseannnnseens reereeens 5-15
DMV11 MICROCODE INTERNAL DATA BASE OVERVIEW ..o, 5-15
Linked Lists ....cccccevveeeninnncnnesettt
a e saesae e saneae - 5-15
The Free Linked List.........cov........eieeerenaeans et ieieereeeeeeererrrrreareaaaaannees 5-16
The Response Linked List ......ouuuuiiiiiiiiiiiiiieieeccceeceirece e 5-18
Buffer Linked Lists ................ SV i iieennliaiieset
e eeeeeeeenaeneeneees O-18

‘Slot Mapping Table.......... ereeerieiensrebesein ieieiiereereerirneesiesrrreeseeanrreees 5-19

5.4.3.1

TSS and GSS Structures.......... .SUUTRTTR e e ret e ethee et e et reraneaay . 5-19
The Global Status Slot (GSS) ...... Visseisresninasntonsesreeerdebeiseurerseeresesersrnnnnnnen 5-19

15432

Trlbutary Status Slots (TSS)............‘..- ..........eiereteeeiveerieeeeeeernnanaaraanea——es 5-19

CHAPTER
6

TECHNICAL DESCRIPTION

6.1
62

INTRODUCTION .................................................................................................
LOGIC DE S CRIPTION oo
e
e
e e
e e e eaeaan

6-1
6-1

6.2.1

Control and Address DeCOAEr . ...
s

6-1

CONTENTS (Cont)

ok ok
bk

The 6502 Microprocessor ........ett

ettt et ea e ettt ettt —taaraaaaanns
TIMINE CITCUILS coevviiiiieeeieeiiiiiiieee
e e e eeereeeeeeeeeeeere i eeaeeeeeseessnnnaeeesesssssnnns
6502 Data and Address Interface.................. eeerrer—— tere
e ————
Address DeCOAETS .....ouuuuiiiiiieiiiiiccie
e
e
ee

B U R

N O\ O\ O
b R B b

Page

6.2.
6.2.2.1

6.2.2.2
1 6.2.2.3
6.2.3
6.2.3.1
6.2.3.2
6.2.3.3

6.2.4
6.2.4.1
6.2.4.2
6.2.4.3

6.2.4.4
6.2.5
6.2.6

6.2.7
6.2.7.1
6.2.7.2

6-1

6-2
6-2
6-3

[/O Data BUs......ccoovveeeniieiiieeeeieeenseeeressessaneeresarestesntsesessatesressatttesaranes
6-3
USY RT ettt
e s e etr e e s e e a e e e s st e e e s s s nees
6-3
USYRT CONtrol ...coooviieiiiiiiieeeeeeee
et
e e e eenan s
6-3
Line Interface Control.......cccocoovviiiiiiiiiiiiii e
6-6
DMV11 Memory ............. eteeeeeteettaeeeeataeneearaaeaenaeaararrraenns erreeeeeaaan——ns ...
6-6
ROM Control StOrage........cuuvveeeiieiieiiiiiiiieeeeeeeeeiieeee
e e eeeeveeeee e e e e eeeeaaans 6-6
RAM ettt
e e e e e e e e e e e e e e e e e e aaaaa—————
6-6
NPR In/Out Registers......cccceeuverernuennnns ettt
ettt et e e et e e et eeeeabaeee s 6-6
LSI-11 BUus INterface ....cccooveviiiiiieeeeeeeeeeccee
et
6-7
LSI-11 Bus DAL Interface ........oooovveeviiiiiiiiiieeeeeeeeeeee e 6-7
CSR Controller............. et eeeeetrttteeeeeestntaeeeeertntaaaeertatr—aaeertartaaaaerraaaeaans
6-8
INterrupt CONtrOIIET . ....covvviiieiiieiieetee ettt
et
eene
6-8
NPR Controller.....coooviiiieeiieeeeeee
e e e ee e e e e
6-9
Memory and Reset Control ............. [UUPUTTURett
e e e e e eeeeeeeeeaeeeaeeeeeeaaaeaaeaaeaeeaeas
6-9
Modem Interface.......................reteerrensssrneeserarrantrnneseserrrnsnstenareseesrtrnrrnnnseesartes
6-9
Integral Modem......... et eteeeeeeeesttsttteeeeeeeesestttiaeeeeeteettstnnaaeaeteeerrrtrnnaaateaertrrnnn 6-10
RECEIVE ..ttt
et
e e e e e e e et e e e aaae e e eenaeaansnans 6-10
TTANSINIL ooevniieiiiee
et ee et eee et e e e e e e reraeeeesseasaseeesaannnseesnennnns 6-10

CHAPTER 7

SERVICE

7.1

SCOPE ...t e ————————————_
7-1
MAINTENANCE PHILOSOPHY ..., 7-1
TROUBLESHOOTING TECHNIQUES FOR MULTIPOINT.......................... 7-1
APProach ......ocoovvieiiiiiiiccice e, eeeeeeeereeeeieeeeeeeeeterereeaeatrarnas
7-1
Error Counters........ccccccuuunnnn... areees e eeeeeetttteeeeetetteeeeetettraaeeeerataaaeraraaaaaeres
7-7
Data Link Error Counters ..........ccccuuu.....ett
ettt ————————————aaaaseeens 7-7
Station Error COUNTETS ..oovvvviiiieeeiieeeeiiieeeieeeeverceee
e e e eeeeetee e s e e e eaeeea e 7-11
ThreShOld Error CoOUNTeTS. . cee ettt
et et e et e eeneeenseanesraneran 7-12
Error Counter Analysis..........cccevvvvvvervnnnnes etteeeeeeeerreere—aeeseeereera—aaaaaeanns 7-13
MAINTENANCQ
E. ...ttt
e e e e e e e e e e e e e e et
sae e s e e aeeeaeeaees 7-16
Maintenance Mode .......coooovviviiiiiiiiiiiiiiiieieee
e, e ——————————————— 7-16
Standard Operating Mode.........coooiiiiiiiiiiiiirrree e 7-16
PREVENTIVE MAINTENANCE (PM) ..o 7-17
CORRECTIVE MAINTENANCE ... 7-17
DMV11 Static Logic Tests Parts 1 and 2........oouvuiiiiiiiiiiiiiiiiveceeeeeee, 7-17
DMV11 Static Logic Tests Parts 3,4, and S....cccooeeeiiiiiiiiiiiiiiiciceeene
e 7-19
DMV11 Functional Diagnostic.......................ettt ——eeeeeeeretr———————aaerarraa 7-21
DMYV11 Microdiagnostic Error Reporting .........cccceecveeeeieviviiiiiiiennns 7-21
Data Communications Link Test Program (DCLT) .......coovviiiiiiennniininnnee, 7-21
DEC/X11 DMV11 Modules.......cccccevuiiiiiiiiiniiiiiiiciiiiiicinicneeee
e 7-25
DM ..t
e e e e e e e e e ee eee
e e ae 7-25

7.2
7.3
7.3.1

7.3.2
7.3.2.1

7.3.2.2
7.3.2.3

7.3.3

7.4
7.4.1
7.4.2
7.5

7.6
7.6.1

7.6.2

7.6.3
7.6.4

7.6.5
7.6.6

7.6.6.1

CONTENTS (Cont)

7.6.6.2

7.6.7

AT
T V

> %>

APPENDIX A

|5 1LY 1 2
S SO USSP
UPPRSRPPP 7-26
Soft Error Reports Under DEC / Xll ............................................................. 7-27
DDCMP IN A NUTSHELL

DDCMP ........................................................................... - A-l
~ Controlling Data Transfers .............. e ehieEiiraeenasaradiieaisieitritareeeesesssanraraaaaren A-1
Error Checking and ReCOVETY ........uuiiiiiiiiiiiiiiieeeee
e eeecerieee e
ee e, A-1
Character Coding .......ccceeeevvvvvvnnnn. eeribieiies eetemsessiesedissesiersesessesertnnnnnransesans - A-2
Data Transparency ......ccccceeeieeviuiinerieeininereecenennnn. itlerereeeirnsntsene
e sssssnatnesssssanns A-2
Data Channel Utilization................... eiaeesereerebederensaeeiantresenererenanessi A-2
PROTOCOL DESCRIPTION .....coovviiiiiiiiiiiieee, e lien i e st eeiesabaneresanreraesanressens A-2
MESSAGE FORMAT .....ooovirieveviveeveeeeeee creveseiiiiseniessinannnerteraaaeaaanes A-4

'APPENDIX B

FLOATING DEVICE AND VECTOR ADDRESSES

FLOATING DEVICE ADDRESSES..............cooovoo...
...... S B-1
FLOATING VECTOR ADDRESSES
...
it B-1
EXAMPLES OF DEVICE AND VECTOR ADDRESS ASSIGNMENT ......... B-5

APPENDIX C

MODEM CONTROL REGISTER FORMATS‘

C.2

MODEM CONTROL REGISTER FORMATS i Eieveriveienesasanan Viveessiessssnmnnadunsnns
RS-449 VERSUS RS-232-C.....cvvvrriiiriiiiieieeiieneninnnn rieessissesssbesinrennnnnninssnreesereranans

APPENDIX D

MODEM CONTROL |

UUU?

C.1

1.1
1.2

C-1
C-4

MODEM CONTROL ..o D-1

Hardware Modem Control................. Ceereertesiarierneesietennuntansasssarasasennseseseananannnss D-1
Modem Control Implemented by the DMVll Mrcrocode ............................. D-1

FIGURES

Figure No.

Title

Page

DMVI 1s Used inPoint to-Point Applications .............. USRIeeeeeeeeeeeeererrr——————— 1-4

DMV11s Usedin Multipoint Applications ..............eeeiecieeceeeeeenns e 1-5
General Summary of DMV11 Command/ Response Structure......veerererererrrennnnnuans 1-7
Local NetwWork TOPOIOZY....uuurieiiiiiiiiiieiiiiiie ittt 2-3
Remote Network Topology..............heietienenanaaesirerriasinentiterransntaneesiirenastaaesenaarannennns 2-4
M8053 Switch Locations .......... e e SPRRRON FERURN veeeeens eveees 2-7
M8064 Switch Locations ......... ereeeenierereeeerrrniantanisaenaraniesv ieederivabesnsnnreieeereeresasrananns 2-8

vil

FIGURES (Cont)
Title

Page

DMYV11 Switch Selectable Features............... et etttieeeeeeeeeeeeeeeeerertrtnt——————————————————1 2-13
Test Connector Insertion for the M8O0S53......oiiiiiiiiiii
e 2-16
Test Connector Insertion for the M8064.............S e tteeeeeeeeererenniaeeeeeeserrersnnnn 2-17
DMV
11 TeSt CONNECLOIS «.euniinniinriieeeeeeeieeeieeieeieeeeerteeeneeereesnens et
———— 2-18
DMV11 Cable Drawings..........ccccvvvveeeennnnne.ereeeeeererraa——— crreeereererrrrnnnaeeeeeaerssresnnees 2722
DMYV11 Remote System Cabling Diagram .........cooeeveviiiiiiiiiiiiiiiiiiiinnieeieeeineeeeeneenee 2-25
DMV11 to DMV11 Integral (Local) Modem Cabling Diagram (Point- to-Pomt) 2-26
Half-Duplex Multipoint Network (Control Station End Node) ...........cccccceni 2-27

et BN BN

]

W= OO

o) e ) We) W) le )

UNHAWNEF

Y I

==
—_—O

]

IV, IRV, IR, IR

DODO00JO0NW

]

bbb

]

D — W= 200G RN

WU WwWwLWwWwWwWwWwuwwwhN

Full-Duplex Multipoint Network (Control Station End Node) ..........ccceveveennnnnnnee. 2-28
Full-Duplex Multipoint Network (Control Station Inner Node) ...........cceeveenneeen, 2-29
DMV11 CSRs Byte and Word Symbolic Addresses ........oocvevviiiiiiiiniininiiinneenn, 3-2
Fixed and Variable Formats for Commands and Responses.............cceeeeeeveveennnnee. 3-2
Microprocessor Control /Maintenance Command Format ........... et e 3-6
Initialization of the DMVI11 .., err——————————————
3-7

Mode Definition Command FOrmat.......ccccceoevvveeeiciiieciiiiieeeciieeeeecveeeeecveeeeeeneess

3-8

Control Command Format.......ccccccvvvviiiiiiiiiiiiiienieeneeceeeeeeeeceeeeeeen.eeeeeeeeeeeeeeeeeeaaraaaaas 3-9
Buffer Address/Character Count Command Format ............................................. 3-17
Buffer Disposition Response FOrmat .............eevveeemeiimeieieieiiieee, 3-19
Control-Out Command Format 18-Bit Mode.........oooviiiiiiiiiiiiieiiirceens 3-26
Information Response FOrmat ...........ooviiviiiiiiiiiiiiiiiiiiiiiineeenireeeeeeeeeeeeeeeeeeeeee 3-27
CSR Interface Control BIts ........coooiiiiiiiiiiiiiiiiriere
et 4-2
CSR Access Window ..............e aressesstsssesieceesesiesersrsrnnnninerriertesaanineeiesassannnsennnenranes 4-4
DMV11 Maintenance Loop Command Format......occoeveeeiiieeiiereeee e 4-24
Interrelationship Between Polling Parameters Q, R, and DELTA T .................... 5-3
Relationship Between Polling Parameters Q, R, and the
Minimum Polling INterval.............ooiiiiiiiii et
e e e 5-4
Relationship Between the Default Values for Q and R for the
Three Polling ACtiVIty LEVELS .oooviiiiiiiiiiiiiiiiiiieeieeeieeeeeeceereereeeeeceeeeeeeetrereeee
e e ee e
5-5
State Diagram of Polling State Transitions ........ e ieveertresresnnsieaseseeeeeeeeteseseeertereenenns 5-7
Data Link and Threshold Error Counters........ccouiiiiiiieiiiiiieieeiiicceieneeneeeeeeeeeeeeeeeeeennn 5-10
Station Error Counters.......oovvveeieieeeeeiiiiiieeeecceeiricee
e eeeeeeeteeeeeerrrrer———————— 5-11
Data MemOTry Map ...cooooiiiiiion ittt 5-16
DMYV11 Linked List Structure FOrmat......oooooeiiiiiiecee e 5-17
Standard Link BIOCK ......coeeeieeeieieeeeie
e e 5-18
G1ODAL STATUS SIOt cuurriiieiiieieeieeeeeee
e
e reee s e se s resee s e e e e e e e ee e e s e e e s e e e saaas 5-21
Tributary Statts SIOt .....ooiveiiieiiiiiiiiiiiiiiiiiiee bbb 5-23
DMV 11 BloCK DIagrami....ccccceiiiiiiiiiiiiiieeccecceeeeeeee
e 6-2
Control and Address DeCOACT .......ouvvviiiiiiiiiiiieeceeeeeeeeeerreeeeeeeeee
e e e eeeeees 6-3
[/0O Data BUS ...ooiiiieiiieeiiecieeeteeec e
6-4
USYRT Timing Diagram .........coovviiiiiiiiiiiiiiiiiiieceeieceeecteeeeeeeeeeeeeeecreeeereee
eeeeeeeereeeens 6-5
Data Memory OrganizZation..........ooieiiieiiriiiiiiieeiceeeee e e eeeeeterttenes
s e s e e 6-7
LSI-11 Bus Interface .........uciiiiiiimiiiiii e e ———————————aa 6-8
Integral Modem RecCeive .......ooevvveiiiiiiiiiciieeeeeeen. e
eereee e e e —————eeeeerrntr——————— 6-11
Integral Modem TransSmit..........eeeeieiieieiiiieiereeeeee
e eee e e e e e
eeeeeeeeeee e
6-12
Example of a Typical Isolation Flow Diagram ...........cccccccciiiiiiiiiiiinninnnnnnn. 7-2
Data Link and Threshold Error Counters.........coouuuiiieriiiiiiiiiiiiieeeeeereeeicee
eeeeeee 7-8
Station Error Counters..........uuuiviiiiiiiieeieieeeeeeeecerreeieee
e e e 7-9

viil

'FIGURES (Cont)

Title

Figure No.

Full-Duplex Seven Tributary Multipoint Network.............. e

7-4

e e e e eeeee e e e 7-14
A-1

B-1

DDCMP Message Formatin Detall..'.‘.................................; ...... e s
UNIBUS and LSI-11 Address Map ......ccccceeeeennnnnn.aeieiicieeeseens ievereeeareeeenerrsnenseeenes

A-3
B-2

‘D-1

Flow Diagram Symbology.......cc...cou.......Vi e

- D-2

Modem Control (Start).............oeeevvvnnnnnee, e
— . eerinashieeesassesssssnnerensens D-5
‘Modem Control (Transmit)......... B
R S URN e ——————————————————— D-6
Modem Control (Transmit 2).................eveerreennns veeeieeeeenss ereeareserie e —————— D-7

D-3
D-4

D-5

D-6
D-7

D-8

Page

DDCMP Data Message Format............cccooeveevineennnnn, eieivatereetaeseieneseeraereserstnnsens

A-1

——— e e

ssressareeas .

Modem Control (Receive) ........cccevuennenee eeveeileienannees es D-8
Modem Control (Modem STAtUS)....cccovvuiiiiiiiiiiiiieieceeeeee et eeeeeeaeee s D-9
Modem Control (Call Timer)................e —— e
e et eeetea e —————— e, D-10
Modem Control (SHULAOWI) . . o v uei i i e
t
eevenns D-11

TABLES

Table No.

Title

DMV
11 OPLIONS....iiiiiiieiiiiiiiieeee
e
e e e e e e e e s e e e eeeeessbaanaeaeeaaeesesseesnnes
DMV 11 Option Packing List .......coiiiiiiiiiiiiiiii
e

Page

1-3
2-2

Typical Host Options of a Bell 208A Data Set (4800 b/s)
Full-Duplex OPeration .....c..ccceeeeireiiiiiiiiiiie
ettt eesree e ee s s eere e e s seeeeessnnees

2-5

—

ON U AW

Typical Tributary Options of a Bell 208
A Data Set (4800 b/s)
Full-DupleX Operation ........ccccciiiiiiiiiiiiiiiiiicieeeeeeeeeeeeeeeeeeeeeeeeee
e eeeeee e eereeeeraeesasseneaeees
2-5
DMV11 Voltage Chart .......ooveeeiiiiiiiieeeeeeeceeee
v feteeeeeerererteeeeeeeeerrnnaa.,
2-6
Device Address Selection.......ccccuvevrvieeiniiiiieiiieniieeeeeeeeieee
e et
e e 2-11
Vector Address SeleCtion........uuueeeeuuieiiiiicicii
s eeeeeereen——— 2-12

AN
L A

Cable DESCIIPLION ...uueeiiiieeeeeeeeeeeeee
e
ee ee e eee eees
eeeeeaaeas 2-14
Modem Option Jumper FUNCLIONS ........oovviiiiiiiiiiiiiee i 2-30
SELQD Bit FUNCHIONS. ...ttt
e e e e e e ee e e e
3-2
BSEL2 Bit Functions........ e eeeeeeeteteeeeiteeeeeieteeeeeereeeeiaieeeeeairreeeeeaatareeeantreeesenrees
3-4
Input Command COES ........coeeeeiiiiiiiciiieeee
e
ee
3-5
Mode Field Codes and FUnCtions..............uuuuuiiiiiiiieiiiiiiiieeeiieiece e 3-8

SEL6 Control Command FUNCLIONS........c..eouirierieriiieieieie e 3-10
Request Key Field Definitions (Control Command) ...........cooeevvvvvirriiiiniivveneenennnee. 3-14
OULPUL COACS ..o 3-21
Return Keys for Information ReSponse .........ccccceeveeiiiiiiiiiiiiiiiiiiiieee e 3-27
DiIagNOSHIC BITOT COES .....ovevieieeeeeeieeeeeeeeeeeeee ettt see e e e se e eaeaes
4-5
User-Defined TSS Parameters........ucciiieiiiiiiiiiicieeeccieeeee e 4-7
User-Defined GSS Parameters .........ceiiiiiiiiiiceiee
et 4-12
Recommended Selection Interval Timer Values .....c.cccoeoeeieiieiiiceiiiiiieiiivnnnineiennn, 4-15
Mode SWILCH SEtHINES ....ciieeiiiiiiei et
ree e e e e e e er e e e e eeeeannannes 4-23
Maintenance Command Functions BSEL2 Bits 0-3 ..ouiriiiiiiiiieieeeeieeieeeeenernenns 4-25

TABLES (Cont)

Table No.
7-1

Title

Page

DMYVI11 DIagNOStiCS..ccceiiiiiiiiiiiiieeeececiecitiieete
e
et e e e e e e s s s s esaens 7-18

7-2

DMV11 Static Logic Test Part 1 Dlagnostlc Summary eeeeeeeeeeetteer
e aaaeererrrann—., 7-18

7-3

DMV11 Static Logic Test Part 2 Diagnostic SUMmary .........ccccccovvvvveieeeeereeeenneen. 7-19

7-4
7-5
7-6
7-7
7-8
B-1
B-2
D-1

DMYV11 Static Logic Test Part 3 Diagnostic SUMmAry .......cccccceveeeeieveeiiiieennnnenen. 7-19
DMV11 Static Logic Test Part 4 Diagnostic SUMMATrY ...........ccoovvvvvvveeeeeeererinnnee. 7-20
DMV11 Static Logic Test Part 5 Diagnostic SUMmMATY ........ccoeeeviviiviiiiiiiinireinneeee. 7-21
DMV 11 Functional Diagnostic SUmMmMATY ........ccceevveeeviveeieeeeeeiieriieenneeenenee. rreeeee
e 7-23
Microdiagnostic Error Codes.............cccccc....... Uett ————————————aaaaaas 7-24
Floating CSR Address DeVICES.......ccccuuuiiiiiiiieiiiiieeieeeee
e, B-3
Floating Interrupt Vector Devices..................... cee———— et —————————aaa
e e B-3
DMYV11 Modem Control Functions........................eeeeesesasessssssesesiressreseressresssssssesses D-2

PREFACE

This manual describes in detail the installation requirements, programming considerations and tech-

niques, microcode operation, technical functions, and servicing procedures, including diagnostic support, for the DMV 11 Synchronous Controller. A variety of appendices are also provided to supplement
the above.
|

Other publications which support the DMV11 Synchronous Controller are:
e

DMVII Print Set (MP-00942)

Electronic Industries Association (EIA) Specifications

DIGITAL Data Communications Message Protocol (DDCMP) Specifications (AA-D599ATC)

CHAPTER 1
INTRODUCTION

1.1

INTRODUCTION

The multipoint DDCMP-DMV11 Intelligent Communications Synchronous Line Controller is a device
which provides efficient high-speed synchronous communications for distributed networks. The
of main CPU
DMV11 uses LSI-11 CPUs as control or tributary stations, while requiring a minimum
the
operating
and
resources. This manual provides detailed information necessary for installing

1.2

DMV11.

INTRODUCTION TO MULTIPOINT

Point-to-point configurations are practical when the message rate of the terminals is high. In many.
cases, however, the message rate of the terminals is very low even though the bit rate may be quite
high. In these cases, sharing a transmission line can significantly reduce the cost and improve the efficiency of a communications network.

Various techniques are used to share transmission lines to improve their utilization. One of these techniques is the use of multipoint lines. In multipoint operation, a single line can be shared among many
nodes. Each node is a station and has a unique address. One station in the network is always designated
as the control station while the remaining stations are designated as tributary stations. Because all stations are connected to the same line, no two tributary stations may transmit at the same time, and each
station must have a means of recognizing which messages it is meant to receive. The address field of the
message header identifies the station to receive the message. The control station governs sharing of the
line by means of polling in order to authorize transmission to the control station. In a polling operation
the tributaries are in effect asked one by one whether they have anything to transmit. To accomplish

this, the control station sends a polling message with a unique tributary address down the line. The
station which recognizes the address responds by sending data or by sending a positive response.

Tributary stations can only transmit to the control station and only in response to a polling message
from the control station. Transmission between tributaries is not allowed as all message traffic must be
routed through the control station. Control stations on the other hand may transmit to any tributary at

any time if the communicating stations are in full-duplex. In fact, multiple messages for different destinations (tributaries) can be sent serially by the control station. Each tributary station then, in turn,
examines the address and accepts only those messages it is meant to receive.

The use of communication lines can be maximized by using full-duplex capabilities at the control station to accommodate many tributary stations on a full-duplex line. In this mode, the control station
keeps the lines full by sending to one or more tributary stations, while at the same time receiving from
'

another tributary station.

1.3 DMV1l GENERAL DESCRIPTION

The DMV11 is a high-performance line controller which operates at speeds up to 56K b/s. It accom|

plishes this by doing DMA transfers.

1-1

There are three available options and they are outlined below.

The DMV11-AA consists of:
An M8053-MA microcontroller/line unit (a quad-high module with multipoint microcode);

- -An H3254 (V.35 or integral modem) module test eonnector;

An H3255 (RS-423-A/232-C) module test connector:
A BC55H cable; and

“An H325 and H3251 cable turnaround test connector.
The DMVll AB consrsts of
An M8053- MA m1croeontroller/lme unit (a quad h1gh module W1th mult1po1nt m1crocode)

An H3254 (V.35 or 1ntegral modem) module test connector
An H3255 (RS-423-A/232-C) module test connector;

‘A BCO05Z-25 cable; and
An H3250 cable turnaround test connector co

An MB064-MA microcontroller/line unit (a quad-high module with multipoint microcode);

The DMV11-AC consists of:

test connector;
or integral modem) module
~ An H3254 (V.35
- A BC5SF cable; and

H3257 and H3258 te_rminatorS.

These three options provide coverage of four different types of interfaces (see Table 1-1).
Features of the DMVll include:

Support of pornt-to-pomt and mult1p01nt operatlon

| Support for remote or local fullduplex or half—duplex conf1gurat10ns o
Support for 12 tr1butar1es and one control station in mult1pomt operat1on
Switch and program selectable operatmg mode and trlbutary address
. Support for mult1ple addressed tr1butar1es
Down-line loadlng and remote load detect capabllmes

Go/No-Go diagnostic testing by the microcode,
ST

Go/No-Go extensive error reporting,
Modem control.

1-2

Table 1-1

DMV11 Options

Option

Line Speed

Interface

DMVII-AA

EIARS232-C

(DMV11 leltathHS)v

Upto 192K b/s

-~ EIA RS-423-A

"Upto 56K b/s

DMVI11-AB

CCITT V.35

Up to 56K b/s

DMV1 l-AC'

Integral modem

356K b/s only

1.4 STANDARD APPLICATIONS
The DMV11 can be used with the integral modem as well as with EIA and CCITT applications. These
applications can be configured as either point-to-point or multipoint networks. Figure 1-1 shows a typical point-to-point application and Figure 1-2 shows a typical multipoint application. For local operations
through integral modems, stations are interconnected by twinax or triaxial cables. The integral modem
can support up to 12 drops in both half- and full-duplex modes. For remote operations, stations are

‘connected through external modems that use common carrier facilities. For specific information on installation of either of the basic DMV11 units and associated options, refer to Chapter 2.

For multipoint applications, the tributary address for each DMV11 in the network is either switch or
program assigned. In the case of switch-assigned tributary addresses, specific switches on each DMV11
define the numerical value of the address to which that DMV11 reSponds The advantage of a switchassigned tributary addressis that it provides data transfer security since the address cannot be changed
by software.
A major advantage of DMV11 multipoint networks is the ability of the main CPU at the control station
to down-line load programs to the CPU at each tributary and start those programs without manual intervention. As a result, DMV11-based multipoint networks are particularly suited for installation at
remote and generally inaccessible locations. For example, DMV11s may be used in satellites, in hard to
reach locations such as weather stations at sea, and in hazardous environments.
1.5
DMVIii1 SYSTEM OPERATION
Operation of the DMV11 communications line controller is initiated and directed by a user program
residingin the main memory. The user program consists of an apphcatlon program and a devrce driver
that serves as an interface between the DMV11 and the cpuU.
|

‘Commumcatlon between the user program and the DMVII 1S accomphshed over the LSI 11 bus
through four control and status registers (CSRs). These four 16-bit registers serve as a bidirectional
port to pass user-program commands to the DMV11, and DMV11 responses to those commands back to
the user program. Each of these registers are word and byte addressable by both the user program and
the DMV11 microcode.
NOTE
Normally only four CSRs are available to the user
program. However, in 22-bit address mode, eight
CSRs are available although only one additional 16bit register is used.
_
|

In this group of four CSRs, the first two have a fixed format and in general serve as a handshake con-

trol for user-program commands and DMV11 responses. The next two CSRs form a port for the ex-

1-3

change of command and reSponses between the user program and the DMV11. Other control fields
provide for initialization, interrupt enabling, reading and executlon of maintenance instructions, data
transfer setup, and tributary addressing.

A user program issues a command to the DMV11

by first requesting the use of the data port. When the

DMV11 grants permission to use the data port, the user program passes the command to the DMV11 in
the CSRs. The DMV11 interprets the command and performs the specified actions. If a response is
required, the DMV11 stores the appropriate response in the CSRs and then informs the user program
that a response i1s present.

Message data received or to be transmitted by the DMVI11 is written into or read from preassigned
buffers in main CPU memory. These buffers are accessed by the DMV11 through nonprocessor
requests (NPRs) to the associated bus address.

LSI-11 CPU

OPERATING

LOCAL

NODE

SYSTEM

LSI-11 CPU
OPERATING

SYSTEM

USER PROGRAM

NODE

USER PROGRAM

DEVICE DRIVER

REMOTE

LSI-11 BUS

’

DEVICE DRIVER

vy
#1

TRIB.

L #2
——
S

LSI-11 BUS

r—

.
j”-——-—-bl
MODEM |j<«—>

L e

|je—> MODEM le

USED

NOT

DMV11

Emm———

#12

DMV11

MK-2485

DMV11s Used in Point-to-Point Applications
TN

Figure 1-1

1.6 COMMAND/RESPONSE STRUCTURES
|
Since the DM V11 is basically an input/output device, it follows that the command/ response set for this
device be categorized as input commands and output responses. Input commands are commands issued
by the user program to the DMV11. Output responses are typically responses to those commands, and
are 1ssued by the DMV11 to the user program.

Some responses are unsolicited, and are used to inform the user program of protocol events and line
errors.

el
s

1.6.1
Input Commands
|
There are four types of input commands. They are listed belowin the usual order ofissuance.

Microprocessor control/malntenance
‘Mode definition,
Control,
Buffer address/character count.

1-4

1-5
T

NOILVLS
HOLIMS HO4

H3sn/WVYHD0Hd
-IS7L1 sSNg

a3y¢-1SEQS|paspUtjulodninysuoryeody
njodiLTINWN| SALVLS101S
— "
=

"~ W3LSAS

niwyado|o
0ZSTHIN

[ Walsas

3'9GON

1.6.2 Output Respo.nses

Output responses serve to inform the user program of normal or abnormal conditions concerning the
data transfer operation. There are three types of output responses:
1.

Control response,

Information response,

Buffer disposition response.

~ Figure 1-3 is a general summary of the functions performed by the DMV11 command /response structure. These commands and responses are discussed in detail in Chapter 3.

“

1.7 PROTOCOL SUPPORT
In DMV11 point-to-point and multipoint networks, all message transfers between nodes are under con-

trol of the Digital Data Communications Message Protocol (DDCMP). All aspects of DDCMP processing are handled by the DMV11 microcode. Message handling at the user-program level only involves

setting up data buffers during transmit operations and accepting data from the DMV11 during receive
operations.

There are no file structure constraints on messages transmitted or received over DMV11 networks;
however, the maximum data message length allowed is 16,383 bytes. Also, there are no restrictions as
to the type of data transmitted or received under DDCMP since all data is transmitted and received in

transparent form.

There are basically three types of DDCMP messages: the data message, the control message, and the
maintenance message.
1.7.1
Data Messages
DDCMP data messages consist of two parts: the message header and the message body. The header
consists of eight bytes of control information necessary for successful transmission of the message. In-

cluded in these eight bytes is the block-check count (BCC) for the header, the byte count of the mes‘sage body, and the tributary address. The header also contains two control bits; one that indicates resynchronization after this message, and one that controls line turnaround. The message body consists of

the message and a BCC for the message. Both BCC characters are used by the DMV11 to validate
messages as they are received.

The header is assembled by the DMV11 and transmitted with the message body to form the data message. The receiving DMV11 uses the header to verify the address and ensure that the message is received in the correct sequence. The header is also used to determine the number of bytes to transfer to
the user program. The header is discarded when the message is successfully passed to the user program.

1.7.3 Maintenance Messages
Under DDCMP, a DMV11 has two data transfer modes: the DDCMP run state, and the DDCMP
maintenance state. In the run state, a DMV11 receives and transmits data messages. In the maintenance state, a DMVI11 receives and transmits on/y maintenance messages. A maintenance message is

formatted much like a data message. It is formed by an eight-byte header followed by a variable length
message body. The content of the body is determined by the user program. Maintenance messages may
consist of: 1) operating or diagnostic programs transmitted by the control station for down-line loading
into the CPU of a specified tributary, or 2) a portion of the contents of a tributary’s CPU memory as
requested by the control station. The request of this information is also handled by a maintenance message.

1-6

1.7.2 Control Messages
|
Control messages are used to manage message traffic. They are eight-byte DDCMP messages which
are passed between control and tributary stations under sole control of the DMV11. Two examples of
control messages are acknowledge (ACK) and negative acknowledge (NAK). ACKs indicate successful
reception of messages while NAKSs indicate unsuccessful reception. Control messages in multipoint also
contain the address field to identify the tributary to which the message is sent or received from.

LSI-11

USER PROGRAM

USER PROGRAM ISSUES
MICROPROCESSOR CONTROL
COMMAND TO MASTER
CLEAR DMV11.

INITIALIZE
DMV11

NOTIFY

| FaiLure

USER
,
PROGRAM |

USER
PROG. VERIFIES
«

INTERNAL

DIAG.

" SUCCESSFUL,
DMV11 RUNNING

SET MODE TO HALF
OF FULL DUPLEX,

USER PROGRAM ISSUES

ESTABLISH STATION
.

AS POINT-TO-POINT

MODE DEFINITION

NODE OR MULTIPOINT |

COMMAND

CONTROL OR TRIBUTARY

STATION

ESTABLISH TRIBUTARIES,
SET USER DEFINED

USER PROGRAM ISSUES

PARAMETERS, INITIATE

CONTROL COMMANDS

PROTOCOL STARTUP

v
ASSIGN TRANSMIT
AND.

RECEIVE BUFFERS
TRANSMIT

USER PROGRAM ISSUES
BUFFER ADDRESS /
CHARACTER COUNT
COMMANDS
RECEIVE

{ |

THE DMV11 PERFORMS THE FOLLOWING

THE DMV11 PERFORMS THE FOLLOWING
TRANSMIT MESSAGE FUNCTIONS:

'RECEIVE MESSAGE FUNCTIONS:

1. PROCESS RECEIVE HEADERS

1. CREATE DDCMP MESSAGE

HEADERS

2. CHECK CRC's

3. ACKNOWLEDGE (ACK) PROPERLY
RECEIVED MESSAGES

2. GENERATE MESSAGE AND
HEADER CRCs
3. TRANSMIT MESSAGES

4. NEGATIVE ACKNOWLEDGE (NAK)

ERRONEOUSLY RECEIVED MESSAGES

i
THE DMV11 PERFORMS THE FOLLOWING

MESSAGE TRAFFIC CONTROL FUNCTIONS: | DOMV171 ISSUES BUFFER
1. MESSAGE SEQUENCING

DISPOSITION RESPONSES,

2. ERROR RECORDING AND REPORTING | CONTROL RESPONSES AND

3. PROTOCOL SUPPORT
4. LINK MANAGEMENT

INFORMATION RESPONSES

MK-2651

Figure 1-3 General Summary of DMV11 Command/Response Structure
1-7

1.8 GENERAL SPECIFICATIONS
|
Environmental, electrical, and performance specifications for all DMV11 configurations are listed in
Sections 1.8.1 through 1.8.3.
1.8.1
Environmental Specifications
|
The DMV11 is designed to operate in a class C environment as specified by DEC STD 102 (extended).

— Operating Temperature

5°C (41°F) to 50°C (122°F»)

— Relative Humidity

10% to 90% with a maximum wet bulb temperature of
28°C (82°F) and a minimum dew point of 2°C (36°F)."

Option

Voltage

DMVI11-AA,AB

+5V@34A
A
+12V@0.380

DMVI11-AC

+5V@335A
+12V @ 0.260 A

A —12V @ 250 mA required by the level conversion logic for both versions is generated off the +12 V

by a switching inverter.

1.8.3 Performance Specifications
Performance parameters are as follows:
— Operating mode

— Data format

Full- or half-duplex

Synchronous DDCMP

— Datarates

Up to 56K b/s

— Tributaries supported

Upto12

DMVI1ls may be connected
to DMP11s/DMV11s, DMR11s, DMCl11s, and any other synchronous
controller running DDCMP protocol.

1.8.2 Electrical Specifications
The DMV11 requires the following voltages from the LSI-11 bus for proper operation.

CHAPTER 2
INSTALLATION

2.1

INTRODUCTION

This chapter provides all the 1nformat10n necessary for a successful installation and subsequent check-

out of the DMV11. Included are instructions for unpackmg and inspection, pre1nstallat1on installation,
and verification of operation.
-

2.2 UNPACKING AND INSPECTION
The DMV11 is packaged according to commercial packing practlces When unpacking, remove all
packing material and check the equipment against the packing list (Table 2-1 contains a list of supplied
items for each conflguratron) Inspect all parts and carefully inspect the module for cracks, loose components, and separations in the etched paths. Report damages or shortages to the shipper and notify the

a\\.

DIGITAL representative.

2.3 INSTALLATION CONSIDERATIONS
Installation of the DMV11 mlcrocontroller/ line unit subsystem should be donein three phases:

Phase I — Preinstallation considerations
Verify system requirements, system placement, and configuration requirements.
Network topology chart
For mult1pomt networks it is absolutely necessary to know the conf1gurat1on of the DMV11
(thatis; control, tributary, HDX, FDX, and so on) locations of tributaries (w/address), and
wherein network they are connected (control Trib 187 Trib 98, Trib 208) or else troubleshooting W1ll be extremely dlfflClllt
|
|
-

Phase II — M'icrocontrOHer/ line unit installation
Configure, install, and verify the mlcrocontroller/hne unit module via the appropriate diagnostics.
|
|

Phase Il - DMV L syst

em testing

Verlfy the DMVll mlcroprocessor subsystem operatlon with the functlonal dlagnostlcs and
system exerciser programs.

2-1

Table 2-1

Option

DMV11 Option Packing List

Parts List

Description

‘DMVI11-AA

RS-232-C/RS-423-A interface containing:
MB053-MA
BC55H
H3254, H3255
H3251, H325
EK-DMV11-UG
MP-00942
ZJ328-RB

DMVI11-AB

Basic remote DMV11 unit
EIA RS-232-C/RS-423-A panel assembly
Module test connectors
Cable turnaround test connector
DMVII User’s Guide
Field Maintenance Print Set
LIB kit

"
|

CCITT V.35 interface containing:
M8053-MA
BC05Z-25
H3250

Basic remote DMV11 unit
CCITT V.35 cable
Cable turnaround test connector
Module test connectors

EK-DMV11-UG

DMV11 User’s Guide

H3254, H3255
MP-00942
ZJ328-RB

DMVI11-AC

Field Maintenance Print Set
LIB kit
Integral modem interface containing:

M8064-MA

Basic local DMV11 unit

BCS55F-10

Integral modem cable

H3254
H3257/H3258
EK-DMV11-UG
- MP-00942
ZJ328-RB

2.4

(

Module test connectorBCS55A terminators
DMV11 User’s Guide
Field Maintenance Print Set
- LIB kit

PREINSTALLATION CONSIDERATIONS

Table 2-1 and the following should be considered prior to ordering a DMV11 communications interface

to ensure that the system can accept the DMV11 and that it can be installed correctly. The steps should
also be verified at installation time.
|

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

Show the actual physical location of the cable trough and indicate any equipment which might cause interference such as

Cable routing —

an X-ray room.
Machine type —

Indicate whether the CPU is a PDP-11/23, PDP-11/70, PDP11/34, VAX-11/780, and so forth. (The network could consist
of a mixture of DMP11s and DMV11s).

Type of station —

Indicate if the station is a control or tributary station.

Physical address —

DDCMP address cah range from 1-255.
2-2

Indicate by room number or other appropriate means, the ac-

Location —

tual physical location of the equipment.

The name given to the tributary if applicable.

Node name —

Operating system and
version —

‘The name of the software operating system such as RSX-11M

DECnet version — |

| D‘ECnettsoftware'version such as DECnet-11M V3.0.

V3.2,

Transmit and re_ceive —

Show transmit and receive lines. Depict end nodes and show

termination. If a patch panel is used, indicate the line numbers
between patch panels.
NOTE

is
The use of patch panels and numbering of the lines

;

recommended.

]

ELEVATOR SHAFT

RX
| TX
DELTA

11/70
DMP11

TRIB 3

RSTS XX.X

SUSPENDED CEILING

DE - XX.X

ROOM 515

[

RX | TM |

RX
| TX

BETA

GAMA

11/34

11/23

DMP 11

DMV 11

TRIB 1

TRIB 2
RSX 11M

RSX11 M

DM - XX.X

~ ROOM 430

1/~

ROOM 412
[

}

—

]

| TX
RX
ALPHA
VAX 11/780
DMP 11

CONTROL
VMS XX.X

DV - XX.X

ROOM 111
MK-2502

Figure 2-1

Local Network Topology

MERRIMACK
MK1-1/K37

VAX 11/780
VAX VMS XX.X
DV-XX.X

*NOTE
1
*NOTE
3
CONTROL

*NOTE
1
PATCH PAN ELS ARE

NASHUA

RECOMMENDED BUT MAY

NOT ALWAYS BE USED IFTHEY
ARE USED, THEIR PHYSICAL

4800

LOCATION SHOULD BE

I NASHUA

4800

IEXCHANGE

INDICATED.

*NOTE
2

11/60
RSTS XX.X

DE-XX.X

*NOTE1
*NOTE
3
TRIB 1

MODEM - 208A
CONTROL STATION OPTIONS

SEE TABLE 2-2

*NOTE
3

TEWKSBURY
TTMW

MODEM - 208A
TRIBUTARY OPTIONS

LOWELL
EXCHANGE

SEE TABLE
2-3

VAX 11/750
VAX VMS XX.X
DV-XX.X

*NOTE1
*NOTE
3

4806 | mavnaro | 4800

TRIB 2

I EXCHANGE l

MARLBORO
.
MR

4800

11/23
RSX-11M

MARLBORO

V.3.2
DM XX.X

EXCHANGE

*®NOTE
1
*NOTE
2

[ ]

MAYNARD

PK3

11/23
RSX-11M

TRIB 4

V.3.2

*NOTE
1
TRIB 3

MK-2848

Figure 2-2

Remote Network Topology

2-4

Table 2-2

Typical Host Options of a Bell 208A Data Set (4800 b /s)
F ull-Duplex Operation

- DEC Recommended Settings

Data Set Options
Transmitter timing

Data set (internal)

Carrier control

Continuous

Request-to-send operation in
continuous carrier mode

Continuous (CB constantly
ON)

Not proVided

One second holdover at
receiver on line dropouts

Not used— NSis strapped OFF
‘within the data set

New sync-option to squelch
receiver clock

Data set ready lead option

CC is ON when the AL button (only)
is depressed

for analog loopback testing

by data terminal
AB connected to AA

Grounding option

Table 2-3 Typical Tributary Options of a Bell 208A Data
Set (4800 b/s)
Full-Duplex Operation

Data Set Options

DEC Recommended Settmgs

Transmitter timing

‘Data set (mternal)

Carrier control

Switched (48.5 ms CA-CB delay)

Request-to-send operation in

Continuous (CB constantly ON) N/A
Switched (8 ms
.5 CA-CB delay)

continuous carrier mode

N/A

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 locpback testing
by data terminal

- CC1s ON when the AL button (only)

Grounding optioh_

AB connected to AA

is depressed

2-5

2.4.1

Device Placement

The DMV11 can be installedin any LSI 11 bus-compatlble backplane such as H9276. On systems that
contain many high-speed direct memory access (DMA) devices, thereis a probability of adverse bus

latency. To help prevent agamst this occurrence, the DMV11 should be placed physwally close to the

processor. As a result, this glves the DMV11 a hlgh DMA priority.

2.4.2

System Requirements

LSI-11 bus loading
The M8053-MA or M8064-MA present two ac loads and one dc load to the LSI-11 bus.

Power requirements

Check the power supply before and after installing the microcontroller/line unit to ensure
against overloading. Power requirements are listed in Table 2-4.
e

Interrupt priority

The interrupt priority is preset to level four.
e

Device address assignment
The DMV11 address residesin the floatmg address space of the LSI-11 bus addresses. The
ranking a331gnment of the DMV11 for bus addressis 24.

The selection of the device address is accomplished by switch
controller/line unit module. Refer to Figures 2-3 and 2-4.
e

packs on the micro-

Device vector address assignment

The DMV11 vectors reside in the floating vector space of the LSI-11 bus addresses. The
ranking assignment of the DMV11 for vector assignments is 46.

The selection of the vector address is

accomplished by a switchpack on the micro-

controller/line unit module. Refer to Figures 2-3 and 2-4.

Table 2-4

DMYV11 Voltage Chart

Module

Voltage Rating

Maximum

Minimum

Backplane

M8053-MA

+5V@34A

+35.25

+35.0

AA?2

+12V@0.380 A

+12.60

+11.40

AD2

+35.25
+12.60

+35.0
+11.40

AA?2
AD?2

M8064-MA
.

+5V@335A
+12V @ 0.260 A

Voltage

[

M8053

E“54~,

E53 |

]

|
MK-2698

Figure 2-3

M8053 Switch Locations

FelE]

M8064

ES9

ES8

]

[
MK-2521

Figure 2-4

M&8064 Switch Locations

2.5

INSTALLATION

When installing the DMVI11 in the LSI-11 bus- compatlble backplane LSI-11 conflgurlng rules must
be followed.

Proceed with the installation as follows by performing the followmg on the slot that wrll contain the
DMVll
1.

© Verify that the backplane voltagesv are_ within the tolerances specified in Table 2-4.
Turn system power OFF and perform resistance checks on the backplane voltage sources to

ground. This ensures that no short circuits exist. Refer to Table 2-4 for backplane pin assignments.

Configure _the"'correct"device address using switchpack settings from Table 2-5.

‘Configure the correct vector address using switchpack settings from Table 2-6.
Verify that the switch selectable features of the DMV11 are configured for the station being
installed. See Figure 2-5.

Insert the approprlate module test connector into the correct microcontroller/line unit con-

nector as specifiedin Table 2-7. Be sure to insert the test connector with “SIDE 1” (etched
on the test connector) visible from the component 51de of the module Refer to Flgure 2-6 and
2-7.

Schematics and outline drawings of each test connector used with the DMV11 are‘p’mvided
in Figure 2-8.

Turn systern power ON.

Load and execute the DMV11 static dlagnostlcs Five errorfree passes of each part is the

minimum for successful operation.

(C)VDMA** — DMV11 static logic test part 1
- (C)VDMB** — DMV11 static logic test part 2

(C)VDMC**— DMV11

static logic test part 3

(C)VDMD**— DMV11 static logic test part 4

- (C)VDME**
— DMVll static logic test part 5

Remove the module turnaround test connector and connect the appropriate cable (see Table
2-7 and Figure 2-9) to the proper Berg connector for the DMV11 option selected. Refer to
Table 2-7 for detailed information on cable requlrements and to Frgures 2 10 through 2 14
for system cabhng confrguratrons

NOTE

When installing panel cables BC55F or BCS5H, it is
important that the panel be properly mounted to the
rear-mounting bulkhead to ensure adequate grounding.

2-9

When connecting the BC55H connector panel, verify that the approprrate modem Jumpers
j
on
the panel are properly configured for the option selected. Table 2-8 lists each of these options
and required jumper configurations.

Integral modem options require that a 75 ohms terminator be connected to each receive line

(BCSSF panel) at each end of a full-duplex and a half-duplex network. These terminators are
availablein 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 BC55F panel. Refer to Fig-

~ure 2-10 for DMVll remote cabling and to Figure 2-11 for DMV11 to DMV11 local cabling.
10.

Insert the appropriate cable turnaround test connector in the end of the cable. Refer to Table
2-7 for the specific test connector. Load and execute the static diagnostics specified in Step 8

using'the external maintenance mode selected to verify the module and cable. Upon obtaining a minimum of five error-free passes, proceed to the DMV11 system test procedures, Section 2.6. Figure 2-8 illustrates the various test connectors usedin the DMV11.

2.6

DMV11 SYSTEM TESTING

The final step in the installation of a DMV11 subsystem is to exercise the DMV11 as: 1) a unit on the
LSI-11 bus; and 2) a link in a communications network.
2.6.1

Functional Dlagnostlc Testmg

Ensure that the specific cable turnaround test connector for the selected DMV11 option is still installed

at the end of the cable. Load and execute the DMV11 functional diagnostics with the external mode
selected. Upon obtaining a minimum of five error-free end passes, proceed to Section 2.6.2.
2.6.2
DEC/X11 System Exerciser
The DEC/X11 system exerciser for the DMV11 can be run in two different operating modes, internal
and external. The internal mode selects faster LSI-11 bus activity. The external mode requires that the
specific modem test connector be installed at the end of the cable. This is the preferred mode of operation. There are two DEC/X11 modules for the DMV11; DMD* and DME*.

2.6.3 Final Cable Connections
The final step in the installation process is to return the DMV11 to its normal cable connections, either
to the appropriate modem or to the distribution panel. The DMV11 system cabling diagramsin Figures
2-10 through 2-14 have been included to help show overall cabling for the various DMV11-XX options.

References to specified locations of the various test connectors during diagnostic testlng 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 (DCLT) be exercised.
2 6.4 DMVII Link Testmg
|
The DMVI1I1 can be exercrsed over a communications lmkby the data communications link test
(DCLT). Itis 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 system that con- |
tains a DMV11.

2-10

Table 2-5

Device Address Selection
LSB.

MSB

15 | 14

12 I 1

9 ]8

s M8053 E63 M8064 E 58

M8053

\.
-

- EB4

M8064 |

o
|

ES9

- SWITCH
NUMBER

3 I 2

Ls1

- 83

760020

760060

760100

| oN

765.'.200

768500
760400

ON
ON
ON

760500

760600
ON

760700
761000

762000

DEVICE
ADDRESS

760040

763000

764000

N
NOTE: SWITCH ON RESPONDS TO LOGICAL ONE ON THE BUS

MK-2584

2-11

Table 2-6
MSB

HEBEER

0 l_o

Vector Address Selection

EBEEREDBE

ofo |

MEOES E54

M8064 E59

SWITCH

LSB

s fs|sfafz2]|1]o

o I ol

~ NUMBER

fi\[uo

S8 1 S7 | S6 ] S5 ] S4 ] sS3

on | on

ON | ON

VECTOR

ADDRESS

“__—fi

300

310

320
330
340
350
360
370
400

500

ON | ON

600

ON | ON | ON

700

NOTE: SWITCH ON PRODUCES LOGICAL ONE ON BUS
MK-2663

2-12

DDCMP ADDRESS REGISTER
TRIBUTARY/PASSWORD

E 119 (M8064)

ZERO = "ON"

E 113 (M8053)

—

ONE IS SET

'MODE WHEN SWITCH
-

—

REMOTE

AUTO

UNIT

LOAD

ANSWER

NUMBER

DETECT

MODE
ENABLE

FOR

ENABLE (OV/ER

BOOTING

POW

BOOT
|

ENABLE

~ E 107 (M8064)

ZERO = "ON"

E 101 (M8053)

SWITCH SETTING FOR THE MODE OF OPERATION.

ON | ON

| ON |

HDX PT TO PT DMC COMPATIBLE

| OFF |

FDX PT TO PT DMC COMPATIBLE

| ON

ON | OFF | ON
|

HDX POINT TO POINT

ON | OFF | OFF |

FDX POINT TO POINT

‘OFF|

| ON |

HDX CONTROL STATION

OFF|

OFF |

FDX CONTROL STATION

HDX TRIBUTARY STATION

'OFF | OFF | ON |

OFF | OFF | OFF |

10
«

FDX TRIBUTARY STATION

HIGH SPEED SWITCH

MUST BE SET
FOR INTEGRAL
MODEM OR WHEN

_§<

“ON" = V.35
“OFF" = EIA
M8053

RUNNING ABOVE 19.2KB

HIGH
SPEED

* UNUSED
ON

E 107 (M8064)

M8064

E 101 (M8053)

OFF = “LOGIC ONE"

ZERO = “ON"
MK-2483

Figure 2-5

DMYV11 Switch Selectable Features

2-13

Table 2-7

Cable Description

Module

Test

Interface

Cable

Connector

RS-232-C

BC55H
Refer to
Figure 2-9
View A

J2
(M8053)

H325

Description

A cable with a 40-pin
‘Berg connector at one

“end. The other end
has a panel that in- cludes two different
cinch connectors, J1
and J2. Connector J2
is used for RS-232-C
to connect to the
modem with external
cable BCO5D-25. The
panel is mounted to
a rear-mounted bulkhead to ensure proper
grounding and ease of
access to external
cable connections.

H325

BC05D-25
Refer to
Figure 2-9
View B

A 7.6 m (25 feet)
external cable that
connects to J2 of the

BC55C panel and an
RS-232-C modem.

RS-423-A

BC55H-03
Refer to
Figure 2-9
View A

J2
- (MRB053)

BC55D-33
Refer to
Figure 2-9
View C

2-14

H3251

Same cable as used
for RS-232-C except
that panel connector
J1 is used with external cable BC55D-33
for connection to the
modem. The panel is
mounted to a rearmounted bulkhead to
ensure proper grounding and ease of access for external
- cable connections.

H3251

A 10.1 m (33 feet)
cable that connects
toJ1 of the BC55H
“panel and an RS-449
modem.

Table 2-7

Interface
V.35

Cable
BC05Z-25
Refer to
Figure 2-9
View D

Cable Description (Cont)
Test
Connector

~ Module
Connector

| H32s0

J1
(M8053)

Description

A7.6m (25 feet)

- ~modem cable with a

40-pin Berg connector

Integral
Modem

BCS5F
Refer to
Figure 2-9
View F

Panel
switch to
HDX position

J1
(M8064)

at one end that connects to J1 of the
M8053. A 34-pin DataPhone DIGITAL Service
(DDS) connector is
installed at the
other end and connects to the modem.

A 0.9 m (3 feet)
cable with a 40-pin
Berg connector at one

end that plugs into
J1 of the M8064. The
panel assembly is installed at the rear“mounted bulkhead for
ease of external connections and to ensure proper grounding.
- Appropriate terminator connectors H3257
or H3258 must be
used. See Figures 2-9

and View F.
Integral
Modem

BC55N-98
Refer to
Figure 2-9
View E

Local link
BCS55F panel

None

A 29.9 m (98 feet)
external twinax cable
used to interconnect
a DMP11 ora DMRI11 to
a DMV11 system for a
selected data rate of

56K b/s.
BC55M-98
Refer to
Figure 2-9
View E

None

2-15

A 29.9 m (98 feet)
external triaxial
cable used for the
same purpose as the
BC55N, but for data
rates above 56K b/s.
The DMV11-AC supports
data rates of 56K
b/s.

CONNECT CABLE BCO5Z FOR V.35 INTERFACE

'H3254
TEST CONNECTOR

H3255
TEST CONNECTOR

FOR RS-232-C OR RS-423-A /

'CONNECT CABLE BCB5H
INTERFACE.

,'/

M80563

T
MK-2508

| Figure 2-6

Test Connector Insertion for the M8053

2-16

CONNECT BC55F

FOR INTEGRAL MODEM

H3254
TEST CONNECTOR
J1

M8064

MK-2699

Figure 2-7 Test Connector Insertion for the M8064

2-17

REC CLK
+

TX CLK +

TXCLK -

TX DATA
-

REC CLK -

*—

R DATA +

RTS

CTS
REC RDY
TER RDY

DATA MODE
TXINT
REC INT

TXINT
REC INT

UL
L

R DATA

*—

_J_[!l

mJ_

TX DATA +

NULL CLK -

NERRRERREE

NULLCLK +

SIDE 1
H3254

VIEW A

H3254 MODULE TEST CONNECTOR (J1 ON M8053/M8064)

MK-2645

(0 [J]o
r

AUX CLK + ——ns

0% 3P
oc
or 9 or
OF
0I5,

O ON POM
OR
OTOUOS

TX CLK DIFF +

AUX CLK -

‘W
-

TX CLK DIFF -

RX CLK DIFF + __.____X
d

RX CLK DIFF = —e——}—@

TX DATA DIFF +

OZOXOYOW

-~ RX DATA DIFF +

TX DATA DIFF -

RX DATA DIFF —
RTS

N
VIEWB
H3250

CTS

-4

-&-

S
S

]

|
J

>
D
—4———-—-—-%

CDET

DTR

% :

DSR

MK-2123

Figure 2-8

DMVI11 Test Connectors (Sheet 1 of 4)

2-18

SEND COMMON
REC COMMON

TER
IN SER

TER RDY

O
O

INCOMING CALL
+

DATA MODE +

SEND DATA +
REC DATA

SEND DATA -

REC DATA —

NULL CLK +

SEND TIMING +

REC TIMING
+
NULL CLK —

SEND TIMING
—
REC TIMING

SEL SIGNAL RATE

SIGNAL RATE IND
¥ SEC SEND DATA
* SEC REC DATA

* SEC RTS
* SECCTS

LOCAL LOOP
TEST MODE
BY

STAND
BY IND
RTS -

cTS —
REC RDY

—

DATA MODE -

—i—
D

TER RDY

-*fi

% NOT REQUIRED FOR DMV11

1
SIDE

EjERERERRRE

SIGNAL QUALITY

SEL STAND

U EBERr P EF BEEE

NEWSIGNAL

'RECRDY
+
-

EREERERRRRRERR

i Ptr-t)

CTH2

TH1
H3255

VIEWC

H3255 MODULE TEST CONNECTOR (J2 ON M8053)

]
MK-2644

Figure 2-8

.DMV11 Test Connectors (Sheet 2 of 4)

2-19

SEND TIMING -

REQ DATA-

REQ TO SEND
REQ TIMING
-

SEND DATA
-

RECEIVE COMMON

—

SIGNAL GROUND

—

4 [ Y 7= o#md\ubm&n [ TN .mfim.)

TEST MODE

d§+3¢302980305é

TERMINAL TIMING +

SIGNAL RATE SEL

INCOMING CALL

REMOTE LOOP

¢ S On
©® w®

RECEIVER READY +

TERMINAL READY +

DATA MODE +

-—

LOCAL LOOP

[
[|

CLEAR TO SEND +

REQ TIMING +

REQ TO SEND +

SEND TIMING +
REQ DATA +

SEND DATA +

SIGNAL RATE INDICATION

SHIELD GROUND

CLEAR TO SEND
-

TERMINAL IN SERVICE

- DATA MODE
-

H3251

TERMINAL READY
-

RECEIVER READY
SELECT STANDBY
—

SIGNAL QUALITY

CVIEWD |

NEW SIGNAL
TERMINAL TIMING
-

STANDBY INDICATION
SEND COMMON

H3251 CABLE TEST CONNNECTOR

MK-2746

Figure 2-8

DMV11 Test Connectors (Sheet 3 of 4)

2-20

H325

SCT
SCR
~

memmen
=
11

SECXMIT

SEC RCV

12 l

CUT TO TEST
| SYNC
NEW
10—-—-—0

SIDE

XMIT DATA

RCV

o) AN
e

NEW SYNC

-E.

DATA SET RDY =t

‘ ° ¢ ¢ o9 6 ®» © 9 6 ©¢ @ @ o

m——_—

‘

S\ CUT TO TEST*

.J NEW SYNC

\‘

H326

E
VIEW

H325 CABLE TEST CONNECTOR (BCEEC AND BCOED)
MK-2124

Figure 2-8 DMV11 Test Connectors (Sheet 4 of 4)

2-21

VIEW A

RS-423-A

RS-232-C

'=W15-

W1
W2
W3
W4 —

— W17
/W18
— W19

W7 43
W8 —

W20

cs [
ce 1

—

—— —

—

,...l.‘.’....'

——— G

]

s e

c3[_]
c2[]

———
—
—
— —
—
P

—

W5 /
W6 —

W16

I............

———— y

c—

W10

=
W11 =
W12 &

W13
=
W14

R1
W21

BC55H-3 (RS-232-C/RS-423-A) INTERFACE PANEL CABLE
MK-2656

VIEW B

25 PIN

- BCO5D-25 (RS-232-C INTERFACE) MODEM CABLE

CINCH
MK-2745

VIEW C

§a_

37 PIN ‘1
BC55D-33 (RS-422-A/RS423-A INTERFACE) MODEM CABLE

Figure 2-9

DMV11 Cable Drawings (Sheet 1 of 3)

2-22

CINCH

MK-2743

Figure 2-9

)©

o‘

—

\\ -~

H
0O00®@e00e

OOee®O®O®
0O0O0Oe®®OG®

DMYVI11 Cable Drawings (Sheet 2 of 3)

2-23

et ot B

oet e e

N D e

BCS5N TWINAX CABLE

BC55M TRIAX CABLE
MK-2742

060 @
R L

Puy pu pem

9 6

ey 4

e ¢ .9 O ¢

Py puy ey gmy

gy pem

L R

e SRR FeA JER

H——TFNNHHT:TTHTTHTT—._E

BCO5Z-25 (V.35 INTERFACE) MODEM CABLE
Mo it it D hor O

3 X
X XXX

VIEW D

MK-2744

VIEW F

FEMALE CONNECTORS

[RECEIVE{®) TRANSMIT
IN_

BCS55F (INTEGRAL MODEM) PANEL CABLE

FDX

MALE CONNECTORS

CONNECTOR PANEL
(FRONT VIEW)
MK-2646

—
SE
—
X

a—

]

re—

H3258
TERMINATOR

TERMINATOR

MK-2244

3 of 3)
" Figure 229 DMVI1I Cable Drawings (Sheet

2-24

RS-232-C/423-A

H325 (RS-232-C)

S J1

MODEM

BC55H

J2 &

M8053
|

_J1

BC55D-33

' H3251 (RS-423 A‘

N\ J2 TEST
i

H3251 (RS-423-A)

CONNECTOR
H3255

V.35 INTERFACE
—

T
J1 TEST

—

DDS

- MODEM

BCO5Z-25

" CONNECTOR

CABLE TEST

MK-2525

vFigure 2-10 DMV11 Remote System Cablivng Diagram

2-25

FULL-DUPLEX

INSTALL 750

TERMINATOR

CONNECTOR (H3257)

BC55F

MB80B4

HDX/FDX SWITCH

MUST BE IN THE FDX

| m—vj

\(down) POSITION ON

M8064
J1

BC55F

INSTALL 75Q

TERMINATOR

% FOR 56K bps USE CABLE BC55N-98
-

CONNECTOR

HALF-DUPLEX

INSTALL 75

f\\
)
J1

TERMINATOR

BC55F

CONNECTOR (H3257)

HDX/FDX SWITCH
|

MUST BE IN THE
HDX (up) POSITION

—

ON BOTH PANELS

- M8064

J1
-

" BCB5F

INSTALL 75%Q

TERMINATOR

% FOR 56K bps USE CABLE BC55N-98

CONNECTOR
(H3258)
MK-2492

'Figure 2-11

DMV11 to DMVI11 Integral (Local) Modem Cabling Diagram (Point-to-Point)

2-26

~ HDX NETWORK

CONTROL

/”OHDXO.:- :
I
TERM
' (H3257) f'-"OFDXO” \//_,51

STATION

—

TERMINATOR

\éSEF

v v—
v

XMIT]

ROV

e = OO

-O O

R\O 4
- -[O

\éSEF
|

|
|

:_ VYRC»’V XMT]

O 1

r ==OmXQ

TERMINATOR
"~ (Ha2s8)

XMIT]

OO |
@

TRIBUTARY

\{CSEF
|

I
RCY

TRIBUTARY

BC55F
MK-2804

Figure 2-12

Half-Duplex Multipoint Network (Control Station End Node)

2-27

FDX NETWORK

TERMINATOR
(H3267)

CONTROL
STATION

l—-—

BCB5F

TRIBUTARY

BC55F

I
|
i

RCV

XMIT

'r—V‘V‘V_’

SO
k| J1
F()’:DXC}&-///E
|

TRIBUTARY

BCB5F

i
|

NOTES:

SOLID LINE REPRESENTS

CONTROL STATION TRANSMIT TO

RCV

DASH LINE REPRESENTS CONTROL

STATION RECEIVE FROM TRIBUTARY

TRANSMIT.

XMIT]l

HOX(C

(:lfo

H3258
\/

)=-H

v v
v
TRIBUTARY

TRIBUTARY RECEIVE.

H3257 |

TERMINATORS

BC55F

BOTH ENDS OF THE TRANSMIT LINE FROM THE TRIBUTARIES NEED TERMINATION
IN ADDITION TO THE ONE TRANSMIT LINE FROM THE CONTROL STATION
MK-2506

Figure 2-13

Full-Duplex Multipoint Network (Control Station End Node)

2-28

armmn

FDX NETWORK
RCV

XMIT

TERMINATORS
(H3257)

TRIBUTARY

BCB5F

'O%XO'

con

RCV.
|

(

DASH LINE REPRESENTS CONTROL

TRANSMIT.

~ CONTROL

STATlON‘ -

BCb5F

)

|
TRIBUTARY

FDXQM

' TRIBUTARY RECEIVE.

xmiT]
@

,
NOTES:
SOLID LINE REPRESENTS

CONTROL STATION TRANSMIT
TO

\/fl

XmIT

@FDX@

(

[RCV

BCB5F

'
n

H3257

—-OHoxO-|=
@'

TRIBUTARY
J1

b
|

PR
\/
H3258
\/
BC55F |

A
|

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 OF THE TRANSMIT LINE FROM
THE TRIBUTARIES

MK-2505

Figure 2-14

Full-Duplex Multipoint Network (Control Station Inner Node)

2-29

Table 2-8

Modem Option Jumper Functions
FUTURE
D
X.21BIS

///:;—-xzoam
S

/8/8/~

/& \o"'?q?qc’chl‘)” S/E/N /) ¥/

&)/

o 0)‘,\7

SSRG
ESE S
)L

/e/§/ &/ &/

/) F/)F/

| IN

12
12

W10 | IN

IN T IN

IN [ IN

IN T IN

W12

W13

W15

| W16 | IN

W17

W19 | IN

W20

W21

IN ] IN]

IN]

F
IN | IN

N
N

| SR

[sa

[110

126

[ sF

[112

| 111

1IN

INTIN | IN

] IN|

INT

NN | IN|

SBB|SRD|[119

1IN

INTIN]

IN]

| IN]

SBA ] SSD]

| IN

IN|

IN]IN | IN|

SCF]|SRR|

IN | IN | IN

T INT

INTIN]IN

[IN|

INTIN]

IN T IN T IN

118
122

| 140

| 105

| IN]

S/ E/E

S/ E/E/ T

|-IN | IN

cf N

| N
| IN

INT

INTIN

T IN

T INTIN

JIN

T IN

JIN
IN

INTIN

W14

DB | ST | 114

T INTIN T IN T IN

| RT

| 115

| N

| 141

INTINTIN]INTIN

| IN|

SCA|SRS [ 121

IN
N

CA
DD

NOT NORMALLY INSTALLED

N
IN|

|TT

| 113

(117

5SS

1116

INJIN|

INT

IN[IN]

TM | 142

IN T IN | IN

IN | IN

IN]

N ] IN]|

SCB|Scs

[121

MAKE BUSY

[INTINT

IN]

INTIN | INTINTIN

101

| SD

703

|RD

| 104

[ Cs

| 106

| oM [ 107

“AB

| SG

|102

| RR

| 109

| TR

| 108

|ic

[125
MK-2725

2-30

CHAPTER 3
o
B
COMMAND AND RESPONSE STRUCTURES

3.1

INTRODUCTION

This chapter defines DMV11 command and response formats in all necessary detail, and describes all
programming sequences relevant to DMV11 operation in the network environment. The CSR command
structure and format of input commands and output responses, as well as data port descriptions, are
described in detail. Discussions include special programming techniques, user access to maintenance
mode, and user interpretation of status/error reporting.
3.2

COMMAND STRUCTURE

The DMV11 command set is structured into two categories; input commands and output responses.
Brief overviews of input commands and output responses, including command codes and the handshaking requirements, are provided in this section.

Transfer of control and status information between the main CPU resident user program and the

DMV11 is accomplished through four 16-bit control and status registers (CSRs). Input commands are
issued to the DMV11 by the user program, and output responses are issued to the user program by the
DMVI11.

NOTE

Normally only four CSRs are used, but in the 22-bit
address mode, eight CSRs are available.

|
3.2.1 Control and Status Registers
byte and
both
are
registers
These
information.
status
and
control
Four 16-bit CSRs are used to transfer
page
I/O
the
in
space
address
floating
the
in
addresses
assigned
are
bytes
word addressable. The eight
as follows:

16XXX0, 16XXX1, 16XXX2, 16XXX3, 16XXX4, 16XXX5, 16XXX6, and 16XXX7.

For discussidn, these byte addresses are designated byte select 0 through 7 (BSELO through BSEL7).

BSEL10 and BSEL11 are only used in 22-bit address mode. BSEL12 through BSEL17 are not used by

the user/DMV11-command structure and are not referred to in this document.

The four word addresses are the even numbered locations and are designated select 0, 2, 4, and 6
(SELO, SEL2, SEL4, and SEL6). The CSR addresses are assigned to the floating address space. The
floating address ranking for the DMV11 is 24 (See Appendix B). The relationship between the symbolic byte and word addresses for DMV11 CSRs, and the actual CSR layout, is shown in Figure 3-1.
Figure 3-2 illustrates the fields in CSR bytes BSELO, BSEL2, and BSEL3 that comprise the fixed
format portion of both user-program commands and DMV11 responses. This fixed format portion serves to identify the command/response type, address of the tributary that the command/response applies
to, and coordinate ownership of the CSRs between the DMV11 and the user program. Detailed bit
descriptions of SELO and BSEL2 are provided in Tables 3-1 and 3-2 respectively.

The four bytes comprising SEL4 and SEL6 contain the fields pertinent to each user-program command
11 response. Detailed descriptions of the SEL4 and SEL6 fields are presented in Sections 3.3
and DMV

through 3.4.

8 7

1
-

BSELT

BSELO

SELO

BSEL3

BSEL2

SEL2

BSELS

BSEL4

SEL4

BSEL7

BSEL

SEL6

BSEL10

| SEL10°

BSEL11

*SEL10
IS ONLY USED IN 22 BIT ADDRE_SS MODE
' MK-2850

Figure 3-1 DMVll CSRS Byte and Word
Symbohc Addresses

;

asELs

BSEL7

Figure 3-2

{

| Namev'

|R®}

;

COMMAND/RESPONSE
FIELDS

| IEl |BSELO
COMMAND TYPE | BSEL2

COMMAND/RESPONSE

|BsEL4

'BSEL6

SEL6

MK-1636

Fixed and Variable Formats for Commands ‘andRéspOnses’ |
Table 3-1

Bits

)

SELO Bit Functions

‘Description

BSELO

Interrupt

1-3

| R‘es:er'vedz

‘When set, thls bitenables the DMV11, upon"as’Serting RDI (bit 4

Enable In (IED) | of BSEL2), to generate an interrupt to vector address XX0.

Interrupt

Enable Out

Reserved

When set, this bit enables the DMV11,

upon asserting RDO (bit

7 of BSEL2), to generate an interrupt to vector address XX4.

3-2

Table 3-1
Bits

SELO Bit Functions (Cont)

Name

Description

Request In

This bit is set by the user program to request access to the data

(RQI)

port. It is cleared by the user program when the data port is not

required for further issuing of commands. The user program may
leave RQI set if successive requests for the data port are pending.

BSELI

Maintenance
Request
|

When set, along with master clear (bit 14 of SEL4), this bit
. causes the DMV11 to enter the maintenance register emulatlon
section of the microcode.

NOTE
Detailed discussion of maintenance register emula-

tion is presentedin Section 4.8.
9-10

Reserved

Diagnostic

Reserved

Mode

“When set, this bit allows diagnostic programs to change the mode
of operation of the DMV11 using the mode definition command
to override the mode switches.

Invoke P/MOP | Invoke primary MOP mode. When set to one, this bit causes the

Boot

DMV11 at this multipoint station to request that the control sta-

tion initiate the primary MOP (maintenance operation protocol)
boot procedure. In point-to-point networks, a DMV11 having this
bit set requests the other station to initiate the primary MOP
boot procedure.

NOTE
The master clear bit (bit 14) must also be asserted
to use mvoke P/MOP.

Master Clear

When set, this b1t initializes the DMVll The clock is enabled
and the RUN flip-flopis set. Master clearis self-clearing.

Run

This bit controls running of the microprocessor. It is set by bus
initialization or master clear. When run is cleared the microprocessor halts.

3-3

Table 3-2

BSEL?2 Bit Functions

Bits

Name

Description

0-2

Control/ Response

These bits define the type of mput command or output response as fol-

Code

lows

Bits

Description’

210

0 0 0

Buffer address/character count

“ '(RCV) command or buffer disposi~ tion (RCV complete) response

Control command or control re-sponse

information response

Buffer dlsposmon (RCV unused)

response

1 0 0

Buffer address/character count

(XMIT) command or buffer dis-

position (XMIT complete) re-

- 1 01

sponse

,Reserved

110 Buffer disposition (sent but not
A ‘acknowledged) response

Buffer disposition (not sent)
response

22-Bit Mode

| Ready In
(RDI)
:

This bit when set 1ndlcates to the DMVII that the buffer addressis in

| the 22-bit format.

)

RDI is the DMV11 response to RQI, indicating to the user program

that it has control
of the CSRs to issue a command. It is cleared by the
“user program when the data port contains the input command. Clearing
RDI returns control back to the DMV11.

5-6

Reserved

Ready Out
(RDO)

Modedefinition command or

0 1 0

RDO is asserted by the DMV11 to indicate that the data ports (SEL4
and SEL6) contain an output response for the user program. The user
program must clear RDO after it has read this information. Clearlng
RDO returns the CSRs to the DMV11.

3.2.2 Input Commands Overview
|
In general, input commands provide the means for the user program to assign, receive, or transmit buf—
fers to the DMVI11. Detailed field dCSCI‘lpthHS and formats of each mput command are prowded in
Section 3.3
There are four types of input commands that can‘ be issued to the DMV11

for execution.

Mlcroprocessor control/mamtenance command
Mode definition;
Control;
Buffer address/character count
| W1th the exception of the microprocessor control/maintenance command, input commands require an
identification code in the first three bits of BSEL2 (see Figure 3-2). These codes, which define each
command and variations of spe01f1c commands within the command set are deflnedin Table 3- 2 and
listed i1 n Table 3-3.
»
|
»
|

NOTE
CSR addresses are expressedin octal

Input Command Type

Table 3-3

Input Command Codes

- 1
'

Bit

Mode definition

Control

]

Buffer address/character count
‘(I‘CCCIVC)
-

0
-

Binar.y Code(BSEL2)

Buffer address/character count |

(transmit)_

N IR

3.2.3 Output Responses Overview
|
|
|
Output responses provide a means for the DMV11 to report various normal and abnormal (error) condltions concerning
the data transfer operation. Three basic responses are provided:
v
|
e

Buffer disposition;

Control;

Information.

The buffer disposition response is used to return both used and unused buffers to the user program.

The control response is used to report error conditions concerning the microcontroller/line unit hardware, data link, physical link, or remote station. It also passes protocol information to the user.

The 1nformat10n response provides 1nformatlon requested by a control command from the user program.

3-5

3.3

DMYVI11 INPUT COMMANDS

This section provides detailed descriptions of each input command. Command formats and data port
usages are illustrated and defined. User-program execution requirements, command variables, and action taken by the DMV11 in response to commands are discussed.
3.3.1

Microprocessor Control/Maintenance Command

This single byte command has two functions; to initialize and cause the DMV11 to start runnmg, and to
cause entry into the microcode maintenance loop when the maintenance request bitis set. At start-up
time under normal operating conditions, thisis the first command issued by the user program in order
to initialize the DMV11.

The format for the DMVI11 initialization register (BSELI) is shown in Figure 3-3. To set the master
clear bit and thereby cause entry into the DMV11 running mode, the user program moves a byte with
an octal value of 100 to BSEL1. As a result, all condition-sensitive logic in the DMV11 is reset for startup, and the start-up diagnosticis executed. When the diagnostic completes satisfactorily, the run bitin
BSEL]1 is set to one. This indicates that the DMV11 is running and the microcode is executing.
Figure 3-4 presents a flow chart describing how to initialize the DMV11. A timeout counter is set to
avoid the possibility of the user program bemg caughtin an endless loopin case the internal diagnostic
does not complete successfully.

BSELT IRUN cLR OO
-

MST

|moDE

|DIAG

| REQJ
MNT

MK-2513

- Figure 3-3

Microprocessor Control/Maintenance
Command Format

3.3.2

Mode Definition Command

Functionally, the mode definition command is used to establish the hlerarchy of a network and the
characteristics of the communications line serving that network. As shownin Figure 3-5, the mode definition command contains two fields; the command type code field in BSEL2, and the mode field in
BSEL6. The mode field contains a code defining the function to be performed by the command.
With the mode definition command, the user program can designate the DMV11 as a control station, a
tributaryin a multipoint network, or as a nodein a point-to-point network. In addition, the characteristics (half-duplex or full-duplex) of the physical communications line connecting the network can be
defined.

The actual mode field codes and the functions implemented by each code are listedin Table 3-4.
Under normal operating conditions, the mode definition command is issued by the user program at
start-up time (after the internal mlcrodlagnostlcs have executed successfully and the run bit is set)
Network discipline requires that each DMV11 in a networkissue a mode definition command that is
appropriate to the network. For example, in a half—duplex mult1pomt network comprised solely of
DMVlls:

The user program at the control station issues a mode definition command with the mode

The user program at each tributary station issues 4 mode def1n1t10n command with the mode
~

field set at four.
field set to six.

3-6

SET MASTER
CLEAR BIT
IN BSEL1

(BSEL= 100)

SET
TIME OUT
\A

COUNTER

«/

‘

> 0.5 SEC

7
7~

RUN

BITSET?

\\BSEL1=200_ )

| DIAGNOSTIC

| ERROR, EXIT
TM
1
T
O
E
R
R
O
R

| CONTINGENCY

EXIT TO

COMPLETE
START UP

‘MK-1638

Figure 3-4

Initialization of the DMV11
3-7

This network discipline also applies to DMV1ls operating in pomt to-point networks W1th other
DMV1l1s, DMPl11s, DMClls, and DMR11s.

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 functions performed by the mode definition command can also be implemented by the mode selection switches on the DMV11 module. The switches must be used to establish mode definition functions
when tributary addresses are switch assigned. The switch setting for performing the mode definition
functions corresponds to the BSEL6 codes listed in Table 3-4.

Once the type of station is set, it can only be changed by a'_master_clcar or a physical change in the

switches. If the type of station is switch assigned, a master clear has no affect. However, the switches
are overridden when the diagnostic mode bit (bit 3 of BSELI)is set and a mode definition commandis
issued.

BSEL3

BSEL7

L
4

[

RDO

| B0

]

1 'oIRQ'

TRIBUTARY ADDRESS

BSEL5

RDI

2
O

'El |BSELO

-1

CMD TYPE CODE

BSEL2

|gg
o

BSEL4

]

SEL4

MODE FIELD | BSELE

|
4

]

SEL6

MK-1639

Figure 3-5

Mode Definition Command Format

Table 3-4
BSELS6 Bit
Positions
2

0O
0O
O
0O
1
1
1
1

0
0
1
1
0
0
1
1

O
1
O
1
O
1
O
1

Mode Field Codes and Functions

Line
Characteristics

Network
Configuration

DMC11-Line
Compatibility?

Half-duplex
Full-duplex
Half-duplex
Full-duplex
Half-duplex
Full-duplex
Half-duplex
Full-duplex

Point-to-point
Point-to-point
Point-to-point
Point-to-point
Control station
Control station
Tributary station
Tributary station

Yes
Yes
No
No
N/A
N/A
N/A
N/A

3-8

3.3.3

Control Command

This command is the pr1mary means of controllmg the operatlon of DMV11 -implemented networks.
The format of the control command1S 1llustratedin Figure 3-6.

At start-up t1mc the user program at the DMV11 control statlon must issue one control command (es-

tablish trlbutary) for each tributary address supportedin the multipoint network. This must be done
after issuing the microprocessor control and mode definition commands. This causes the microcode to
create a tributary status slot (TSS) in the DMV11 data memory for each tributary in the network.

Similarly, the user program at each tributary must issue a control command (establish tributary) for
each tributary to be established at that statlon This causes a TSS to be created at that station for each
trlbutary it cstabllshes

In pomt-to-pomt networks a control command (cstabhsh trlbutary) must be 1ssued at both stations. The

tributary address fieldin this case must be a one. ThlS resultsin the creationof a smgle TSS structure
PN
i

at each station.

The DMV 11 microcode at the control station and at all tributary stations uses these TSS structures to
coordinate protocol operation over the network between the control/tributary pair. User programs, at
the control station and at the cstabhshcd trlbutarles access these structures to obtam operatlonal information such as:

e
e
e
e

The number of messages transmitted and received,;
The number of selection intervals and timeouts;
The number
of reply timeouts; and
The number of various types of errors.

//,«-'-\ .

sseL7|

14
-

| l |

POLLING |
|
STATE

UNLATCH
POLLING
STATE

]

DISABLE

10
-~

|Tss |yss |TSS
7

| | |

|IBSEL2

BSEL4
SEL4

—————

REQUEST KEY OR

TSS ADDRESS
3

|BSELO

]

‘

IEl

baA

1 | .WR'G
READ|
CLR

Pa» 12 11

LATCH

)

E
i

|
DATA

BSELS

)

—

BSEL6

SEL6
0

COMMON
BUFFER

PoOL
~

ENABLE
COMMON
BUFFER

‘POOL
MK-1640

Figure 3-6 Control Command Format

At the start-up of tributary stations having multiple software assigned tributary addresses, the user program at each station issues as many control commands as there are established tributaries at that node.
However, in networks where tributary addresses are switch assigned, only one control command specifying the switch assigned address is required as part of the start-up sequence. Any other nonzero tributary address in a control command is overridden by the switches.
The control commandis also used by the user program at the control station to specrfy a unique set of
polling parameters for each tributaryin the network.
During normal operation, the control station microcode can determine the polling level of any tributary
in the network and adjust the polling frequency of that tributary as necessary (see Section 5.2).
The control command permits the user program to perform a number of control functions using the
same command format.
In general, each function implemented by this command requires issuing of a

single appropriately formatted control command. SEL6 is used to define the various functions of the
control command (see Figure 3-6). Table 3-5 provides a detailed description of each bit or field.

Table 3-5 -SEL6 Control Command Functions
Bit | Name

Description

0-4 | Request Key

These five bits are encoded requests frem the user program. When this field is

used, bits 5 through 7 must be cleared Request keys are encoded as shownin
Table 3-6.
.

|Read TSS/GSS | A control command with this bit set, allows the user program to read two consecutive locations (two bytes) of a tributary status slot (TSS) or global status
slot (GSS) without modifying it.

The TSS to be read is specified in BSEL3 and the location within the TSS is

When the DMV11 receives a control command to read a TSS or GSS location,
it passes the requested information to the user program through an information

response (see Section 3.4. 3) However, the requested information is placed on
an internal queue before it is passed to the user program. As a result, the information requested may change before the user gets it. This is particularly true
for number of messages transmitted /received and selection intervals.

|Read and

- A control command with this bit_set, allows the user program to read and clear

Clear TSS/GSS | specific locations in a TSS or GSS.

The TSS to be read is specified
in BSEL3 and the location within the TSS is
‘specified in bits 0-4 of BSEL6. Notice that a read and clear command bit 6 of
BSELSG6 is also set. This gives a base octal value of 100 to which the specific
TSS address is added. To read and clear a GSS location, BSEL3 is zero.

Only the error counter sections of the TSS (7-17 octal) and GSS (15-17 octal)
can be accessed with a read and clear command.

3-10

specifiedin bits 0-4 of BSEL6. Notice that bit 5 is also set to indicate a read
GSS. To read a GSS locatlon BSEL3 is zero.
-

Table 3-5

Bit

Name

SEL6 Control Command Functions (Cont)

Description

Accessing a_ny othér locations results in a p__rocédural e_fror. |

Valid octal values for BSEL6 for the read and clear function are listed below:

Octal Value

) TSS'Locatimi .

107

110
111
112
113
114
115
116
117

|
-

116

117

Data messages received
~ Selection intervals
Data errors outbound
Data errors inbound
~ Local buffer errors
Remote buffer errors
Selection timeouts
Local and remote reply tlmeouts

GSS Location

Octal Value
115

Data messages transmitted

Remote station errors

Local station errors

Global header blockcheck and

maintenance data blockcheck
|
errors

These errors are covered in
detail Section 5.3.

When the DMV11 receives a control command to read and clear a TSS or GSS

location, it passes the requested information to the user program through an information response, and then clears the location.

As in the case of the read TSS/GSS the information is placed on an internal

queue andis subject to change before the user gets it. However, by readmg and
clearing, the user can keep a cumulatlve total.

Write TSS

A control command w1th this bit set, cnables the user program to write into spe01f1c locations in an assomated TSS or GSS.

The TSS to be written into is speC1fledin BSEL3 and the specific location within the TSS orGSSis specifiedin bits 0-4 of BSEL6. To write to a GSS loca|

tion, BSEL3 is zero.

Notice that bit 7 of BSELS'is alsd set to indicate a Write TSS. This gives a base

octal value of 200 to which the specific TSS address is added.
3-11

Table 3-5

Description

The data to be written is contained in BSEL4 and BSELS. There are eight TSS
and five GSS parameters that can be written:

TSS PARAMETER
.

Transmit delay timer (XDT)

BSELS6

230

Initial polling urgency (Q) and

polling rate (R) for active
state

231

Initial polling urgency (Q) and

polling rate (R) for inactive
state

232

Initial polling urgency (Q) and

polling rate (R) for unresponsive
state

233

Nodata message count (non-inact)
and unresponsive timeout count
(TO-UNRESP)

234

Dead timeout count (TO-DEAD)
and maximum message count (MMCQC)

235

Selection interval timing
counter

Babbling tributary counter

See Table 4-2 for details

236

GSS PARAMETER

BSEL6

Number of sync-characters to

precede nonabutting messages

233

Preset value for streaming

tributary time counter
3.
|

234

Polling algorithm update
~interval (DELTA T)

235

Polling rate for dead tribu-

taries (DEAD T)
5.

236

Fixed polling delay (poll

delay)

See Table 4-3 for details
-

237

NOTE

Some parameters are 8-bits in length. Thus, in those
cases two parameters are indicated. All user accesses are on two byte boundaries.
i

Bit | Name

SEL6 Control Command Functions (Cont)

Table 3-5

Bit

SEL6 Control Command Functions (Cont)

Name

Description

Enable

A control command with this bit set allows a specified tributary (BSEL3) to use
the common receive buffer pool. Usage of the common pool is based on a com~mon pool quota. This quota is determined for the specified tributary by adding -

‘Common
Pool

the octal value in BSEL4 to the current quota. If this results in a value equal to
or greater than 377 octal, a procedural error results and the quota is reset to
376. However, a tributary may be set up for unlimited use of the common pool
by setting BSEL4 to 377. Each time a tributary uses a common pool buffer, the

quota is decremented by one. When the quota reaches zero, the tributary is prevented from using the common pool. See Section 3.3.4 for more details on common buffer pool.
|
The common pool is checked first. If no buffers are available or the quota is
zero, or the buffer is too small, the private receive buffers are checked.

Disable
Common Pool

A control command with this bit set, disables the use of the common receive
buffer pool for the tributary specified in BSEL3.

The quota previously established for this tributary is cleared to zero.

1011

Reserved

Unlatch Polling State

Latch Polling State

A control Command with this bit set, causes the polling state level of the tributa-

ry addressed by BSEL3 to go to the active polling state. Control of the polling
activity for the specified tributary is then returned to the polling algorithm.

A control command with this bit set, establishes the polling state of the tributary addressed by BSEL3. The polling state is determined by bits 0 and 1 of
BSELA4. These bits are encoded as follow:
Bits
00
01
10
11

1&0

Polling State
Active
Inactive
Unresponsible

Dead

3-13

Table 3-6

Octal
Code

|
Name

No ReqtieSt' o

Request Key Field Definitions (Control Command)

Description

This code allows the1ssumg of a null eontrol command for the purpose

- of returning control of the CSRs to the DMV11. The no request code
~1s.used when RDIis set but thereiis no command to issuc (see Section
- 4.2.3). This is effectively an NOP command.

~ The

- NOTE

enable/disable

common

pool

and/or

(TSS) data structure. This must be accomplished before any com-

Tributary

~ This control funetiofi'ihitiates the creation of the tributary status slot

Establish

mand is issued that uses a tributary address.

The user pfogram at the control station must issue one establish tri-

- butary control command for each tributary supported in the network.
The tributary address is designated in BSEL3.

|
TN

//“""N\‘v :

latch/unlatch polling state bits in BSEL7 can be
usedin conjunction with this request key.
o

NOTE

In a point-to-point network, this control command,
with a tributary address of one in BSEL3, should be
o 1ssued at each statlon to establlsh the requnred TSS.

The DMVll h'as 12 available .TSS blocl(s Each block has 64 — $-bit

locations for storing status and other mformatmn necessary for main-

| ..tammg commumcatlons over the data link.

As a result ef establishing one or more tributary addresses, the

DMVI11 creates a global status slot (GSS). This GSS is part of the

overall status structure at each station.
‘NOTE

This control command function can also be used
during normal operation ofa multipoint network to
- reestablish previously deleted tributaries. In such
cases, the tributary address must be reestablished at
both the control station and the pertinent tributary
station.

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.

3-14

Request Key Field Definitions (Control Command) (Cont)

Table 3-6
Octal

Code

Name

Description

Delete
Tributary

This control function removes a specified tributary from operational
by eliminating its associated TSS. Prior to issuing this comstatus

mand, the user program must first halt the tributary being deleted.
[See request key 05 (request halt state)]. The TSS can only be reestablished by using an establish tributary function. Only 12 addresses
|

may be established at any one time.

Request Start-up | This control function initializes the designated TSS and initiates the
State

DDCMP start-up sequence for that tributary. BSEL3 specifies the tributary address. Request start-up state must only be issued to tribu-

taries that'are in the halt state.

When the start-up sequence is completed, the DMV11 notifies the
user program by issuing a control response. When this response (run
state) is received by the user programs at the tributary and control
station, message traffic can begin between these two stations (see
|

Table 3-7).

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 Maint
State

This control function places the tributary designated by BSEL3 into
|
|
the DDCMP maintenance state.
A tributary placed in the maintenance state can only transmit and receive maintenance messages. Both the control station and tributary
must be in the maintenance state in order for maintenance message
|

traffic to occur.

The maintenance state must only be issued to tributaries that are in
the halt state.

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 con|
trol command.

3-15

Table 3-6

Octal
Code

Request Key Field Definitions (Control Command) (Cont)

Name

Description

Request Halt
State

- This control function places the tributary designated by BSEL3 into
the DDCMP halt state. All outstanding buffers are returned.

- When aitribu.tary’ is halted at the control station, the tributary 1S no
longer polled. When a tributary is halted by its own user program, it

no longer responds to polling.

The TSS for the halted tributary remains unchanged at both the con-

- -trol and tributary stations. The halted tributary can be restarted by
issuing a request start-up state control command.

NOTE
The halt state may also be used on a global basis to
‘return all common pool buffers. This is accom~

plished

BSEL3 when

using a

tributary

address

of zero

issuing the request halt state.

Tributaries
must not be using the common pool when
this function is issued. If the common pool is in use,
a procedural error of 136 is generated (see Table 3-

7).

Read Modem

Control

This c‘ofitrolfunctioh,permits the riser program to write the contents of

- BSEL4 into the DMVI11 modem register. (See Appendix C).

Write Modem

This control function causes the DMV11 to read the modem register

Control

and pass this information to the user program through SEL4 by way
of an information response. (See Appendrx C.)
NOTE
Request key codes 6-17 and 22-37 are reserved.

3.3.4 Buffer Address/Character Count (BA/CC) Cornmand

This command provides the user programs at both the control station and tributaries with the means to
assign, transmit, and receive buffers.

The format for this command is shown in Figure 3-7. Note that the command has two forms to facilitate
separate management of transmit and receive buffers. These two forms are distinguished by the type
code in BSEL2. A type code of zero is used to allocate receive buffers and a type code of fouris used to

allocate transmit buffers.

The tributary addressis specrfred by BSEL3 and the buffer addressis contalnedin SEL4 and bits 14

and 15 of SEL6. The remammg 14 bits of SEL6 contain the character count in posmve notation. A
character count of zero is illegal.

In addition to allocating receive and transmit buffers, the buffer address/character count command is
used to allocate common receive buffers by specifying a tributary address of zero in BSEL3. These
buffers can only be used by tributaries authorized to do so by the enable common pool bit of a control
command.

When the user program has a message to transmit, it informs the DMV11 of the size and address of the
message buffer. This is done by the buffer address/character count command on a one buffer per command basis. A tributary address of zero when assigning transmit buffers results in a procedural error.
When the user program is receiving messages, it assigns receive buffers on a one buffer per command

basis using the buffer address/character count command. These buffers may be in the common pool of
buffers or be private buffers. Each BA/CC command used to assign a buffer to be the common pool
must contain a zero in BSEL3. Although this command assigns buffers to a common pool, actual allocation i1s performed through the control command by enabling access to the common pool on a per tributary basis.
|
|
|
|
'
o
-

BSEL3

[

asgL7| B/A | B/A
17

:
13

HIGH

RDO

TYPE
CMD
,
P
'

RDI

LOW

BYTE

18 BIT ADDRESS MODE

SEL4

BSEL6

SELGE

| LOW IBYTE 1

|BSEL2

BSEL4

CHARACTER COUNT

|BSELO
BSel

BUFFER ADDRESS

HIIGH S!X BITSl

IEO

BYTE

CHARACTER COUNT
12

5
'

o | R¢
|

BUFFER ADDRESS

ADDRESS
_ TRIBUTARY DDRE

BSELS

|
MK-1641

BSEL3

BSELS

)

]

HIGH BYTE

UNULSED
14

_
13

22BIT]

RDI

)

CMD TYPE

BUFFER ADDRESS

BSEL4
|sELs

SELE

N
> | UNUSED

BUS ADDRESS BITS 16-21

_LOW BYTE

1
6

CHARACTER COUNT

| LOW BYTE

SELo

[BsEwo
BSEL?

'
cope*

M(1)DE

HIGH SIX BITS
11

RDO

CHARACTER COUNT
12

s
'
UNUSED
L

BUFFER ADDRESS

BSEL11

TRIIBUTAR?( ADD.RESS

|
BSEL7|

]

|R«

BSEL6

BSEL10

SEL10
0

22 BIT ADDRESS MODE

* 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
3-17

MK-2512

allocation
In multi‘point‘networks, user programs at both the control and tributary stations can handle L
.

.‘

in two ways:
of receive buffers

The first method involves the allocation of receive buffers from a common pool of buffers.
the common buffer pool enabled, receive buffers are assigned to the pool through the
With
buffer address/character count command on the basis of one buffer for each command isto the common
“sued. Each buffer address/character count command used to assign a buffer

pool must contain a zero in BSEL3. Although this command assigns buffers to a common
pool, actual allocation toa tributary is done through the control command by enabling access
to the pool and assigning quotas (see Table 3-5).

the user program can directly allocate private receive buffers based on
2. Inthe second method,

anticipated message traffic using one buffer address/character count command for each private buffer allocated. In this context, a private buffer is defined as a receive buffer assigned
to a specific tributary for its exclusive use, and in all cases the address of that tributary must
be in BSEL3 of the assigning command.

Private buffers and buffers from a common pool can be used jointly at control and tributary stations in

are availa multipoint network. Under these circumstances, theadvantages provided by both methods
to the user program.
able

The preceding information involved standard 18-bit addressing. However, the DMV11 may also oper-

Priva‘te"buffe_rs can be set for unanticipated messages and/or abnormally large messages.

ate in 22-bit addressing mode. The DMV11 allows the software to set the 22-bit address mode of operation. The bit indicating 22-bit address mode is set by the user program as part of a command when
issuing transmit or receive buffers to the DMV11. The state of this bit is retained to indicate to the

3.4 DMV11 OUTPUT RESPONSES

has a set of three responses it can use to either reply to user-program comThe DMV11 microcode

- mands or to inform the user program of error conditions. These responses are:

‘e

Buffer disposition response;

Control response;

Information response.

As with input commands, each output response is identified by a typecode in bits zero, one, and two of

BSEL?2.

SRS
NOTE
conissued
response
each
networks,
In multipoint
that
tributary
the
of
address
the
BSEL3)
tains (in
point-to-point
in
However,
response.
the
to
relates
- networks, equivalent responses always contain a one
R

inBSEL3.-

3-18

/,/”m\\

DMV11 the number of bits in the buffer address.

3.4.1
Buffer Dlsposmon Response
|
This response is used to return both used and unused buffers to the user program. Figure 3-8 shows the
format of this response and identifies the five methods of message disposition in BSEL2. BSEL2 bits
zero, one, and two are encoded as follows:
Bit2

Bitl

Bit0

Buffer Disposition
Receive buffer complete

- Transmit buffer complete

Transmit buffer sent but not acknowledged

Transmit buffer not sent

Receive buffer unused
|

—

Receive Buffer Complete — When a message is received successfully and stored in the assigned main
memory buffer, the DMV11 microcode notifies the user program by issuing a buffer disposition response with a type code of 000. Data in the receive buffer is not valid until such a response is issued.
This response is issued for both common pool and private buffers.
|

Receive Buffer Unused — When the protocol for a specified tributary is haited, the DMV11 automatically returns all associated private receive buffers to the user program. To do this the DMV11 issues a
buffer disposition response with a type code of 011 for each outstanding buffer held by the specified
tributary. This pertains to private buffers only.
:

S
:

BSEL3

)

B:SELS:

4
1

3
|

)

BseL7 | B/A|

RDO

B/A

| 16

CHARACTER COUNT
HIGHSIXBITS

S
o
/

RSEL3

BSELS

;

—|

HIGH BYTE

SEL

UNUSED
|

HIGH SIX BITS

RDI

BSEL6

SEL6

)]

IEl

22BIT|

cMD
TYPE
~

|MODE

_ LOW BYTE

UNUSED

)

L
6

SEL2

CH'ARACTER COUNT

| LOW BYTE

SEL4

ggge

BSEL11

o
1

|BSELO
BSEL4

BUS ADDRESS BITS 16-21
1

MK-1642

| BSEL2

—BUFFER lADDRESS
1
FODE"
.

|sELs

RDO

BSEL2

|BSEL4

|BSELO
SEL2

IEl

CODE*

. LOW BYTE

IEO

CMD TYPE

CHARACTER COUNT

—
1

CHARACTER COUNT

UNUSED

156

LOW BYTE

o | R
'

—t

BUFFER ADDRESS

BSEL7 |
BSEL11]

i
e o
'
!
TRIBUTARY ADDRESS

+—

BUFFERADDRESS @

18 BIT ADDRESS MODE

)

RDI

'
.

|
.

BUFFER ADDRESS
. HIGH BYTE

|RA|

TRI?UTARY ADDRESS

SEL10

22 BIT ADDRESS MODE

% BUFFER DISPOSITION RESPONSE TYPE CODES:
1.

RECEIVE BUFFER COMPLETE = 000

RECEIVE BUFFER UNUSED = 011

3. TRANSMIT BUFFER COMPLETE =100
4. TRANSMIT BUFFER SENT NOT ACK'd

=110

5. TRANSMIT BUFFER NOT SENT = 111
S

Figure 3-8

Buffer Disposition Response Format

3-19

MK:2515

The DMV11 returns all common pool buffers if the user program issues a request halt state control
command with a tributary address of zero. When all common pool buffers are returned, the DMV11
issues an information response indicating that the process is finished.
o

Transmit Buffer Complete — When a message is transmitted successfully, the DMVll notifies the user
program by issuing a buffer disposition response with
a type code of 100. Successful transmission
means that the receiving station has acknowledged receipt of the message.

Transmit Buffer Sent But Not Acknowledged - When the protocol for a specified tributary is halted, the
DMV11 automatically returns all transmit buffers currently being processed by that tributary to the
user program. To do this, the microcodeissues a buffer d1spos1t10n response with a type code of 110 for
each buffer sent but not acknowledged.
o
NOTE
Durmg protocol operation, after seven unacknowledged transmissions of a message occur, the transmit threshold error is exceeded and the DMV11 issues a comtrol response indicating this error (see
~ Section 3.4.2). The DMV11 continues to retransmit

the message and responsibility for terminating
the
transmission belongs to the user program

Transmit Buffer Not Sent - DMV11 maintains a queue of buffers to be transmltted for each tributary
address establishedin the network. When the protocol for a given tributaryis halted, the DMV11 returns all unused buffersin the associated queue to the user program. To do this, the DMV11 issues a

buffer dlsposmon response wrth a ‘type code of 111 for each transmit buffer remarnmg in the queue.

The other CSRs used by the buffer d1spos1t10n response are BSEL3 SEL4 and SEL6. The function of
these CSRsis as follows

0 -

BSEL3 specrfres the trrbutary address associated w1th the. buffer drsposrtlon response

SEL4 and bits 14 and 15 of SEL6 contam the 1:8-b1t buffer address for the buffer being
completed or returned. For 22-bit address mode, bits 0-5 of BSEL6 are used with SEL4.

SEL6 (or SEL10in 22-bit address mode), bits zero through 13, compose a 14-bit character

count allowing for a maximum buffer size of 16,383 bytes. A character count returned by a
buffer dlsposmon response is designatedin positive binary notation.
»

Protocol18 halted for a trrbutaryin one of three ways when:
1.

The user program rssuesa control,command halting the tributary._

v 2. A DDCMP STRT message is received while the tributary is invv the run state. 3.

A DDCMP mamtenance message 1s recerved temporarlly haltlng the protocol, while receive
buffers are being returned.

3.4.2

Control Response

A control response is an unsolicited response issued by the DMV11 when an error is deteeted or when
protocol information must be passed to the user program. The format for the controlresponse is shown
in Figure 3-9. BSEL3 contains the tributary address or character count, SEL4 and bits 8 through 13 of
SEL6 contain the bus address for the affected buffer, and BSEL6 indicates the type of information
being passed to the user program. In 22-bit address mode, SEL6 contains buffer address bits 16-21 and
3-20

BSEL10 indicates the type of information being passed to the user program. These types of information
are indicated by an octal code and are listedin Table 3-7. Basically there are four categories of information.

1.
2.
3.
4.

System events
Protocol events
Network errors
Procedural errors
Table 3-7

Octal

Output Codes

Code

Category

002

Network
Error

Information

Receive threshold error — This error is reported to the user program when the

number of consecutive receive errors equals seven. These receive error types
are:
|
|
1.
2.
3.
4.
5.
6.
7.

Message header blockcheck.

Message data blockcheck.
|
NAK in response to a DDCMP reply message.
Buffer temporarily unavailable.
Receive message overrun.

Message header format error.
When a message is too long for the
available buffer.

Each time the receive threshold error is reported, the counter is reset to zero.

004

Network |

Error

Transmit threshold error — This error is reported when the number of consecutive transmit errors equals seven. These transmit errors consist of four
types and occur when:

A STRT message is transmitted but not acknowledged within the timeout

A STACK message is transmitted but not acknowledged within the

period.

timeout period.

ANAKis receivedin responseto a transmission with a reason code other
~

than REP response.
A data message is transmitted but a reply was not received within the
timeout period.
,

Each time a transmit threshold error is reported the transmit error counter is

reset to zero, except when remaining in the DDCMP ISTRT and ASTRT
states.

006

Network
Error

Select threshold error — This error is reported when the selection interval time-

out counter for a given station has timed out seven times. The selection inter-

val is the time allocated for a tributary to respond to a poll or for a point-topoint station to respond to a transmission.
Each time the select threshold error 1S reported the selectlon error counter is
reset to zero.

3-21

Table 3-7

Octal
Code

Category

Output Codes (Cont)

Information

NOTE

For more information on receive, transmit, or selection threshold errors, see Section 5.3.3.
010

Protocol
Event

Start message received while running — This response indicates that a
DDCMP STRT message was received from the station specified in BSEL3

while this station was in the DDCMP run state.

In normal operation, this message is used by a centfol station to inform a tri-

butary that the control station has started protocol operation for that tributary.

The start message is also used to resynchroniZe the logical link between a con-

When the DDCMP STRT message is received: 1) the DMV11 responds with

start message received while runmng, and 2) the protocol at the tributary halts

trol station and tributary. This might be necessary when message traffic is inhibited because of threshold errors or receive or transmit overruns.

and all buffers are returned.

At this time the logical link may be restarted (request start-up state control

command).

Pretocol |

Maintenance mess’agereceiVed while running (or ISTRT or ASTRT)‘ This

Event

response indicates that a DDCMP maintenance message was received by a
‘specific tributary while it was in the DDCMP run, ISTRT, or ASTRT state.
This causes the tributary to notify the user program, return all unused buffers
to the user program, and halt the protocol. The DMV11 then places the tributary in the maintenance state. The message that caused this event is lost.

Protocol

‘Maintenance message received while halted — This response indicates that a

TM,

012

014

Event

DDCMP maintenance message was received by a specific tributary while it
~was halted. This places the tributaryin the mamtenance state. The message
that caused this event is lost.

Protocol
Event

Start message received while in the maintenance state — This response indicates that a DDCMP STRT message was received by a specific tributary
while it was in the maintenance state. No further action is taken by the microcode. The user program determines what action to take based on this response.

022 | System -

| Tributary polling state dead This response informs the user program at a

016

- Event

control station that the spe01f1ed tributary pollmg state has gone to the dead

| state.

This response has no meaning in point-'to?-point networks.
3-22

Table 3-7

Output Codes (Cont)

Octal
Code

Category

Information

024

Protocol

Run state — This response informs the user program at all stations that a specific tributary has entered the run state. The DMV11 microcode issues this
response as a result of receiving a request start-up state control command and
having completed the DDCMP start-up sequence.

Event

026

Network
Error

Babbling tributary — This response is issued to a user program to record the

occurrence of a babbling tributary. It is only used by half-duplex point-to-point
and multipoint control stations.
A babbling tributaryis one which continues to transmit valid control messages
or message headers beyond the end of a programmable timeout (babbling tributary timeout counter).
The babbling tributary response usually indicates a malfunction at the transmitting station or too short a period for the timeout counter.
- Recovery from this error condition generally requires human intervention at
the remote station.

030

Network
Error

Streaming tributary — This response is issued to a user program to indicate

that the remote station failed to release the channel at the end of the selection
interval. It is only used by half-duplex point-to-point and multipoint control

stations.

The streaming tributary timeout period is determined by the programmable
streaming tributary timeout counter. The timeout period starts at the end of
the selection interval.
This error condition generally indicates a modem malfunction at the remote
station or an inappropriate choice of the streaming tributary timeout counter.

Recovery from this error condition generally requires human intervention at
the remote station.

32-76

Reserved

100-276 Procedural
Error

Procedural errors — This response is issued by the DMV11 when the user program violates the procedures designated for interfacing the DMV11. In all
cases, the event that caused the procedural error is ignored by the microcode.
All procedural errors except octal codes 140, 300, and 302 are reported immediately to the user program. Those procedural errors are placed on the response queue. Specific procedural violations are 1dent1f1ed by the octal codein
BSELS6 as follows.

3-23

Table 3-7

Code

Category

Information

Code

- 100

Description

A command other than a mode defimtron commandis issued before
the mode has been established.

102

Invalid type code usedin a command

104

Invahd mode change (for example the mode of a trlbutary station is

106

A nonglobal ommandis issued to an unestablished trrbutary

110

A nonglobal commandis 1ssued havmg a trrbutary address of zero.

112

Attempt to delete or place an unhalted tributary in the start or

114

to establish more than 12 tributaries.
Attempt

116

Attempt to establish an already established tributary.

120

- An invalid request key is usedin a control command.

changed to point-to-point).-

maintenance protocol state.

Octal

Output Codes (Cont)

122

Attempt to ass1gn a buffer for an unestabhshed trrbutary

124

Attempt to assign a buffer for a halted tributary

126

Attempt to assign a buffer havmg a byte count of zero.

130

Attempt to assrgn a transmlt buffer W1th a tr1butary address of zero.

132

Attempt to writeor read and clear a reserved area of a tributary or

| 1.34

Attempt to use the reserved brts in BSEL7 of control command

136

Attempt to return all common receive buffers while the common

140

Attempt toraise the common pool buffer quota to a value hrgher than

global status slot.

buffer poolis being used.

376 octal

142276 Reserved

300

Procedural | Buffer too small — This response informs the user program that the assigned
Error

receive buffer is not large enough for the current incoming message. Recovery
from this error condition is detailed in Section 4.6.2.2.

3-24

Table 3-7

Output Codes (Cont)

Octal |
Code |

Category |

Information

302

Procedural

Nonexistent memory This response is issued to the user program when the

Error

DMV11 microcode attempts to access a CPU ‘memory location that does not

)

respond. The address that caused the error is returnedin BSEL4, BSELS, and
BSEL7 for this response

This error is fatal to the station that initiated the memory access causing the
error. That station must be halted and restarted (see Section 4.6.2.1).

NOTE
- With the exception of the buffer too small and non-

existent memory errors, the only control response
L

fields used by the DMV11 when posting a procedural error are the type code field ir BSEL2, and
the output code field in BSEL6. All other fields re-

main as set by the user program in the command

- that originally generated the procedural error.

304

System
Event
|

(

Modem disconnected — This response informs the user program that an on-tooff transition of the EIA signal data set ready (DSR) was detected. Such a
transition indicates that the modemis drsconnectrng from the communications

Since this is a global response, the content of BSEL3 is zero.

1306

System

Queue overflow— This response indicates that the free hnked listis empty (see

Event

Section 4.6.2.3).

This error typrcally indicates that for some reason output responses are bemg

(

called for faster than the mrcrocode can process them |
-

e
NOTE
o
This is a global fatal error and the DMVI] initializes itself after attempting to return all informa-

tion (response queue only).

~ Once this response is issued, the user program has three seconds
to retrieve the
‘next pending response from the CSRs. Three seconds are allowed for each
e

pending response before the internal mlerodragnosucs clear the CSRs (see

Section 4.3 for restarting).

310 | System
-

Event

| '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 DMV11
~ was in the process of receiving a message. Since thisis a global response, the

| content of BSEL3 is always zero.

3-25

BSEL3

BSEL7

;

)

TRIBUTARY ADDRESS

RDO

‘
—_

'
}

'
4

'
}

}

IEl | BSELO

RDI M8DE CMD TYPE CODE

'
!

'
}

IE 0

*BUS ADDRESS BITS 16-21
14

o | R

|
2

BSEL5
|

N
= }

—

EVENT/ERROR
CODE

;

}

|
1

SEL2

BSEL4
SEL4

BSEL6

SEL6 |
o)

18 BIT ADDRESS MODE

| *ONLY APPLICABLE FOR NON-EXISTENT MEMORY
AND BUFFER TO SMALL ERRORS

MK-2510

Figure 3-9 | Control-Out Command Format 18-Bit Mode
3.4.3

Information Response

An information response is issued by the DMV11 microcode in reply to a request for information by the
user program. The format for this response is shown in Figure 3-10. The type code for this response is

010 binary as indicated in BSEL2. BSEL3 contains the tributary address and SEL4 contains the information requested by the user program. If the information requested is from a GSS, BSEL3 contains a

ZEro.

An information response contains a return key code or a TSS or GSS address in BSEL6 bits 0-4. The
return key codes are used in response to control request keys or protocol events and are encoded as
shown in Table 3-8. The TSS address is used in response to a read or read and clear TSS/GSS command in conjunction with bits 5 and 6 of BSELS6.

Information responses to read or read and clear TSS/GSS .control commands, use the TSS address

field (bits 0-4 of BSELG6) to indicate the TSS or GSS location read. Bits 5 and 6 indicate whether the
response is to a read or read and clear input command.
3.5

TSS/GSS ACCESS

When a user program accesses a TSS or a GSS through a control command, the DMV11 reads the
designated location and passes the data to the user program by means of an information response.
Reading a TSS or GSS location results in two bytes of data from that location being stored in BSEL4
and BSELS of the information response, with the low-order byte in BSEL4, and the high-order byte in
BSELS. If a TSS is being read, BSEL3 in the information response contains the associated tributary
address. However, data read from a GSS is passed to the user program through an information response
having a zero in BSEL3.

As described in Table 3-5, any TSS or GSS location can be read, and specific locations can be read and
cleared. Along with the TSS/GSS address, BSEL6 in an information response also contains two singlebit fields that indicate a read or a read and clear TSS/GSS. These information response fields are
described in detail below.

TSS/GSS address

This field (BSELS®6, bits zero to four) contains the octal address of the
TSS/GSS location from which the data in BSEL4 and BSELS is read.

Read TSS/GSS

When set, this bit (BSEL6, bit five) designates that BSEL4 and

BSELS5 contain the data requested by the read TSS/GSS control command.
|
-

3-26

Read and clear TSS/GSS

When set, this bit (BSELS6, bit six) designates that BSEL4 and BSELS

contain the data requested by the read and clear TSS/GSS control
command. In addition to reading the requested TSS/GSS location, the
DMV11 also clears that location.

NOTE
The TSS/GSS locations accessible for writing,
reading, and reading and clearing, are listed and described in Table 3-5.

7
7

6
.

5
T

BSEL3

]

BSEL5

’

|
156

.
14

.
13

|
10

.
9

IEI | BSELO

]

BSEL4

DATA

CLR

1SS

READ
TSS

|BSEL2

CMD TYPE CODE | gg2

READ/

)

—t
NOT USED

RDI

DATA

BSEL7

IEO

RDO
|

’

RQl

TRIBUTARY ADDRESS
|

SEL4

+—t—+—

RETURN KEY OR
TSS ADDRESS
|

BSELS
SEL6
0
MK-2391

Information Response Format

Figure 3-10

Table 3-8
Octal
Code

Name

Return Keys for Information Response

Description

Return Modem | This code indicates that BSEL4 and BSELS contain the requested modem
Status
status (see Appendix C).

Buffer Return
Complete

This code indicates that the process of returning all buffers is completed.

Buffers are returned to the user program under the following circumstances:

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

Theuserprogram issued a control command containing the request key
request halt state causing all private buffers assigned to the tributary
designated by the address in BSEL3 to be returned.

The user program issued a global control command (BSEL3 = zero)
containing the request key request halt state. This causes all unused
common pool buffers to be returned.

rJfl

Information responses containing the return key code for buffer return
complete always have zeros in BSEL4 and BSELS.

3-27

CHAPTER 4
PROGRAMMING TECHNIQUES

4.1

INTRODUCTION

Proper design of user programs for operation with DMV11-based multipoint and point-to-point networks requires that consideration be given to a number of programming topics. This chapter discusses
the following programming topics.

These programming topics deal with interfacing the user program and DMV11 microcode by using the
DMV11 command/response structure.

TM~ ",

Command discipline and handshaking;
DMV11 start-up; Criteria for determining user-defined parameters;
Error counter access;
Error recovery procedures;
Booting a remote station.

4.2

COMMAND/RESPONSE DISCIPLINE AND HANDSHAKING

The command/response interface between a DMV11 and the user program is accompllshed through
the DMVI11 CSRs that are addressed through the CPU 1/0 page.

Since the DMV11 runs in a multiprocessing mode with the associated CPU, the passing of commands
and responses through this interface must be highly disciplined to eliminate the possibility of a race
condition (the user program setting bits in the CSRs after the microprocessor has read those bits). This
interface discipline requires that the user program follow two separate procedures; one for issuing commands, and one for retrieving responses.

Figure 4-1 illustrates the control bits involved in CSR interface discipline. These bits are located in the
DMV11 CSRs BSELO and BSEL2. Examination of Figure 4-1 shows that BSELO contains two control
bits named interrupt enablein (IEI) and 1nterrupt enable out (IEO), bits zero and four respectively
These bits when set, serve to enable the mlcroprocessor to interrupt themain CPU under two circumstances:

-1.

When the CSRs become available for theissuing of a command after access is requested by

| 2

When the microcode has a response to be retrieved from the CSRs by the user program (IEO

the user program for that purpose (IEI and RDI).

and RDO).

NOTE

Interrupt mode must be used, otherwise, receive
overruns and transmit underruns are very probable.

4-1

NOTE
It is recommended that the interrupt enable (IEI
and IEQ) bits be set when interfacing with the
CSRs. It is imperative that they be set when
operating at 56K b/s because the microprocessor
is halted momentarily on every access to the
CSRs.

“The procedures descrlbed below defmmg CSR interface discipline are based on operatlon in the inter-

rupt mode. As a consequence, the IEI and IEO bits should be set by the user program prior to using the
CSR interface.

Ral

.‘

RDO

- |EO

RDI

IEl

| BSELO

BSEL2
MK-2380

Figure 4-1

4.2.1

CSR Interface Control Bits

Command Discipline

At start-up time, before the user program can execute any command it must initialize the DMV11.
This1s accomplished by the program settmg the master clear bitin BSELI and wa1t1ng for the DMV11
“to set the run bit.
|
Once the DMV11 is initialized, commands may be issued. All commands are 1ssued by the user program in two successive steps. The first step requests the use of the data port. The second step identifies
the command type and the data port information for the appropriate command, and notifies the
DMVI11 that the commandis in the CSRs. The specific content of each data port is further defined
under each command descriptionin Sectlon 3.3. The handshakmg procedure for mput commandsis as
follows (see Figure 4-2).
:
|
The user program requests the use of the data port to issue a command by settmg request in
(RQI) bit 7 of BSELO. The user should. also set bit 0 of BSELO, interrupt enablein (IEI), at
the same time (using the same 1nstructlon) to allow the DMV11 to 1nterrupt the CPU when
the data port is available. An interrupt is generated to XX0 when RDI is asserted by the
DMV11.
|
NOTE
" The 22-bit modeis used when the software supports
22-address bit buffers. A “one” in bit 3 of BSEL2
indicates to the DMV11 that the software supports
the changein the command format required to support 22-address bits.
‘When the data port is available, the DMV11 informs the user by settmg readyin (blt 4 of
BSEL2) and generating an interrupt to vector XXO.

On detecting RDI bit set, the user can: 1) load the appropriate mformatlon into SEL4 and
SEL6, and 2) load the input command code into bit 0-2 of BSEL?2.
If a single commandis to be issued, the user program clears RQI. If a series of commands

are to be issued, RQI can remain set until just prior to the loading of the last command into
the CSRs. By leaving RQI set while issuing a series of commands, the user program is as-

sured of having access to the CSRs after the next response.

4-2

|
CHAPTER 4
PROGRAMMING TECHNIQUES

4.1

INTRODUCTION

Command discipline and handshaking;
DMV11 start-up;-

These programming topics deal with interfacing the user program and DMV 11 microcode by using the
DMYVI11 command/response structure.

Criteria for determining user-defined parameters;
Error counter access;
Error recovery procedures;
Booting a remote station.

4.2 COMMAND/RESPONSE DISCIPLINE AND HANDSHAKING
The command/response interface between a DMV11 and the user program is accompllshed through
the DMV11 CSRs that are addressed through the CPU I/0 page.

Figure 4-1 illustrates the control bits involved in CSR interface discipline. These bits are located in the
DMV11 CSRs BSELO and BSEL2. Examination of Figure 4-1 shows that BSELO contains two control
bits named interrupt enablein (IEI) and interrupt enable out (IEO), bits zero and four respectively
These bits when set, serve to enable the mlcroprocessor to interrupt the main CPU under two circumstances:

When the CSRs become available for theissuing of a command after access is requested by
the user program for that purpose (IEI and RDI).

When the microcode has a response to be retrieved from the CSRs by the user program (IEO
and RDO).

NOTE

Interrupt mode must be used, otherwise, receive
overruns and transmit underruns are very probable.

4-1

NOTE
It is recommended that the interrupt enable (IEI
and IEQO) bits be set when interfacing with the
CSRs. It is imperative that they be set when
operating at S6K b/s because the microprocessor
is halted momentarily on every access to the

CSRs.

‘The procedures descrrbed below defining CSR interface discipline are based on operatlon in the inter-

rupt mode. As a consequence, the IEI and IEO bits should be set by the user program prior to using the
CSR interface.

~§ RQb

1EO |

RDO

RDI

IEl

§ BSELO

BSEL2
MK-2380

Figure 4-1

4.2.1

CSR Interface Control Bits

Command Discipline

At start-up time, before the user program can execute any command, it must initialize the DMVII.
This1s accomphshed by the program settmg the master clear bitin BSELI and waiting for the DMV11
to set the run bit.

Once the DMV11 is 1n1t1ahzed commands may be issued. All commands are 1ssued by the user pro—
gram in two successive steps. The first step requests the use of the data port. The second step identifies
the command type and the data port information for the appropriate command, and notifies the
DMV11 that the commandis in the CSRs. The specific content of each data port is further defined
under each command descriptionin Sectron 3.3. The handshakmg procedure for 1nput commandsis as
follows (see Flgure 4-2).
:
|

The user program requests the use of the data port to issue a command by setting request in
(RQI) bit 7 of BSELO. The user should also set bit 0 of BSELO, interrupt enablein (IEI), at
the same time (using the same mstructlon) to allow the DMV11 to mterrupt the CPU when
the data port is available. An interrupt is generated to XX0 when RDI is asserted by the
DMV11.
|

NOTE

- The 22-bit mode is used when the software supports
22-address bit buffers. A “one” in bit 3 of BSEL2
“indicates to the DMV11 that the software supports

the change in the command format required to support 22-address bits.

When the data port is available, the DMV11 informs the user by setting ready in (bit 4 of

BSEL?2) and generating an interrupt to vector XXO0.

On detecting RDI bit set, the user can: 1) load the appropriate 1nformatlon into SEL4 and
SEL6, and 2) load the input command code into bit 0-2 of BSEL?2.

If a single commandis to be issued, the user program clears RQI. If a series of commands
are to be issued, RQI can remain set until just prior to the loading of the last command into
the CSRs. By leaving RQI set while issuing a series of commands, the user program is assured of having access to the CSRs after the next response.
.

4-2

4.2.2
Retrieving Responses
The DMV11 issues responses in two steps. If RDI is not set, the data pertinent to the responses being
issued is loaded into BSEL3, SEL4, and SEL6. Once this is complete, the DMV11 sets the ready out
(RDO) bit and the command code in BSEL2. An interrupt through vector XX4 is generated if the IEO
bit is set. Generally, processing an output response involves the following steps:

The user program checks for RDO set. This is done by waiting for an interrupt.

To use the output interrupt capability, the user program must set the output interrupt enable
bit in BSELO immediately after it detects that the run bit has been set following master
clear.

When an RDO set condition is detected, the user program should move the contents of
SEL2, SEL4, and SEL6 into a working storage area and clear RDO in BSEL2 as soon as
possible. When RDO is cleared, the data port (SEL4 and SEL6) is released to the DMV11
for more input or output processing.

4.2.3 CSR Interface Interactions
|
User-program access to the CSRs is under complete control of the DMV11 microcode. Access to the
CSRs is granted upon request when the user program has a command to issue, or when the microcode has a response for the user program. Figure 4-2 illustrates the nature of the access window available to
the user program under interrupt control, when issuing a command, or retrieving a response. As previously indicated, the user program requests use of the CSRs for the purpose of issuing a command by

setting RQI. The DMVI11 microcode makes the CSRs available for the issuing of a command, only
when it is not using them, by setting RDI and interrupting the user program through the floating vector
XXO0. As a result, the time period between the setting of RQI by the user program and the granting of
the CSRs through an interrupt is indeterminate.

When commands are being issued from a queue, and the last command has been issued with the user
program still having ownership of the CSRs (RDI is set), there is a mechanism for returning the CSRs

to the DMV11. In such cases, the user program can issue a control command with the code for no
request in the request key field, thereby, signalling the microcode to ignore the contents of the CSRs.
In this way, the possible reading of erroneous data by the DMV11 microcode is avoided.
If the user program is to issue a single command, RQI should be cleared prior to issuing the command
to the CSRs, 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 circumstance, when a command is issued and RDI is cleared, the
microcode relinquishes the CSRs to the user program as soon as they become available. The time period between clearing RDO upon completing one command, and an interrupt 1n1t1at1ng the next command, is also indeterminate.

When the DMVI11 has a response to be passed to the user program, it sets RDO then interrupts the

main CPU through the floating vector XX4. At this point, the user program has ownership 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

retrieving responses should limit the CSR access time required to load a command or retrieve a re-

sponse.

4.3

DMVI11 START-UP

Starting a DMV 11 requires that the user program perform a series of steps to; 1) configure the DMV11
to operate within the context of the associated network, 2) establish user-defined parameters, and 3)
initiate protocol operations at the DMV11.

4-3

USSR

NO FURTHER COMMANDS

] 70 1ssUE

%*DMV1 1
-

USERCLEARSRQl

‘

SETS
.

% USER CLEARS RDI TO

D
NOTIEY DMV11 THAT COMMAN
A

[ USER HAS TOTAL CONTROL OF CSRs | jias BEEN ISSUED
s

TO ISSUE COMMAND

(IE])

|
o
preee—

%DMV 1
| SETS

USER HAS TOTAL CONTROL OF CSRs

RDO
oz

RETRIEVE

—.

* USER CLEARS RDO TO NOTIFY DMV11 THAT

| RESPONSE HAS BEEN RETRIEVED

RESPONSE

OF CSRs
¥ DMV11 HAS OWNERSHIP
"MK-2514

Figure 4-2

CSR Access Window
,

Configuratnon Procedure

4. 3 1

The sequence to configure a DMVll control and trrbutary statron for network operat1on 1s formed by
the following steps
1.

Set the master clear b1t and wait for the run bit to set. (See Sectlon 3.3. l)
When the run bit1s set, read BSEL4 and BSEL6

If the contents of BSEL4 equals octal 33, and the contents of BSEL6 equals octal 305, the
start-up diagnostics have successfully executed and the DMV11 is running. Any other value
in BSEL6 indicates that an error condition was detected by the start-up diagnostics. The val‘ues and meanmgs of these diagnostic error codes are hstedin Table 4-1.

If the DMVll operat1onal modeis software selectable set the mode for that device by is‘suing the appropriate mode definition command (see Section 3.3.2). If the mode for the
DMV11 is selected by internal switches, this step can be ignored.

4.3.2

Specnfymg User-Defined Parameters

After a DMV11 is configured, the user parameters are specified. User-defined parameters include parameters used by the polling algorithm, and parameters specific to protocol operatron In addition,
those user parameters that are specific to the operation of tributaries, are storedin the tributary status
slot (TSS) associated with each tributary. Parameters that are pertinent to overall system operat1on are
storedin the global status slot (GSS) for the control or tributary station.

4-4

—

Table 4-1

Diagnostic Error Codes

BSELG6

BSEL4

Description

101

N/A

Branch test has failed and the microcode is spinning in a }loop.

102

~N/A

103

N/A

| 6502 intérnal resister test has failed
and the microcode
is spinning
in a loop.
Load and store instructions test has failed and the microcode is spinning in a

loop.

104

N/A

105

N/A

Compare instructions test has failed and t»he microcode is spinning in a loop.

Increment and decrement inStruCtions test has failed and the microcode is
‘spinning in a loop.

106

N/A

Shift and rotate instructions test has failed and the microcode is spinning in
a loop.

107

N/A

110

N/A

Logic instructions test has failed and the microcode is spinning in a loop.
Add with carry, subtract with carry, set and clear decimal mode instruc-

tions test has failed and the microcode is spinning in a loop.

111

N/A

Stack push and pull instructions test has failed and the microcode is spinning in a loop.

112

N/A

113

N/A

Subroutine instructions test has failed and.the microcode is spinning in a

loop.

Ram scratchpad, CSR, and NPR address resisters addressing test has failed
and the microcode is spinning in a loop.

114

N/A

Ram scratchpad, CSR, and NPR address resisters data test has failed and

the microcode is spinning in a loop.

115

N/A

116

N/A

True interrupt test has failed and the microcode is spinning in a loop.
Ram data and addressing test has failed and the microcode is spinning in a

loop.
117

N/A

Ram alternating data test has failed and the microcode is spinning in a loop.

120

N/A

Indexed indirect addressing mode instruction test has failed and the microcode is spinning in a loop.

121

N/A

Line unit message test has failed and the microcode is spinning in a loop.

305

The microdiagnostics have completed without errors.

4-5

As described in Section 3.3.3, user parameters are specified through control commands configured to
address a TSS or a GSS. These control commands write the data contained in BSEL4 and BSELS into
the locations specified by bits 0-4 of BSEL6 (TSS or GSS address). In establishing polling and protocol
parameters, the user program has the option of accepting the default for a parameter or setting the
parameter to some predetermmed value. Chapter 5 details the criteria to be usedin determmrng optimum values for the various polling parameters. Criteria for determining the remammg parameters,
which generally concern the operation of the communications link, are presentedin Sectlon 4.6.
|

NOTE

Although the majority of user-defined parameters

-are 16-bits in length, some are single byte parame-

ters. If the user program wishes to change one of the
single-byte parameters and accept the default for the
other, both parameters must be written. Thisis necessary because both TSS and GSS user-defined parameters are written on 2-byte boundaries.

The process of establishing user-defined parameters is presented as two separate steps:
1.

Specifying user-defined parameters for control and tributary station TSS structures.

Specifying user-defined parameters for control and tributary station GSS structures.

4.3.2.1 Specifying TSS Parameters — TSS parameters that can be spee1f1ed by the user program are listed in Table 4-2 by name, BSEL3 address, size, and default value. A functional summary of each

parameter is also grven The actual order of setting TSS parameters through appropriately configured
control commands is arbitrary. The complete command series to specify these parameters for TSS
structures at a multipoint control stat1on is l1sted below |

Issue a series of control commands to set the value for the transmit delay timer. This1s re-~

ferred to as the preset value.

Issue a series of control eommands to establish the pollmg parameters Q and R for the three

polling levels.

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

Issue a series of control commands to specify values for the maximum transm1tted 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.

4-6

<«

Table 4-2

User-Defined TSS Parameters

TSS Addr

Name

(Octal)
BSEL6

Size

230

Default
(Octal)

(Bits)

Description

XDT
PRESET

Preset value for

the transmit delay
%

timer. This parameter provides a
fixed delay between
transmission of data
and maintenance

231

Q (Unresp)

233

232

377
The initial value of

Q (Active)
Q (Inactive)

OO0 OO

messages. The default value of 0 =
no delay.

polling urgency (U)

for the tributary:
The TSS for a tributary must be assign-

ed a Q value for
each of the three
activity levels;
active, inactive,

and unresponsive.
This parameter is
applicable only to
TSS structures at
the control station.

232

The rate (R) by

OO0

231

R (Active)
R (Inactive)
R (Unresp)

100

which the urgency

(U) is increased for
the tributary.

233

The TSS for the tributary must be
assigned an R value
for each of the
three activity
levels; active, inactive, and unresponsive. Both the
Q and R for a given
tributary are established through a
single control command. Therefore, if

4-7

Table 4-2

TSS Addr
Name

(Octal)
BSEL®6

User-Defined TSS Parameters (Cont)

Size |
(Bits) |

Default
(Octal)

Description
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.
|

NDM-INACT

234

Number of no data
messages required to
go inactive: This is
the number of consecutive polls to be
made (without receiving a data message) before changing the activity
level of that tributary from active to
inactive. This para-

meter is applicable
only to the TSS
structures at the
- control station.
TO-UNRSP

234

Number of timeouts
to go unresponsive:
The number of consecutive polls of a
tributary (without
response) before
changing the activity level of that
tributary from active or inactive to
unresponsive. Both
the NDM-INACT and
TO-UNRSP counts are
established through
a single control

Table 4-2

Name

TSS Addr
(Octal)
BSEL6

User-Defined
TSS Parameters (Cont)

Default

Size
(Bits)

(Octal)

Description

command. Therefore,
if one parameter is
tobesettoa
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.
TO-DEAD

235

MXMC

235

Number of timeouts
to go dead: The number of consecutive
polls of an unresponsive tributary
(consecutive selection timeouts) before the activity
level of that tributary is changed from
unresponsive to
dead. This parameter is applicable
only to the TSS
structures at the
multipoint control
station.

Maximum transmitted
message count: This
parameter is a count
of the maximum number of abutting data
messages to be
transmitted by a
station before it
deselects itself.
This count applies

to the TSS structures at both control and tributary

4-9

Table 4-2 User-Defined TSS Parameters (Cont)

Name

TSS Addr
(Octal)

BSEL6

Size

Default

(Bits) | (Octal)

| Description
stations in multipoint networks as
well as point-topoint stations. Both
TO-DEAD and MXMC

for a given tribut-

SEL

TIMER
o

236

3 (seconds)

(5670 Octal)

ary are established
through a single
control command.
Therefore, 1f 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 pointto-point stations,
the polling parameter TO-DEAD is 1gnored.

Selection interval
timer: This timer
is started when a

message 1s transmitted with the
select flag set, and
halted when a valid
reply is received or
the line is resynchronized. The selection timer is used
as a reply timer for
full-duplex pointto-point networks
(see Section 4.4.1).
It is used as the
select timer at multipoint. control sta‘tions, and in halfduplex point-to-

point networks. This

Table 4-2

User-Defined TSS Parameters (Cont)

TSS Addr

- Name

(Octal)

Size

Default

BSEL6

(Bits)

(Octal)

Description
counter counts in

milliseconds (ms)
from 1 to 65,535 ms.
&

BAB
TIMER

237

16
~

6 (seconds)
(13560 Octal)

Babbling tributary
timer: This timer

is used to detect

a babbling tributary (see Section
442).Ina
multipoint network
this parameter is
applicable only to
the control station.
However, this parameter is applicable
to both stations in
point-to-point networks operating
half-duplex.

4.3.2.2 Specifying GSS Parameters - As previously indicated, when one or more tributary addresses
are established at a DMV11, the DMV11’s 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 4-3
by name, BSEL3 address, size, and default value. A functional summary of each parameter is also
given. If a value is not specified for a parameter the microcode uses the default value. Specifying a

user-defined GSS parameter requires that BSEL3in the pertinent control command contain zero. The
control commands necessary to specify GSS parameters for a multidrop station are listed below:

A control command to specify the number of sync-characters (NUM SYNC) that are to pre-

cede nonabuttmg transmlt messages.

A control command to set the streaming tributary timer.

Three control commands to establish values for the global polling parameters delta time

(DELTA T), dead tributary (DEAD T), and poll delay, respectively.

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 being accessed in BSEL3. Similarly, each control command specifying a GSS parameter must have BSEL3 set to zero. Since the prior steps have covered user-defined TSS and GSS

4-11

parameters (see Table 4-2 and 4-3) at the control station, the two steps listed below complete the parameter-specifying process at the tributary stations:
1.

to set the maximum transmitted
Issue a series of control commands at each tributary station
message count (MXMC) for each active TSS in a tributary station TSS structure. Note that
the 8-bit value for MXMC must be placed in BSELS5 of each command (the tributary station
microcode ignores BSEL4 in these commands). The procedure for determining an optimum
value of this parameter for tributary stations is the same one used for control stations.

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 nonabutted transmit messages. The procedure for determining an optimum value of this parameter for tributary stations is the same

Table 4-3

Name

GSS Addr

Default

(Octal)

(Bits)

| BSEL6

NUM SYNC | 233
|

User-Defined GSS Parameters

Size

(Octal)

’

10 (low

speed)

15 (high
speed)

Description

This global value speci-

Number of sync-characters:

fies the number of synccharacters that are to
precede nonabutting transmitted messages. This
parameter applies to all
stations. Low speed is
defined as less than 56K
is 56K
b/s, and high speed

234
|

|
16
|
|

Streaming tributary timer
preset: This timer is used
‘to detect a streaming tri-

6(sec.)
(13560
Octal)

STREAM
TRIB
-

one used for control stations.

butary (see Section 4.4.3)
and Table 3-7). In a mul-

tipoint network, this
parameter is applicable
only to the control sta-

tion. However, in point-

to-point networks, this
parameter is applicable to
both stations.

DELTAT

235

200 (ms)

Delta time: This is the

(310
Octal)

polling algorithm update
increment. This global
parameter, which-is ap-

4-12

- plicable only to multipoint control stations,
by the polling
is used
|

Table 4-3 User-Defined GSS Parameters (Cont)
GSS Addr

(Octal)
BSEL6

Name

Size
(Bits)

Default
(Octal)

Description
algorithm to calculate
polling urgency (see
- Section 5.2.1). The default value of 200 ms is
the minimum value for this
parameter. DELTA T is also

used as the interval for
updating the transmlt
delay timer.
DEADT

236

10 (sec.)

Dead timer: This is the

(17500

interval between polls for

Octal)

dead tributaries. This
global parameter applies
only to multipoint control
stations.

POLL
DELAY

237
|

0 (no
delay)
~

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 occurs immediately
following the current
poll.

4.3.3 Protocol Operation
At this point the DMVI11 has been conflgured for operatlon in the associated network andis ready for
protocol operation. The steps required to initiate protocol operation are:
1.

Place established tributariesin the ISTRT state byissuing one control command containing
the request key. Request ISTRT state for each tributary address.

If the DMV11 is a station in a point-to-point network, one control command must be issued
containing the request ISTRT state request key with a tributary address of one in BSEL3.

The DMVI11 confirms that the'protocol is operational at each tributary by issuing a control
response (one for each control command issued) contamlng the protocol event code for
DDCMP run state entered.

4.4 CRITERIA FOR DETERMINING COMMUNICATIONS LINK PARAMETERS
User-defined TSS and GSS parameters fall into two categories: polling parameters that provide for
user-program control over the dynamic activity of the polling algorithm, and communications link pa-

4-13

rameters that provide the user program with the ability to regulate data traffic over the physical communications line. Referring to Tables 4-2 and 4-3, the commumcat1ons lmk parameters include:

‘Selection interval timer,
Number of sync-characters,
Babbling tributary timer,

Maximum transmitted message count,
Streaming tributary timer,
Transmit delay timer.

Values for the selection interval timer and the number of sync-characters are interrelated, as are values

for the babbling tr1butary timer and the maximum transm1tted message count. These interrelated parameters are describedin Sections 4.4.1 and 4.4.2.
4,4.1

Setting the Selectionlnterval' Timer

The function performed by the selection interval timer at a DMV11 depends on the mode selected for
that DMV11. In full-duplex point-to-point networks, this timer is used as a reply timer for the purpose

of message accountabllrty This timer serves as a selection interval timer when the mode for the associated DMV11 is one of the following:
e
e
e

A full-duplex control station;
A half-duplex control station;
A half-duplex point-to-point station.

In this capacity, it performs the link management function and provides 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 actively connected to the same physical com-

munications link. This applies to half- and full-duplex multipoint networks as well as half-duplex point-

to-point links. On half-duplex links, only one transmitter can be active at any time, and on full-duplex
links, only one slave transmitter can be active at a time.
|
A station on such links can transmit when it is selected or granted ownership of 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

a select flag means that the transmitting station ceases transmitting after the current message is sent.

A selection timer detects the loss of a select flag by timing the interval required to receive the longest
message from a station. A timer is started when a station is selected and reset when valid messages are
received from that station. When the timer interval is exceeded at the sending station (a message was
not received during the perlod of the timer) 1t is assumed that messages with the select flag were either
transmitted or receivedin error.
.

At this point, the statron that originally sent the messages with the select flag set assumes ownership of
the link. This station resumes 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 avoid possible deadlock race conditions. For both multipoint control stations
and half-duplex point-to-pomt stations, the criteria for determining the value for a select timer includes
such factors as:
Maximum message length;
Number of sync-characters
Line speed;
Line turnaround time;
~Message processing delays.

4-14

As indicated in Table 4-3, the GSS parameter number of sync-characters has two defaults; one for lowspeed operation (10), and one for high-speed operation (15). The operational speed range of a DMV11
is specified by the line unit low-speed/high-speed switch which is placed in the appropriate position
when the DMV11 is hardware configured (see Section 2.5). In the low-speed position the DMV11 can
operate at line speeds up to 19.2K b/s, and in the high-speed position the device line speed is 56K b/s.
It is recommended that the default for the appropriate line speed range be taken for this parameter.

Some recommended values for a selection interval timer are given in Table 4-4. The calculations shown

in Table 4-4 include the following overhead factors:

256 bytes of data
28 bytes of header, sync, and pad characters

Total

284 bytes

The formula used to derive the values listed in Table 4-4 is:
8 bits per byte

284 bytes per message + RTS/CTS delay = Timer value (in seconds)
|

baud rate (bits
per second)

NOTE

Most modems include an RTS/CTS delay that must
be included in the calculation of the value for the selection interval timer. When operating with an external (EIA) modem, the typical delay used is 150 ms.
The delay used when operating with the integral
modem is 100 us.

Table 4-4 Recommended Selection Interval Timer Values

Bits Per Second

Calculated Timer Value for a 256 Byte Message

4.8K

473 ms + 150 ms = 623 ms (700 ms)

- 9.6K

236 ms + 150 ms = 386 ms (400 ms)

56K

40.5 ms + 0.1 ms = 40.6 ms (50 ms)

The values listed in Table 4-4 represent absolute minimums. In most cases, specific applications require
additional delay time over these values to prevent a timeout during reception of a valid message. Requirement for additional delay time can be caused by processing delays that occur when receiving from
a non-DMV 11 device, or by line delays encountered when dealing with satellite networks. When determining this value, keep in mind that it represents the time 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 before taking error recovery actions. This timeout is
necessary to recover from outages and the distortion of messages by the link. This timeout also prevents
the protocol from being deadlocked.

4-15

4.4.2 Settmg the Babblmg Tnbutary Timer
This user parameter is applicable to half-duplex and full- duplex multrdrop network control stations. A
babbling tributary is a tributary that continues to transmit valid DDCMP messages after a programmable timeout has expired, thereby, denying equal access to other nodes. This situation is controlled by
the babbling tributary timer which monitors the total time period a tributary continuously transmits
without relinquishing the communications line. When this period exceeds the timeout period of the
babbling tributary timer, the user program is notified through a control response. The control response
contains the code for a babbling tributary along with the identity of the offending tributary. When a
babbling tributary is detected, the control station takes no action beyond this notification.

A major consideration in determining a value for the tributary timer is the total time interval that a
given tributary requires to end a selection interval. Determining the value for this timer is similar to
that for the selection interval timer because the same range of factors are used as criteria for calculating the value. The main difference in the two determinations is that the total number of message
bytes should be used in babbling trlbutary timer parameter calculations rather than the number of
bytesin the longest message.

A value for the maximum transmitted message count parameter must also be considered in conjunction

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 selection period. This

is done by forcing the select flag when the count of messages received from a tributary equals the value
“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 case, the period established for the timeout of a babbling tributary timer should be long enough
to ensure that timer expiration definitely indicates an error condition. In addition, the parameter assigned to the maximum transmitted message count should also be considered when establishing the period of the babbling trrbutary counter.

4.4.3

Setting the Streaming Tributary’ Timer

A streaming tributary is a tributary station on a multipoint line (or an associated point-to-point station)
that continues to assert the carrier signal on the link after it has relinquished ownership of the link. In
normal operation, ownership of the linkis returned to the control station when it receives a select flag or
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 trrbutary station, a defective point-to-point station, or a malfunctioning modem.

The streammg tributary is started when ownershrp of the link1s granted to the control station by the

remote station, and stopped when the carrier is dropped by that station. When a streammg trrbutaryis
detected, through expiration of the streaming tributary timer, the user program is notifiedin the same
manner as with a babbling tributary. The control station does not transmit until the carrier is dropped.

Determination of a value for the streaming tributary timer requires consideration.of suoh factors as

settling time of the communications line and modem delays. As with determining periods for the selec-

tion interval timer and the babbling tributary timer, the period specified for this timer should be long
enough to preclude premature expiration of the timer. For most network applrcatrons the default of one
secondis sufficient.

4-16

/fl—w\.\\

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 trmer As shown in Table 4-2, the
default value for both cases is three seconds

4.5 ERROR COUNTER ACCESS
The DMV11 is equipped with a large compliment of error counters de&gned to isolate a wide range of
error conditions. The TSS for each established tributary contains seven error counters along with three
statistical counters that provide background information for error analysis. In addition, the GSS at each
station contains four error counters that tabulate errors that are global to the station. The three TSS
statistical counters are 16 bitsin length, and the threshold error counters are three bitsin length. The
remaining TSS/GSS error counters are eight bits long.
4.5.1
Reading the Counters
Both TSS and GSS counters are accessed through an approprlate control command with the content of

the requested counter or counters being returned through an information response (see Sections 3.3 and
3.4). Through the control command, the user program has the option of reading, or reading and clearing
the counters. When doing error analysis it is recommended that a user program read and clear these
counters to assure a zero-base for subsequent sampling of the counters. If copies of the counters are
being maintained in main CPU 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 so
that access to these counters is restricted to reading

only.

The DMV11 error and statistical counter structure is designed to be complimentary. As an example of

this complimentary structure, consider the data errors inbound counter and the data messages received
counter. The data errors inbound counter tabulates the errors related to the validity of message reception such as block-check errors, whereas, 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
derived from these two counters.
4.5.2 Counter Skew
When performing error analysis, there is a potential of skew between counts due to read time lags and
the requirement that counters be read one at a time. The probability 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 discrepancy between the number of selection intervals and the

number of selection timeouts. A skew could result from additional selection timeouts occurring while
the counters are being read.
If circumstances require 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 due to read time lags.

4.6 ERROR RECOVERY PROCEDURES
|
|
Within a DMV11-based network, there are three basic levels of error recovery involving the user program:

Procedural violations where only the user program is notified.

Recovery from errors requiring protocol shutdown initiated by the user program.

Fatal errors resulting in system shutdown with minimal notice to the user program.

Referring to Table 3-7, procedural error codes from 100 to 140 are reported to the user program with
no recovery required. The remaining
two procedural errors (codes 300 and 302) involve error recovery
levels two and three respectively. All network errors require recovery through protocol shutdown, and
the control response (queue overflow) could result in network shutdown.

4-17

Recovery from Network Errors

4.6.1

In all cases, recovery from network errors requires-that the protocol be halted at the trlbutary or stat1on
recording the error. Two similar but separate procedures are recommended for recovery from threshold
errors, and babblrng and streammg tributary errors. These recovery procedures are descrlbed below.

4 6 1. 1 Recovery from Threshold Errors —~ DMVll threshold errors are detailedin Sectlon 5.3.3. The
recommended recovery procedure to be initiated by the user program at the station recording the errors
is presented below:

1. Halt the protocol (see Table 3-6).
2. Read the error counters to determme the nature and cause of the threshold error cond1tlon If
the error results from a shortage of receive buffers, correct the condition. If the transmit or
selectlon threshold1S bemg exceeded check the operatlonal condition of the remote station.
3.

When the cond1t1ons causmg the eITors have been el1m1nated restart the protocol (see Section 4.3.3).

4.6.1.2 Recovery from Babbling and Streammg Trlbutary Errors — Babblmg or streaming tributary er-.
rors are created when their respective timers are exceeded. Therefore, a timeout can result from an
actual error condition, or because the period of the timer is too short for the type of message activity on
the line (see Sections 4.4.2 and 4.4.3). A suggested recovery procedure to be used when encountermg
|

1. Halt the protocol.
2. Check the»Value of timer parameters an‘d increase ifthe value is not appropriate.

\\~

these conditionsis:

Restart the protocol (see Section ‘4.3.3).

4. If this errOrcondition p’ersists, reconfigure the station as specified by Section "4.3.1. l
5. When the cause of the trmeout or1g1nates at the remote station, action must be taken at the
remote station to ascertain and correct the fault. The local station is at fault only if the Values

'of the timer parameters are inappropriate.
4.6.2

Recovery from Procedural Errors

The three procedural Errors that requlre a recovery procedure are:
1.
2.
3.

Nonexistent memory error.
Buffer too small error.
Queue overflow error.

The recovery procedure for each of these errors is detailed in Sections 4.6.2.1 through 4.6.2.3.
4.6.2.1 Recovery from a Nonexistent Memory Error — Nonexistent memory errors occur when the
DMV 11 tries to access an allocated receive or transmit buffer having an invalid bus address. When this
error is detected, the DMV11 posts a control response to the user program containing the invalid address (see Section 3.4.2). It is up to the user program to determine whether the nonexistent address
concerns a transmit or receive buffer,
|
NOTE

- Depending on microcode processing circumstances,

- the nonexistent memory address returned to the user
program could have been mcremented to the next sequential location.

4-18

The suggested recovery procedure for this error is as follows:
Halt the protocol for the tributary or station recording this error to initiate return of all outstanding buffers.

If the error concerns a buffer from the common pool, the user program shouldissue the global halt command to initiate return of all outstanding receive buffers from the common pool.

Restart the protocol and reallocate buffers as necessary.

Persistent recurrence of this error indicates a possible main CPU or DMV11 malfunction.

NOTE
If the network line speed is 56K b /s, the requests for
retransmission generated by a nonexistent memory

address can result in the overflow of the DMV11 response queue causing a fatal system error (see Sec-

tion 4.6.2.3).
4.6.2.2
Recovery from a Receive Buffer Too Small Error — When the DMV11 receives a message, it
first checks for the availability of a buffer from the common buffer pool linked list, and if one is avail-

able, it uses that buffer. If the common buffer pool is empty or not enabled, the private buffer linked
listis checked. If a prlvate bufferis not available, the receiving station NAKs the incoming message.
The steps taken by the DMV11 microcodein this process are listed below.
Is the message number in sequence? Yes, continue; No, ignore message.

Is the common buffer pool enabled? Yes, continue; No, go to Step 6.
~Is the common buffer pool quota = 0?7 Yes, go to Step 6; No, continue.

‘Is a common pool buffer available? Yes;’ continue; No, ge 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 unavailable.

Is private buffer too small? Yes, send NAK — buffer too small; No, use this buffer.
[s private buffer available? Yes go to Step 7; No, send NAK — buffer too small.

NOTE

The DMV11 does not scan the common pool or pri-

vate linked list structures looking for a buffer of sufficient size. Rather, 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 is

dependent on whether the allocated bufferis from the common pool or is a private buffer. The applicable recovery procedures are explained below.

Common pool buffer too small
1.

Assign a private buffer of sufficient size to the receiving tributary through a buffer ad-

dress/character count command (see Section 3.3.4).

4-19

Both private and common pool buffers too small

1. = Halt the protocol for the offending trlbutary to initiate return of all outstandmg private buffers.

Restart the protocol

Assign a private buffer of sufficient size to the receiving trlbutary through a buffer ad|

dress/character count command (see Section 3.3.4).

Private buffer too small, and common pool not enabled

4.6.2.3

If buffers from the common pool are available to other tributaries, and are of sufficient size,

enable common pool buffers for this tributary (see Section 3.3.4).

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

above.

Recovery from a Queue Overflow Error — This error is alwaysfatal to the DMVll recordmg

the error since it forces automatic shutdown of the device. The basic cause of this error is the in-

when the internal response queue overflows because the DMV11 generated responses faster than the
user program could retrieve responses from the queue. This error can also occur if an inordinate number of receive buffers have been allocated. One cause of response queue overflowis the occurrence of
repetitive nonexistent memory errors in hlgh-speed networks (see Section 4.6.2.1).

When this error occurs, the DMV11 posts the most current entry inthe response queue to the user
program. The user program then has three seconds after being 1nterrupted to retrieve the response. If it
is retrieved during this three second window, the next response is posted. As long as the user program
retrieves each response within this window, the process continues until the internal response queue is
empty These responses can then be analyzed to determine the cause of the queue overflow.

After the last response has been posted, or the three second response period has exprred the DMV11

shuts itself down. At this point, returning the DMV11 to operational status requlres that the start-up
procedure be initiated from the beginning (see Section 4.3).

4.7

BOOTING A REMOTE STATION

DMV 11-based networks provide the user program, at the multlpomt control station or point-to-point
station, with the ability to boot the main CPU at a remote station that has been shut down due to power
outage or software malfunction. There are three ways this boot function can be performed:
-

Remote load detect: The control station starts the primary MOP boot procedure for a remote
station.

Power-on boot: The first poll received after power up at the remote station causes the
DMV11 at that station to request that the control station start the MOP boot procedure.

Invoke primary MOP: The user program at the remote station causes the DMV11 to request
that the control station start the primary MOP boot procedure.

4-20

f//*\,\p

availability of link blocks from the free linked list (see Section 5.4.1.1). Typically, this error results

‘NOTE
Control station is either a multipoint control or
point-to-point station that is transmitting (over the
link) the boot or requested program to the remote
station. Remote station is either a multipoint tributary or point-to-point station that is receiving
(over the link) the boot or requested program.
|
NOTE
Power-on boot, remote load detect, and invoke
primary MOP are not mutually exclusive. All three
features could be used in a particular application.
Primary MOP boot procedures require that the DMV11 be switch-configured in the manner 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 presented in the following sections.
4.7.1
Steps Leading to a Remote Load Detect Boot
The steps taking place at the DMV11 remote station and its host CPUin response to an enter MOP
mode message from the control station are:

The DMVI11 NPRs a tight-loop routine into main memory.

The DMVI11 transfers control
to the routine through the power fail /restart vector. This routine inactivates the CPU to prevent any intervention during the NPR process.

The DMVI11 then sends a primary MOP request program message to the control station. The
control station respondsin turn with a primary MOP memory load with transfer address message containing the boot or related program to be loaded into main memory at the remote
station.

The DMVI11 NPRs that program into main memory, then starts executing the program.

At thls point the remote station is operatmg in the manner intended by the down line loaded
program.

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

2.
3.
4.

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

The remote station recogmzes the address and passwordin the message, then inactivates its
host CPU.

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 containing the boot or related program to be loaded into the host CPU at the
- remote station.

4.7.2 Steps Leadmg to a Power-On Boot
When power is restored after a shutdown at a remote station, the DMV11 performs the same steps used
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 an MOP
request program message. The same sequence used in the remote load detect boot procedure is then
followed.
|

4-21

Steps Leadmg to an Invoke Primary MOP Boot

4.7.3

This boot operation is initiated when a user at a remote station sets the boot and master clear bitsin the
DMV11 initialization reglster (see Sectron 3. 3 l) The steps taken by the DMV11 are the same as with
a power-on boot.

DMV11 Switch Settmgs for the Boot Functlons -

4.7.4

At remote stations, in networks supporting the primary MOP boot funot1ons the switches must be configuredin a specific way in order to properly perform the boot funot1ons (see Section 2.5 and Table 26).

NOTE

The swrtch setting procedures described below apply
~ only ‘to trlbutary stations in a multipoint network
and one nodein a pomt-to-pomt network

The unit number (zero or one) of each DMV11 must be appropr1ately set. Th1s number allows the boot

program, once it is loaded into the host CPU, to 1dent1fy the speo1frc DMVll (w1th1n the host’ S floating
address space) performmg the boot.

NOTE

When primary MOP booting is supported in a net- work, the operating mode of each tributary station
eligible for booting must be set in the switches rather
than through the mode definition command.

The operating mode of a DMVI11 is specified by setting the mode enable switch to one (OFF) (switch
number 1 of the boot enable switch pack), and setting switches numbered 6,7, and 8 to the required

operating mode. The settings for these switches are listedin Table 4-5.

4.7.4.1 Switch Settings for the Power-On Boot Function — To enable the power-on boot function at a
remote station, switch number 4 of the boot enable switch pack (power-on boot) must be set to one
(OFF). In addition, the tributary address of this station must be set in the DDCMP address switch
|

pack.

4.7.4.2 Switch Settmgs for the Invoke Primary MOP Boot Functnon. The DMV11 switch settings for
the invoke primary MOP boot function are the same as those for the power-on boot function. However,
the settmg of the power on boot swrtch has no affect on the invoke prrmary MOP boot.

An addrtronal feature of the 1nvoke prrmary MOP bootis that it may allow the trrbutary address of the
remote station to be software assigned instead of switch assigned. This featureis only valid for remote
stations that are multipoint tributaries. To use this feature, the following conditions must exist.
®

The trrbutary address/passwordin the DDCMP address swrtch pack must be zero.

The user program at the remote statron must have estabhshed the trrbutary usmg ‘the control
command (establish tributary). If the remote station is using the multiple address trrbutary
optron the tr1butary address used for bootmg must be the first one established.-

‘ Invoke primary MOP boot with the software-as- signed tributary address does not work if the poweron boot switch is enabled.

4-22

Table 4-§

Mode

Switches
6

Line

Mode Switch Settings

Charactenstlcs

| Network

- Configuration

DMC11 Line

Compatibility

ON ON ON

Half-duplex

Point-to-point

OFF ON ON

Full-duplex |

Point-to-point |

e Yes

ON OFF ON

.Ha.lf-duplex

Point-to-point

OFF OFF ON

Full-duplex

Point-to-point

OFF

Half-duplex
|

OFF ON

|
|

OFF OFF OFF

N/A

Full-duplex

Multipoint trlbutary
B

N/A

statlon

e Mult1p01nt trlbutary

statlon

N/A

m———

‘Yes

station

Half-duplex
|

'Multipoint control

ON OFF OFF

Multipoint control
‘station

OFF | Full-duplex

4. 7 4.3 Swntch Settings for the Remote Load Detect Boot Function— To enable the remote load detect
boot function at a remote tributary station, switch number 5 of the boot enable switch pack (enable
remote load detect) must be set to one (OFF). For the remote load detect boot function, the switchspecified trlbutary address also serves as the password whichis containedin the enter MOP mode mes-

sage.

When using boot functions in point to-point networks, the tributary address/ password switches can, for
security purposes, be set to a unique value since the address of a pomt-to-point nodeis always known to
be one.

4.8 MAINTENANCE REGISTER EMULATION
|
The DMV 11 is placed into maintenance mode when the user program sets bits 0 and 6 at the same time
in BSEL1. When this happens, the microcode enters a maintenance loop and sets MNT RDY (bit 7 of
BSEL?2) to indicate that the microcode is ready to receive a command as defined by bits zero through
three of BSEL2 (see Figure 4-3). The functions of these commands are described in Table 4-6.

NOTE
The microdiagnostics must complete tests 1-12 as a
minimum before allowing entry into the maintenance
loop.
In the maintenance mode, the functions of the CSRs are redefined as follows:
e

BSELDO, bits zero and four, enable the respective microprocessor LSI bus interrupts as defined by bits one and two. See Figure 4-3.

4-23

ne loop command function. Bits four
the maintenance
BSEL?2, bits zero through three, defi

and five are used for interrupting the CPU if an internal microprocessor interrupt occurs.

These interrupts are enabled by BSELO. Bit seven is set by the microprocessor when the
to receive another command function in bits zero through three
~ maintenance loop is ready
e

SEL4 contains a DMV11 memory location for function codes one through five (Table 4-6).

SELG6 contains data written or read for functions one, two, and six. It also contains a 16-bit
SR

address for functions three and four.

SELI10 contains the upper address bits for functions three and four. Only the low byte of this

;

CSR is used.

NOTE

BSEL10 and 11 are only used in 22-bit mode.
BSEL12 through 17 are not shown because they are
never used by the user/DMV11-command structure.

BSEL

MNT

REQ |

RUN| CLR

MNotwse

BSEL3

MNT

uP | 4P | EN

INTA|INTB| ~a- |BSELO

“Ar | B

ATION

| BSEL2

PETEREPPR

[t |wrj

|mov]

NOT USED

DMV11 MEMORY LOCATION (LOW BYTE)

}BSEL4

1, 2 & 6 or 16 BIT ADDRESS FOR FUNCTIONS 3 &4
* DATA READ OR WRITTEN FOR FUNCTIONS

[ oo 0

DMV11 MEMORY LOCATION (HIGH BYTE)

BSELS5|
BSEL7

MTR

;

BSEL11

]

‘
14

NOT USED
13

}

L
11

|
9

UNUSED
8

gg:EFfiufig?g%sg E'TS
4

SEL4

FOR FUNCTIONS 1-5

‘

SEL10

BSEL1C
|

0
.

Figure 4-3 DMVI11 Maintenance Loop Command Format

4-24

MK-2517

Table 4-6
Octal Code
Bits 0-3

Maintenance Command Functions BSEL2 Bits (-3

Function

Reserved to avoid unpredictable results.

Read DMV11 memory location specified by SEL4 (16-bit address specifying a
byte). SEL6 contains the data read from this location.

Write DMV11 memory location specified by SEL4. SEL6 contains the data

written to this location.

Read 256 bytes of DMV11 memory. SEL4 points to the starting DMV11 memory address from which the information is read. The lower eight bits of this address are ignored. SEL6 and BSELS point to the starting LSI-11 bus address
where the information is stored.
|

Write 256 bytes of DMV 11 memory. SEL4 points to the starting DMV11 memory address to which the information is written. The lower eight bits of this address are ignored. SEL6 and BSELS point to the starting LSI-11 bus address
from which the information is read.

Set the 6502 microprocessor’s program counter to the value contained in SEL4.
This is used to start the microprocessor executing code at the
location specified by SEL4.

DMV 11 memory

Set internal loop and null clock for functional diagnostics. Null clock is in-

‘

itialized for 56K b/s.

Set maintenance interrupt flags and clear interrupt disable in processor status.

4-25

——

CHAPTER 5

ASPECTS OF DMV11

" MICROCODE OPERATION

5.1
INTRODUCTION
The functionality of the DMV11 results from the microprocessor and its associated microprogram or
microcode. This chapter discusses the microcode as it relates to:
e

The polling algorithm,

Error recording, and

The internal data base.

5.2 DMV11 POLLING ALGORITHM
In a polling operation the tributaries are in effect asked one by one whether they have anything to
transmit. To accomplish this, the control station sends a polling message with a unique tributary address down the line. The station which recognizes the address responds with data messages or a positive
response.

With DMV11s, polling is based on a priority scheme which is derived automatically and applied dynamically by the microcode. To control pollmg and data message transmission, the DMV11 uses the
followmg 1nformat10n

The tributary’s recent poll history,

The tributary’s user-defined parameters,

The tributary’s protocol state.

In regard to protocol state, the multipoint network control station polls all established tributaries that
are not in the DDCMP halt state. The protocol state of every established tributaryis maintainedin the
associated TSS by the control station. When a tributaryis eligible for polling, all outstanding transmit
messages for that tributary are sent as the poll, up to the limit imposed by the maximum transmitted
message count. If no transmit messages are available for a tributary eligible for polling, the DMV11
automatically transmits the appropriate DDCMP control message.

The DMV11 polling algorithm determines which tributaryis to be polled next, based on each tributary’s polling urgency level. The DMV11 polling algorithm employsthe user-defined TSS and GSS polling parameters as the basis for categorizing tributaries into polling levels. The polling algorithm also
determines the rate at which polling urgency is increased within each polling level. A tributary’s polling
level is based on its recent response history. This classification mechanism, combined with periodic incrementing of polling urgency, results in the most active tributaries being polled most often.

5-1

Calculating Polling Urgency

5.2.1

The polling urgency (priority) of each tributary is periodically calculated when a global timer expires.
This periodic calculation enables the algorithm to enforce minimum poll intervals for each tributary,
and to account for competition between tributaries. The minimum poll interval prevents unneeded polls
|

from:

Delaying other tributaries,

e
~ e

Interfering
with output traffic, and
Causing unnecessary processor overhead.

The polling urgency of a tributary is calculated as a linear function of time elapsed since the last poll.
The calculation is truncated when the maximum value of 255 is reached. The three parameters in this
»

calculation are:

Q _ the initial value 'of thelpic,)llin'gurgency (U);

DELTA T - the polling algorithm global update interval.

R - the rate at which Q is to be increased;

SRR

Figure 5-1 shows the relationship between these three parameters. The appropriate choice of Q and R
can give a variety of behaviors. For this reason, the choice of these values must be based on the desired
performance. One method of choosing Q and R is to define the minimum poll time and the time to
reach maximum poll urgency, and to use these values in the following equations.
~ Minimum poll time = 128-Q (DELTA T)
|

‘R

Time to reach maximum poll urgency = 255-Q (DELTA T)
R

of the global timer and its value depends on the line speed. It must
DELTA T is the user-defined period
be smaller than the smallest nonzero minimum poll interval, but no smaller than 200 ms. The poll priority is calculated for each tributary and stored in a single byte of the TSS. With every DELTA T, the
polling urgency byte is updated. The following interpretation is placed on its value. These values are
represented as base lines on the graph of Figure 5-1. Figure 5-2 shows the relationship of the polling
~

urgency to different values of Q and R.

0 - 127

128
129 -254
D

255

Do not poll the tributary. The minimum poll interval has not expired.

Minimum poll interval has expired. The tributary is eligible for polling.
Minimum poll interval is cheeded. The tributary is eligible for polling. The

“higher values indicate increasing priority in the event of competition between

tributaries.

BRI

cy Competition between tributaries at this priis reached.
Maximum poll urgen
ority is round-robin.

This method of determining values for Q and R 1s applicable when static behavior 1s desired. For many
at a
applications, however, dynamic behavior can improve performance by polling active tributaries
inactive tributaries.
rity
than
faster rate and with a higher prio

T time period, the control stationpolling a‘ilgo_rithm updatés the ur‘genéyzof each
During each DELTA
(excluding dead) to the

operational tributary by adding the value of R for the appropriate polling state
urgency value of each tributary. This updating sequence is performed on the TSS data base in the order
5-2

in which tributaries were originally established 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 original order of tributary establishment), starting at the TSS following the last tributary polled
In this process, the tributary having the highest value of U from active, 1nact1ve and unresponsive
tributariesis selected as the next tributary to be polled.
TIME TO

MAXIMUM URGENCY

25—

———

- --

—— — ——— ——

. — ————

ELIGIBLE FOR
POLLING

E’

MINIMUM

POLLING INTERVAL

el e D
o

127

NOT POLLED

®)
PN

R
¥

_’I I‘_At |

| L|

]

At =200ms
MK-2643

Figure 5-1 Interrelationship Between Polling Parameters Q, R, and DELTA T

If an urgency of 255 is detected during the process ‘of scanning the TSS data base for the tributary
having the highest value of U, the scan process is halted and that tributaryis immediately selected for
polling. Once the selected tributaryis polled its urgency reverts to the assigned value of Q for its polling level

Dead tributaries are polled at a rate determined by the user-defmedparameter DEAD T. One dead
-

tributaryis polled at each expiration of the DEAD T tlmerand the scan of dead tributariesis resumed
from the last dead tributary polled.

5-3

255

I
|

|
|

I
I

127

==
g

Vi£

DELTA

9.55

8.0

6.375

3.2

—t—4.0

200ms

|
||

>/I
I

|I'

5
O
=

|
'
3

12.0

16.0

200ms

MK-2647

Figure 5-2 Relationship Between Polling Parameters Q, R, and the Minimum Polling Interval

modified states. There are
The dynamic polling algorithm uses Q and R values based on dynamically
B
.
four of these states:
1.

Active — The polling algorithm maintains a tributary as active when it responds to polls with

data messages.

Inactive — The polling state of an active tributary is changed to inactive when it responds to a
consecutive number of polls with nondata DDCMP messages. The count of consecutive nondata messages received from a tributary in the active state is designated by the user-defined
parameter NDM-INACT (number of no-data messages required to go inactive).
Unresponsive — A tributary currently active or inactive is changed to the unresponsive state
to a consecutive number of polls (each poll results in a
when it fails to respond in any way
e polls without responses is designated by the
consecutiv
of
count
The
selection timeout).
of timeouts to go unresponsive).
(number
SP
TO-UNRE
parameter
user-defined

Dead — A currently unresponsive tributary which continues to be unresponsive to consecutive

polls is changed to dead. This occurs when the number of selection interval timeouts designated by the user-defined parameter TO-DEAD (number of timeouts to go dead) is exceeded. Unlike tributaries in the other polling states, dead tributaries are always polled on a

by the user-defined global
round-robin basis with the period between polls being determined
|

parameter DEAD T (dead timer).

5-4

When specifying the parameters controlling polling levels, the user has the option of accepting the
defaults (see Table 4-2 and 4-3), or selecting specific values in place of these defaults. 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 imposed by a control command remains in effect, irrespective of
tributary performance until polling control is handed back to the polling algorithm by the user program. Thisis done byissuing an unlatch polling state control command

Figure 5-3 shows the relationShip of the the default values of Q and R for the active, inactive, and
unresponsive polling states.

255

ACTIVE Q=255,

R=0

ELIGIBLE FOR

POLLING

/
&

O/P
i

Z
LLl

)

S 128

—

)

—
e

NOT POLLED

_
200ms

1.2sec

p—

2.60sec

DELTA T =200ms

MK-2842

Flgure 5-3

Relatlonshlp Betweenthe ‘Default Values for Q and R for the Three
Polling Act1v1ty Levels

5-5

Each tributary has Q and R values defined by the user for the active, inactive, and unresponsive polling

states. This allows dynamic modification of the behavior of the polling algorithm. The basic mechanics

of the polling algorithm are: |

e
e

If a tributary always responds to a poll with daté, it remains in the active state. The polling

urgency in this case is calculated using the Q and R values specified for the active state.

If a tributary responds to a user-defined number of consecutive polls with no data messages

(ACKs only), polling is changed to inactive. The polling urgency is then calculated based on
|

the Q and R values defined for the inactive state.

If a tributary in either the active or inactive states fails to respond (times out) to a predetermined number of consecutive polls, the polling state is changed to unresponsive. The Q
and R values defined for the unresponsive state are used to calculate the polling urgency.

If a tributary in the unresponsive state fails to respond to a predetermined number of consecutive polls, the polling state is changed to dead. The user is notified of this transition by a
control response. Polling dead tributaries is very time-consuming because it usually requires
a timeout. Therefore, dead tributaries are not polled on a priority basis. Instead, a global poll
interval is defined for dead tributaries. Each time this timer expires, a single dead tributary
is polled. If at any time a dead tributary responds to a poll with a data message or ACK, its
o

“

state is changed to active.

NOTE

A tributary (not in the active polling state) is automatically returned to the active state when it responds to a poll with a valid data message. It also

becomes active when the user program allocates a
transmit buffer to it.

Figure 5-4 is a state diagram describing the transitions between polling states. The actual transitions are
|

dependent on the particular polling parameters.

~ The user may control sending of all polls by defining a poll delay interval that must expire before a poll
can be sent.

The program-selectable parameters which pertain to polling are included in Table 4-2. These parameters are set by the user with a write TSS command.

5.2.2 Criteria for Determining Polling Parameters

Although there are no absolute rules for determining polling parameters, there are general guidelines
for deriving them. These guidelines are presented in sections 5.2.2.1 through 35.2.2.4.

5.2.2.1 Determining a Value for DELTA T - For most multipoint network applications, the default
value of 200 ms for DELTA T is adequate. The default value of 200 ms is the smallest permissible
value for DELTA T and represents the actual time required for the microcode to update the urgencies
of 12 tributaries. However, in specific cases a higher value of DELTA T might be recommended. An
example of this is a network formed by low traffic devices.

5-6

|
NOTE
The minimum polling interval defines the time required for a tributary to reach the urgency threshold
value of 128. This is the minimum time required to
be eligible for polling. The maximum polling interval

cannot be determined. This is because it is a function
of line speed, message traffic, the number of tribu-

taries in a network, and the polling states of those
tributaries. These variables make it impossible to
predict the time at which any tributary in a network
is polled.
|

ACTIVE

(NDM — INACT)

(TO — UNRSP)

UNRESPONSIVE

\ (To - DEAD)

MK-1956

Figure 5-4 State Diagram of Polling State Transitions
5.2.2.2

Determining Values for Q and R - For
a given value of DELTA T, the minimum polling inter-

val is a function of the user-defined parameters Q and R. For example, if a minimum polling interval of
3.2 seconds is desired for a tributary (assuming DELTA T = 200 ms), the parameters Q = 0,and R =
8 satisfy this requirement. With these values of Q and R, the time to reach maximum polling urgency is

5-7

6.375 seconds (see Figure 5-2). Notice that if Q = 64 and R = 4, the minimum polling interval remains at 3.2 seconds but the maxrmum pollmg urgency increases to 9.55 seconds
NOTE
Reachmg maximum pollmg urgency represents maximum eligibility for polling, but does not guarantee
that a tnbutary wnll be polled

Figure 5-3 graphs the relatronshlp between thedefault values of Q and R for each of the three pollmg
states. When all tributariesin a network have the default value for Q and R, and all tributaries are in
the active polling state, the manner of pollingis round-robin.
5.2.2.3 Determining a Value for Poll Delay — The user-defined global parameter poll delay imposes a
fixed delay between control station polls. This provides a mechanism for regulating message traffic
without changing the values of Q and R for individual tributaries. During this delay, transmission from
the control station to tributariesis halted for the interval defined by the poll delay timer. This interval
begins when the tributary just polled deselects itself.

The ab1]1ty to regulate message traffrc through a single parameter is valuablein mult1pomt networks.

Thisis especially true where DMV11s are configured together with slower character 1nterrupt communication devices such as DUP11s. The value selected for poll delayin these circumstances is a function

of the character handling rates of the non-DMV11 devices.
-In remote mult1pomt networks where the distance between the control station and tributaries varies

significantly, thereis a greater chance of transmit and receive errors. Thisis due to the differencein
communication line settling time in such a network. In such instances, the settling time for the most
distant tributary station should be used for determining a value for poll delay

For DMV11 1mplemented hrgh-speed local networks, this parameter is unnecessary The default value

(zero) for poll delayis usedin these networks.

5.2.2.4 Determining a Value for DEAD T - This global parameter establishes the rate at which dead

tributaries are polled. Dead tributaries are polled on a round-robin basis, W1th one tributary polled at
each expiration of the dead tributary timer.

Polling dead tributaries can significantly impact network line utilization. The shorter the period of this
timer — the greater the impact. For a given value of DEAD T, the impact decreases as system line
speed increases. When determining a value for this parameter, the primary goal is to minimize the impact on network line utilization.

The value 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 and the selection interval timer are
equal, only dead tributaries are polled. For most system applications, the period of the dead tributary
should be from three to ten times greater than that of the selection interval timer. This of course depends on line speed. The default value for DEAD T is ten seconds. Thisis about three times the default
value for the selection 1nterva1 trmer

5.3

ERROR COUNTERS

In multipoint networks many tributaries tie to the same transmission line. Because of this, it is more
difficult to determine which link, ifany, is causmg errors. To aidin troubleshooting, the DM VIl maintains extensive error counters. Every DMV11 in the network (the control station and each tributary)
uses error counters to record errors. This allows user programs at any DMVll in the network to determine ‘overall error rates and to detect a malfunctioning 11nk
5-8

Data link errors are indicated to the DMV11 by DDCMP negative acknowledge messages (NAKs).
Each NAK contains an address field and a reason code that identifies the source and reason for the

NAK. In general, when an error is detectedin an incoming message, the station that receives the message sends a NAK to the station that sent the message. By recording NAKs sent and NAKs received,
each point or tributaryin the networkis able to compile statistics on the condition of the link established between the two stations. DDCMP error recording 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 dlrectlons on the link.

There are three main categories of error counters used by the DMVII data link counters, station

counters, and threshold counters. Data link counters and threshold counters are maintained for each
trlbutary/ control station pair on a physical link. These counters are locatedin the tributary status slots
of the data memory (Figure 5-5). Station counters are maintained for the physical link as a whole, and
are located in the global status slots of the data memory (Figure 5-6). Unless otherwise stated, all
counters increment to a maximum value and hold that value until cleared.

5.3.1
Data Link Error Counters
Data link counters are of two types; cumulative and background The cumulative counters are 8-bit
registers which latch at 255. The background counters are 16-bit registers which latch at 65535. The
cumulative data link counters record and total all occurrences of an error and group them into the following categories.
Data errors outbound,
Data errors inbound,
Local reply timeouts,
TN

Remote reply timeouts,

Local buffer errors,
Remote buffer errors,

Selection timeouts.
Background data lmk counters are used to provide a statlstlcal 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.

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 established tributary. Data link
counters are cleared by:
A master clear of the

A control command to establish the tributary, or

DMVI],

A user-issued control command to read and clear the TSS error counters.

5.3.1.1

Data Errors Outbound - This 8-bit group counter records NAKs received for data errors oc-

curring on the communications channel outbound from this station. There are three types of outbound

errors for which this counter records NAKSs received; header blockcheck (OHBCC), data field block-

check (ODBCC), and reply response (OREP). Three separate flag bits indicate which type of outbound
error 1s being counted.

OHBCC (outbound header bldckcheck) is set when a NAK with a reason code of one is re-

ceived for a header block-check error for either data or control messages.

5-9

ODBCC (outbound data field blockeheck) 1s set when a NAK w1th a reason code of two is

| recelved for a data field block-eheck error.

| OREP (outbound reply response)is set when a NAK wrtha reason code of threeis. recelved
for a reply message response

TRIBUTARY STATUS SLOT (TSS)
ADDRESS (OCTAL)

RESERVED

RECEIVE THRESHOLD ERRORS
~

- SELECTION THRESHOLD ERRORS

10
)

TRANSMIT THRESHOLD ERRORS

—

DATA MESSAGES TRANSMITTED

'DATA MESSAGES RECEIVED

—

SELECTION INTERVALS

—

DATA ERRORS OUTBOUND

Ioaepiooscei oHBcce |
~ DATA ERRORS INBOUND |
!"lREP"I IDBCC !'IHBCC
RESERVED
'RESERVED

13
|

LOCAL BUFFER ERRORS
| LBTS I,LBTU
RESERVED

REMOTE BUFFER ERRORS
I RBTS‘I'RBTU
RESERVED

SELECTION TIMEOUTS

RESERVED
17

LocAL REPLY TIMEOUTS

!~I‘RTS l NRTS
|

REMOTE REPLY TIMEOUTS

MK-1960

‘/—~—~

| Figure 5-5 | Da’taLink and Threshold Error Counters

5-10

GLOBAL STATUS SLOT (GSS)
ADDRESS (OCTAL)

REMOTE STATION ERRORS

| I rsTR | rseL | RmHEE |ROVRN
16

LOCAL STATION ERRORS

ILOVR
17

LUNDR | LMHFE | LOVRN

GLOBAL HEADER BLOCK CHECK ERRORS

MAINT. DATA BLOCK CHECK ERRORS
MK-1959

Figure 5-6

5.3.1.2

Station Error Counters

Data Errors Inbound — This 8-bit group counter records occurrences which normally result

from data errors on the communications channel inbound to this station. Three separate bits indicate
specific error types associated with this counter.
e

[HBCC (inbound header blockcheck) is set when messages having header-block check 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 mult1p01nt 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 NAKSs with a reason code of two are to
be sent for data field block-check errors.

IREP (inbound reply '_response) is set when NAKs with a reason code of three are to be sent
for a reply response.

5.3.1.3

Local Reply Timeouts — This 8-bit counter records occurrences which result from:

The loss of commumcatlons between two statlons while the one recordmg this error has data
to transmit, or

The choice of an inappropriate value for the reply timer.

Specifically, this error counter records the sending of a REP message.
¢

5.3.1.4

Remote Reply Timeouts — This 8-bit counter 4records occurrences which result from:
The loss of commumcatlons between two stations wh11e the remote station has data to transmit, or
,

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

Spe01f1ca11y, 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.

5-11

5.3.1.5 Local Buffer Errors — This 8-bit counter records the fact that the user program at the station
recording the error failed to properly allocate receive buffers to data messages from the remote station.

Two separate bits indicate the specific errors associated with this counter.

e
e

LBTU (local buffer temporarily unavailable): is set when a buffer is temporarily unavailable.

a reason code of eight is to be sent.
This condition indicates that a NAK with

LBTS (local buffer too small) is set when a local buffer is too small for the incoming mes-

sage. This con’dition‘in.dicates that a NAKwith a reason code of 16 is to be sent.

5.3.1.6 Remote Buffer Errors — This 8-bit counter records the fact that the user program at the remote
station failed to properly allocate receive buffers to data messages from the station recording the error.
Two separate bits indicate the specific errors associated with this counter.
e

RBTU (remote receive buffer temporarlly unavallable) is set when a NAK with a reason
code of eightis received.

RBTS (remote receive buffer too small) is set when a NAK with a reason code of 16 is received.
|
<

8.3.1.7
®
e

Selection Timeouts — This 8-bit counter records the occurrences which result from:
Loss of commumcatrons with a remote station,
'
Data errors on the communications channel to or from the remote statlon and

The chorce of an 1nappropr1ate value for this station’s select timer.

Th1s counter is used only by half-duplex point-to-pomt or multipoint control stations.‘Two separate bits
indicate the spec1flc errors assoc1ated with this counter.
®

NRTS (no reply to select) is used to record selection intervals in Wthh no transmission is

received from the tributary, and no attempt to transmit 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 an attempted transmission.
e

IRTS (incomplete reply to select) is used to record selection intervals which were not proper-

‘ly terminated. Specifically, it records the expiration of the select timer preceded by receipt
of a valid control message, recelpt of a valid header, or detection of an attempted transmission. An attempted transmission is indicated by:

—
—
—

The presence of a carrier signal,

The receipt of a DDCMP synchronization sequence, and
The receipt of an SOH, ENQ, or DLE.

5.3.1.8 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. Messages sent as a result of retransm1ss1on are
not includedin this count.

5.3.1.9 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. Messages received out of sequence or in error are not
mcludedin this count.

5-12

5.3.1.10 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 half-duplex point-to-point and multipoint control stations. It can be used as a statistical base when evaluating the number of selection timeouts.

5.3.2

Station Error Counters

Station counters are 8-bit counters which latch at 255 and record unusual occurrences. These occurrences may be the result of:
~
e
o

A hardware or software fault at this station,
A hardware
or software fault at a remote station, or

A data error on the communications channels undetected by the header block-check 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 block-check errors, and
Maintenance data field block-check errors.

Station counters are cleared by:
®

A master clear of the DMVII or

A user-issued control command to read and clear the GSS error counters.

5.3.2.1

Remote Station Errors — This 8—bit counter rvecords occurrences caused by a,’fa_ult in a remote

station or by an undetected data error on the channel inbound to this station. Four separate bits indicate

- the specific errors associated with this error counter.

ROVRN (remote receive overrun) is set when a NAK with a reason code of nine is received
for a receive overrun.

RMHFE (remote message header format errors) is set when a message is received which has
a header format error. This condition 1nd1cates 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 containing an address field which does not match the address of the currently 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-dependent maximum transmission interval is exceeded without releasing the channel
(babbling tributary), or 2) the channelis not released followmg the end of a selection interval
(streaming tributary).
|
:

5.3.2.2 Local Station Errors - This 8-bit counter records occurrences caused by a fault in a local station or by an undetected data error on the channel outbound from this statlon Four separate bits indicate the specific errors associated with this error counter.

LOVRN (local receive overru'n, NAK sent) is set for local station receive overruns. This condition indicates a NAK with a reason code of nine is to be sent.

5-13

LOVR (local receive overrun, NAK not sent) is set by a receive overrun when a NAK is not

sent. For a multipoint tributary, this happens if an overrun occurs while receiving a header
For other statlons 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 whena NAK with a reason code of 17 is
received to indicate a message with a header format error was sent by this station.

5.3.2.3 Global Header Block-Check Errors — This 8-bit counter records the occurrence of header
block-check errors that are not recorded on a per tributary basis. Specifically, it counts header blockcheck errors for maintenance messages and for messages to trlbutarres where the address freld does not

5.3.2.4

‘

match the station address.

Maintenance Data Field Block-Check Errors — Thrs 8-bit counter records the occurrence of

data field block check errors for maintenance messages.
Threshold Error Counters

5.3.3

Threshold error counters are used to determine if a persistent fault exists. A per51stent fault is one
which occurs seven consecutive times. Whenever a threshold counter reaches its maxrmum value (7),
the user program is notified by a control response

In the DDCMP run state, threshold counters are cleared when the user is notified. In this way the user

is continually informed of a pers1stent fault. In the DDCMP ISTRT and ASTRT states, threshold
counters are not cleared when the user is notified. In this way the user is not continually mformed of an
inoperative remote statlon

A point-to--point station maintains a smgle set of threshold counters. A multipoint control station maintains a separate set for each trrbutary A multipoint tributary maintains a single set unless it supports
multiple tributary addresses in which case it mamtams a s1ngle set for each estabhshed trrbutary address.

5.3.3.1

Transmit Threshold Errors - Th1s 3- b1t counter T3 1ncremented (if less than seven)in the fol|

,'/

lowmg mstances

lv\

There are three types of threshold error counters: transmit, receive, and selectlon

The DMVll is in the lSTRT state when a STRT message 1s sent

The DMVll is in the ASTRT state when aSTACK message is sent or

The DMVII is in the run state and a NAK w1th a reason code other than three (REP response) 1S recerved or when sendmg a REP message

The transmlt threshold error counter is cleared
o

Upon entering’the ISTRT, ASTRT, or run states.

Whilein the run state one of the followmg occurs:

A transmlt threshold error is reported

A NAK, ACK, or data message is recerved acknowledgmg a new message, or

—

A NAK ACK, or data message IS recelved when no messages are outstandmg

5-14

HEXADECIMAL
0000

,
SCRATCH PADS

16 BYTES

Q-BUS CSRs

8 BYTES

—

SCRATCH PADS

OUT NPR ADDRESS
3B

SCRATCH PAD

3 BYTES

—

-—

—

1 BYTE

IN NPR ADDRESS

3 BYTES

.
SCRATCH PAD

BYTE

—

GLOBAL STATUS SLOT
-

‘

—

32 BYTES

—

64 BYTES

256 BYTES

MICROPROCESSOR STACK

64 BYTES

BUFFER AND OUTPUT QUEUE
500

98 ENTRIES

|
8 BYTES/ENTRY

TRIBUTARY STATUS SLOTS

800

——

SLOT MAPPING TABLE (SMT)

12 ENTRIES

*
|
64 BYTES/ENTRY

MK-2485

Figure 5-7

Data Memory Map

A DMVI1I lmked listis made up of five kmds of linked lists.
1.

The free linked list — A list of empty lmk blocks used by the mlcrocode to form the remaining

The response linked list — A queue of responses for posting to the user program.

The{comm_on.buffer,pool linked list — A list of the accessing information for each receive
buffer assigned to the common pool. There is one link block for each assigned buffer.

Receive buffer linked list — A list of receive buffer accessing information. One of these is

maintained by the microcode for each established tributary having prlvate recelve buffers
a551gned Thereis one link block for each buffer.

Transmit buffer linked list — A list of transmit buffer accessing information. One of these is

maintained by the microcode for each established trlbutary having transmit buffers assxgned
Thereis one linked list for each buffer.
|

54.1.1 The Free Linked List - The free llnked list from which all other linked lists draw link blocks, is
mamtamedin the lower section of data memory called the buffer and output queue (BOQ) (Figure 57). These 832 bytes translate into a total of 104 link blocks available for use by the operational lmked
lists. In this way, the free linked list functions as a finite resource for the operational linked lists.
- 5-16

kinds of linked lists.

5.3.3.2

Receive Threshold Errors — This 3-bit counter is incremented (if less than seven) when a NAK

with one of the following reason codes is sent.

—\O
OO0 W N =—

Reason Code

Description

Header block-check error.
Data field block-check error.
REP response.

Buffer temporarily unavailable.
Receive overrun.

Message header format error.

This counter is cleared when:
e

Entering the ISTRT, ASTRT, or run states,

A control message with a correct header blockcheck is received without a header format er-

ror,

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

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

format error, or

5.3.3.3 Selection Threshold Errors — This 3-bit counter is only used by multipoint control stations and
half-duplex point-to-point stations. It is incremented (if less than seven) when a selection timeout occurs.

It is cleared upon receipt of a message with the select bit set, or while in the run state and a selection
threshold error is reported.

5.4 DMV11 MICROCODE INTERNAL DATA BASE OVERVIEW
Functionally, the

DMV 11 internal data base provides the mechanism for managing:

The assignment and completion of transmit and receive buffer‘s,

The queuing of DMV11 résponSes, |

The assignment of TSS structures to established tr1butar1es for the storage and mamtenance
of tributary and global status mformatlon

A map of this data base is shown in Figure 5,-7. The data base is implemented by three basic structures:
o
e
e

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

Each of these are described below in terms of organization and function.
5.4.1
Linked Lists
A linked listis an open-ended data list made up of fixed-length blocks linked by pointers. Each of these
blocks (link blocks) contain seven bytes of data and a one byte pointer to the next link blockin the list.
The pointer in the last blockin the listis a terminator value. Figures 5-8 and 5-9 illustrate the standard
format for DMV11 linked-list structures.

5-15

START OF LIST
POINTER

]
l

| LINK BLOCK
POINTER A

DATA

LINK BLOCK
POINTER B

DATA

LINK BLOCK

/r,—.\\

“

" POINTER C

”

DATA

o E
oF

o N

_are—.

3778

1———[

TERMINATOR

DATA

MK-1961

Figure 5-8

DMYV11 Linked List Structure Format

As previously stated, a linked list is equipped with two list pointers; one that points to the start of the
list and one that points to the end of the list.When a link block is removed from the free linked list, the
start of the 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 operational linked lists, it is added to the end of the free linked list

and its internal pointer is set to the terminator value of 377 octal. In addition, the internal pointer in the
next to last link block is changed from 377 octal to the address of the link block just added. The startand end-of-list pointers for the free linked list are maintained in the station GSS.

5-17

~ Link blocks are removed from the free linked list and added to the receive, transmit, or common pool
buffer linked lists when the user program issues control or buffer address/character count commands
for that purpose. Similarly, link blocks are removed from the free linked list and added to the response
linked list when the DMV 11 microcode posts a response to the user program. If the last link block is
removed from the free linked list, the start of the list pointer is set to the terminator value of 377 octal
to indicate there are no more link blocks available. In this event, the next request for a link block generates the fatal error QUEUE OVERFLOW. For this reason, the buffer allocation strategy for a user
program must be designed to assure an adequate number of link blocks.

LINK POINTER

POINTER TO THE NEXT LINK BLOCK

MESSAGE..NU‘MBER: | oooMP MESSAGE NUMBER
BSEL 3

TRIBUTARY ADDRESS

BSEL 4

'BUS ADDRESS

BSEL 5

BUS ADDRESS

BSEL 6

CHARACTER COUNT LOW

BSEL 7

CHARACTER CNT HIGH, BA 16, BA 17

BSEL2

TYPE CODE AND BA 18-21
MK-2497

Figure 5-9

Standard Link Block

5.4.1.2 The Response Linked List — This linked list functions as a queue of buffer disposition, control,
and information responses to be posted to the user program. The format of the link block for each of
these three responses is shownin Figure 5-8. When preparing a link to convey a control or information
response, the microcode clears all unused bit positions in the link block to zero. However, link blocks
restored to the free linked list remain unchanged

The start-of-list and endof—list pomters for the response linked list are maintainedin the station GSS.

5.4.1.3 Buffer Lmked Llsts — A buffer llnked list1s provrded for each type of message buffer allocated
by a user program. These are:
e

Common pool receive buffers,

@
®

Transmit buffers.

Private receive buffers, and

Each link block in a buffer linked list provides the location and size of a buffer in main memory.
The Common Buffer Pool Linked List — This linked list provides a queue of receive buffers available to
all established 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 Section 3.3. 3) The start- of-list and
end-of-list pomtersfor this lmked list are mamtamedin the station GSS

Recelve Buffer Linked List — This linked list serves as a queue of private receive buffers One list 1s

maintained for each tributary established at a multipoint station. For point-to-point stations, one listis
maintained at each station. The start- and end-of-list pomtersfor each receive buffer linked list are
maintainedin the associatedtributary’s (or station 's) TSS.

5-18

Transmit Buffe'r'vLinkfed List — This list functions as a queue of transmit buffers. One list is maintained

- for each tributary established at a multipoint station. For point-to-point stations, one list 1s maintained

at each station. The start- and end-of-list pointers for each transmit buffer linked list are maintained in
the associated tributary’s (or station’s) TSS.
Con
et e
o

A unique feature
of this link block
is the message number field. When a message is transmitted from a
buffer
the header
, of that message contains the DDCMP message number (in the message number
field). The microcode uses this field to locate the buffer for a message that has been NAKed after

transmission and, therefore, must be retransmitted.

SRR

5.4.2 Slot Mapping Table =~
=
T
Under DDCMP, the 8-bit message header address field permits a maximum of 255 unique tributary
addresses in a multipoint network. However, the DMV 11 microcode limits the number of established
tributaries to 12. In order to implement DDCM
a tributary
P, in a DMV11-based multipoint network
can have a TSS address in the range of 1 to 255. However, only 12 of these tributarie
may
s be estab-

lished at any one time.

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 locations in DMV11 data memory; one location
for each of the 255 possible tributary addresses, and one location to address the GSS. ;

The function of the SMT is to map an 8-bit tributary address into one of the‘ 12 av‘éilablé TSS structures. When a tributary is deleted,
its TSS and SMT entry is released for reassignment. When 12 tributaries are established and an attempt is made to establish a 13th, a procedural error is posted to the user

5.4.3

TSS and GSS Structures

The TSS and GSS structures occupy separate sections of data memory. The GSS is a single 64-byte
section while the TSS structure consists of twelve 64-byte sections.

5.4.3.1
~

The Global Status Slot (GSS) — Functionally, the GSS is used to:

Maintain control and status information specific to the operation of the microcode,

Record event counts and error conditions that are global in nature, and

Store global parameters.

The majority of the GSS is devoted to microcode control and status information. A detailed map of the

GSS is shown in Figure 5-10.

Access to the GSS is accomplished on word boundaries. A user program can read any GSS location
through the control command. The content
of the addressed location is transferred to the user program

through an information response. A user program can read and clear only GSS station error counters.
The four global parameters (locations 34 through 37) are written by the user program through the control command.

S5.4.3.2 Tributary Status Slots (TSS) - A TSS contains four general categories of tributary informaBN
—

tion (Figure 5-11):

Protocol and tributary status,
Error and statistical counters,

-Message exchange variables, and
Polling parameters.

5-19

Protocol and Tributary Status — This category includes information on the tributary’s protocol state, its
status with relation to the logical communications line, its protocol status, and its polling status. Tributary polling status is maintained only at a multipoint network control station. Although this information is only pertinent to, and used by the DMV11 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 meaning of specific error counts. These
counters can be read and cleared by the user program. The function of each of these counters is described in detail in Section 5.3.

Message Exchange Variables — This category includes a range of variables used by the microcode to
control the transmission and reception of message data. This includes the common buffer pool quota
|

assigned by the user program.

A group of timers is also included in the message exchange variables. These timers can be preset to a
specific timeout value by the user program, and directly concern message traffic transferred on the
logical link by the associated tributary. These timers are referred to as:
Transmit delay timer,

Selection interval timer,*
Maximum transmitted message count,
Babbling tributary timer.

The link management functions performed by these timers is detailed in Chapter 4.

Polling Parameters — These parameters are user-defined values that are used by the polling algorithm
to conduct dynamic polling activity in multipoint networks. The functions performed by the DMV11
polling algorithm, and the criteria for determining the values for each parameter, are discussed in detail
in Section 35.2.

* A timeout value for the selection interval timer is maintained in each tributary’s TSS, but the actual timer is maintained in
the GSS.

5-20

GSS

ADDRESS

ADDRESS
(HIGH)

RCVPTR

~ RPM CNTR

XMTPTR

ACKTIM

(LOW}

TSP

ACKTIM

(HIGH)

2 | nNasp

- MODEM
MODE

BUFPTR.

| s/oOF

UNUSED

E/OF

FLAG REGISTER "D"

4 |

s/0Q

(HIGH)

~E/0Q

{

s/oC

6 | TIMER STATUS
S
-~

'LOCAL STATION ERRORS
| LOVR |LUNDR| LMHFE |LOVRN

S/R TIMER (LOW) .

REMOTE STATION ERRORS

RSTR | RSEL |RMHFE|ROVRN

_E/OC

~ CLEAR TO SEND TIMER (LOW)

(HIGH)

GLOBAL HDR BCC ERRORS
MAINT DATA BCC ERRORS

~ B/CW TIMER (LOW)

MK-2489

Figure 5-10

;Globalu.Status Slot (Sheet 1 of 2)

5-21

//.»w‘:w\ .

ADDRESS

30|

XHDR (1)

’

GSS

GSS
ADDRESS

POLL DELAY TIMER (LOW)
(HIGH)

(2)
21

(3)

31 | POLL UPDATE POINTER

(4)

DEAD SCAN

(5)

32 |

(6)

USYRT HANG TIMER

RCVHDR (1)

(3)

B
27

35|

(5)
|

CARRIER WAIT

TIMER COUNTER

NUMBER OF SYNCS
RESERVED

(2)

CARRIER LOSS TIMER

DELTAT

(LOW)
(HIGH)

(6)

36|

R TIMER (LOW)

(HIGH)
D TIMER (LOW)

(Low)

DEADT
~

POLL DELAY

(HIGH)
(LOW)
(HIGH)

(HIGH)

MK-2490

Figure 5-10

Global Status Slot (Sheet 2 of 2)

5-22

TSS

ADDRESS

DATA MESSAGES
RECEIVED

TRIB STATUS FLAGS

NAK REASON =

- SELECTION

~ TRIBUTARY ADDRESS

INTERVALS

~ POLL STATUS FLAGS

DATA ERRORS OUTBOUND

POLL STATUS FLAGS =

- POLL RATE (Ri)

RESERVED| OREP| ODBCC|
13

POLL PRIORITY (Vi)
" RESERVED

RECEIVE THRESHOLD ERRORS
TRANSMIT THRESHOLD ERRORS

MESSAGES

IHBCC

LOCAL BUFFER ERRORS |

REMOTE BUFFER ERRORS
RESERVED
| RBTS|

~ SELECTION THRESHOLD ERRORS
DATA

IDBCC|

RESERVED
| LBTS
| LBTU

MAX MSG COUNTER
' COMMON POOL QUOTA

DATA ERRORS INBOUND
RESERVED| "IREP|

OHBCC

SELECTION TIMEOUTS
RESERVED |

RBTU

IRTS| NRTS

'LOCAL REPLY TIMEOUTS

REMOTE REPLY TIMEOUTS

TRANSMITTED

MK-2722

Figure 5-11 Tributary Status Slot (Sheet 1 of 2)

5-23

TSS

ADDRESS

20 | N HIGHEST MSG NUMBER XMITD

PRESET VALUE FOR TRANSMIT

A HIGHEST MSG NUMBER ACK'D

DELAY TIMER

T NE‘XT MSG NUMBER TO XMIT

21»

Q VALUE FOR ACTIVE STATE

T
TPTR ADDR OF LNKBK FOR MSG

R VALUE FOR ACTIVE STATE

22 | X LAST MSG NUMBER XMIT'T

Q VALUE FOR INACTIVE STATE

XPTR ADDR OF LNKBK FOR MSG X

R VALUE FOR INACTIVE STATE

OL T/0)
X REPLY
23 | CX (CONTR

Q VALUE FOR PDEAD STATE.

S/DX START OF XMIT BUF QUEUE

R VALUE FOR PDEAD STATE

24 | E/OX END OF XMIT BUFFER QUEUE

R HIGHEST MSG NUMBER RCV'D

S/OR START OF RCV BUFFER QUEUE

TRANSMIT

DELAY

-->

#T/0

--->

|INACTIVE STATE
PDEAD STATE

DEAD STATE

MAXIMUM MESSAGE COUNTER

36 | SELECTION INTERVAL
TIMING COUNTER

TIMER

#NDM

35 | #T/O --->

E/OR END OF RCV» BUFFER QUEUE
26

| NO DATA MESSAGE COUNTER

BABBLING TRIBUTARY
TIMING COUNTER

T/O COUNTER

MK-2721

Figure 5-11

Tributary Status Slot (Sheet 2 of 2)

5-24

CHAPTER 6

TECHNICAL DESCRIPTION

6.1

INTRODUCTION

This chapter provides a block dlagram functlonal deser1pt10n of the DMVll

The DMVII is basrcally a 6502 mlcroprocessor ‘with 1nterfaces to the LSI-11 bus and the outside
world. The 6502is supported with 16K bytes of ROM and 2K bytes of RAM. All I/Ois memory map-

ped to the 6502 data bus. The LSI-11 1nterfaceis made up of three areas:

1.
2.

NPR circuitry — Full 16-bit d1rect memory access is allowed on data transferred i_n and out.
Independant address registers
are provided for the in and out NPR addresses.

CSRs - Eight 16-bit CSRs are mapped into the on-board RAM. Any time the LSI-11 bus

accesses these reglsters the 6502is halted, the on-board 8-bit RAMis demultrplexed to 16

bits, and 16-bit datais read from or written to the RAM.
3.

Interrupt circuitry— The 6502 can cause mterrupts to the LSI-11 bus if enabled by the CPU
Two separate bits allow two levels of 1nterrupt to be initiated.

The outside world interface is handled by a USYRT using a serial transmit and a serial receive line.
The USYRT interfaces directly to the 6502 data bus; activated and monitored by data and status transfers using a predefined address space.

The remaining circuitry which makes up the DMV11 is used to support the USYRT, the LSI-11 interface, the DDCMP protocol, and the modem interfaces.

6.2 LOGIC DESCRIPTION

|
For discussion purposes, the DMV11 loglc1S d1v1ded into the blocks shownin Flgure 6 1. The errcultry
and functions represented by each of these blocksis describedin Sections 6.2.1 through 6.2.7.
6.2.1

Control and Address Decoder

This block contains the 6502 mrcroprocessor timing circuits, 6502 data and address mterfaces and

address deeoders

6.2.1.1 The 6502 Microbprocessor — The 6502 microprocessor is a'40-pin microproeessor with a full 16-

bit address bus, an 8-bit bidirectional data bus, and two interrupts.

-»

The 6502 is organized around two primary buses: the address bus and the data bus. The address bus is
used to transfer the address generated by the microprocessor to the address inputs of memory. The data
bus consists of an 8-bit bidirectional data path. All data and instructions are transmitted on this bus.

The 6502 provides a sync-signal to indicate when it is fetching operation code from program memory.
The timing of all data transfersis controlled by a two-phase clock (two nonoverlapping square waves)
referred to as Phase 1 and Phase 2. The address lines and read/write line stabilize durrng Phase 1 and
datais transferred during Phase 2.
|

DATA

: MEMORY

ADDR

IV
|

|=|.controL |

JdJl
;%ISH

j=—>{

Koam

ADDRESS
DECODER

|
|

RECEIVE
FIFO

|«

DATA
BUS

| INTERFACE|
——————q

| Mobem

|ucontROL |}

—

IMEMORY ]:
>{ AND RESET| |
ICONTROLI

MK-2523

DMV11 Block Diagram

6.2.1.2

Timing Circuits — The source for all the timing signals necessary for the DMVII is a 20 MHZ

The 6502 microprocessor clock— This clock rate is 1.67 MHz and1is generated by d1V1d1ng
|
the 20 MHz clock by 6 and then by 2. The 1.67 clock has a duty eycle of 50% %2 ns.

The 1ntegra1 modem receiver 20 trmes clock This1s a 1 11 MHz clock (20 times 56 KHz
recerve clock rate) whrehis generated by d1v1d1ng the 20 MHz clock by 9 and then by 2.

The 1ntegra1 modem 2 times clock— Thisis a 111 KHz clock whrchis generated by d1v1d1ng

crystal. The trmmg srgnals produced from this source are:

,/-wr-f-:\\

Figure 6-1

the 1.11 MHz clock by 10.

ROM timing is provrded for strobrng the 8K by e1ght ROMS This 51gnalis produced by delaying the
Phase 2 signal from the 6502 by approximately 160 ns. This allows enough time for the address to
stabhze at the ROMS before strobe tlme

6.2.1.3 6502 Data and Address Interface - This circuit consists of buffers for the address and data
outputs of the 6502. The buffers have .5 mV low-level inputs so that they are eompatlble wrth the 6502
|

driving requirements.

6-2

6.2.1.4 Address Decoders — Addresses for the address bus come from two sources: 1) the 6502 for
normal microcode execution, and 2) the LSI-11 bus for accessmg the CSRs. Accessing the CSRs has
priority over the 6502 addressing requirements. CSRS are in RAM and are accessed by the CPU
through the LSI bus interface.

Decoding of the address bus is accomplished by three sepafate circuits:

Block address decodef A programmable array logic (PAL) is used to decode the ROM

[/0O, and RAM blocks; and to demultiplex the RAM to 16-bit words when CSRs are ac-

cessed by the LSI-11 bus.

1/0 decoder — This decoder breaks a part of the I/0 page up into eight 256-byte .sections

when the block address decoder selects I/0O. No attempt is made to decode to a specific ad-

dress within a section. Thus, multiple addresses w1th1n a given section decode the same de-

vice.
3.

NPR current address decoder — This circuit decodes to a specific address in the RAM address space. Thus, the NPR address registers are mapped directly from RAM. They reflect
the contents of the NPR address locations assigned in RAM (Figure 6-2).

| CONTROL STORAGE

CSR

POINTER

\‘.
P

apbR

RESET

HALT

DMA DECODE AND CNTRL

VERSITAL INTF. ADPT

BAUD

DECODE | USYRT

GENERATION

20 X 56K

CLOCK

—

MAIN MEMORY

o
|‘“:“"|

INTERRUPT LOGIC

| ADDR

MICRO-

|PROCESSOR

‘

16 BIT ADDRESS BUS

—

(8)

DATA BUS
‘

,
MK-2622

Figure 6-2

Control and Address Decoder

6.2.2 1/0 Data Bus
This block contains the USYRT, USYRT control, and line 1nterface control. See Figure 6-3.

6.2.2.1 USYRT - The USYRT is an LSI subsystem for synchronous communications. It provides the
necessary logic support by way of parameter registers for DDCMP. Within this discipline a wide range
of support such as programmable error detection, character recognition, complete serialization, deserialization, and buffermg of datais provided.

6.2.2.2 USYRT Control — This circuit consists of a 74LSZ45 data transceiver, a 741.S373 tristate input data latch, a 741.S244 tristate output data buffer, a 74L.S373 tristate address latch, and three

6-3

741.S74 controlling flip-flops and associated gates. For discussion, the operatlon of the USYRT control
logleis d1v1ded into wr1te and read operatlons See Figure 6- 4

Write — The 6502 asserts the address, selects the USYRT, and generates the write signal approximately
140 ns after Phase 1 high is asserted. The data is available from the 6502 approximately 115 ns after
Phase 2 high is asserted. The USYRT on the other hand, requires data 50 ns prior to the assertion of
data port enable (DPENA). To achieve the necessary timing relationship, the address and data are
latched into buffers and strobed into the USYRT by the eontrolhng flip-flops which are clocked by
signals generated from Phase 1 and Phase 2 timing. Phase 2 is used to gate DPENA to guarantee the
USYRT minimum requirement of 250 ns for DPENA.

Read — Again the 6502 asserts the address and selects the USYRT asina write cycle. The USYRT is

strobed by the controlling flip-flops and data is made available from the USYRT. In a read operation,
the address select lines must be held for 30 ns after DPENA (the 6502 does not guarantee this). Therefore, the addressis latched when Phase 2 highis asserted.

SWITCH

ogic

[reeess——n

(8
(8)

201

BO2

SELECT

o
USvAT

SATEIE),

CONTROL

USYRT |

R/W

(8
(8)

SR
[ Tso

ADDRESS (3)>

ADDRESS (3) >
MK-2499

Figure 6-3

I/O Data Bus

6-4

eU_I|I_bIE——8:&,5.,\*.e

[ioa1IL|1|1_

|T1.<—LHH§VAAdSSNoN3wIHdo1Oau3S5T/s~5 N1.O)B.[IJLIY1AM\aHIaLwVAI1—"[ |{ O/LVIv|jG|Y55oanLvaav4ay{\|| [. 1
(0523

6-5

6.2.2.3

Line Interface Control — The line interface control section of logic can be broken into two sub-

Switch logic — This section consists of two sets of switches and their associated buffers. One

sections: 1) switch logic, and 2) programmable interface adapter.

set of switches has its configuration latched into its butfer at boot time. The configuration of
the other set of switches is latched at DDCMP interface selection time.

Programmable interface adapter’ (also referred to as VIA) — This circuit consists of a 6522

chip and is used to control and monitor the various interface signals to the modem interface
|

logic.

Referring to the 6522, the PB section and bit O of the PA section are used as an output register only.
Bits one through seven of the PA section are used as an input register. CAl and CA2 are used to monitor, by way of the 6502 interrupt, modem ready high and clear to send high. PB7 is used to generate
modem clocks when self-testing with loopback connectors. CB1 is used to produce eight clock pulses at
a time when instructed to do so by the 6502 microcode. These clock pulses are used to flush the USo

YRT receiver of data after carrier drops.

DMVI11 Memory

6.2.3

This block may be divided into three sections as follows.

ROM control storage, |

NPR in/out registers.

"
~

RAM,

6.2.3.1 ROM Control Storage — ROMS are used for storing operation codes for the 6502 micro-

Processor.

In order to provide the most immediate access to data for the 6502, 200 ns ROMs are used and the

ROM clock is continuously applied to the chip enable (CE) pin of the ROM. 745241 buffers are used

on the output so that no more than 9 ns additional delay is introduced.

6.2.3.2 RAM - This is the data memory for the DMV11. It is organized functionally as shown in
the RAM into even and odd sections for the sole purpose of having
Figure 6-5. The hardware organizes
16-bit CSRs. When CSRs are accessed from the LSI-11 bus, even, odd, or both sections of the RAM
are enabled. Two 741.S245 transceivers are used to enable and direct data to and from the LSI-11 bus.
The other two 7418245 transceivers are used for drive buffering and to disable the RAM from the 8-bit
microprocessor data bus when the CSRs are accessed by the LSI-11 bus. The microprocessor is halted
when the CSRs are accessed by the LSI-11 bus.

6.2.3.3 NPR In/Out Registers — This circuitry'consists of two 16-bit registers (one for address in, and

one for address out), and two 16-bit registers (one for data in, and one for data out). Extended address
capabilities are included in two scratchpad registers which are four words deep by four bits wide.

to set up the address for an
The microprocessor loads the appropriate NPR ‘addre’ss‘register (in or out)
NPR data transfer. This address is then enabled onto the LSI-11 bus during the address enable cycle.

During a read cycle, the data-in register is loaded from the LSI-11 bus and read by the 6502. During a

write cycle, the data-out register is loaded by the 6502 and read by the LSI-11 bus.
6-6

HEXADECIMAL

0000
-

e
SCRATCH PADS

16 BYTES

Q-BUS CSRs

—

SCRATCH PADS
|
A

8 BYTES

—

OUT NPR ADDRESS

‘

3 BYTES

—

SCRATCH PAD

1 BYTE

IN NPR ADDRESS

o
80

SCRATCHPAD

32 BYTES
—

3 BYTES

BYTE

' GLOBAL STATUS SLOT

—

SLOT MAPPING TABLE (SMT)

1FF

MICROPROCESSOR STACK
BUFFER AND OUTPUT QUEUE

c00

98 ENTRIES

TRIBUTARY STATUS SLOTS

—

64 BYTES

256 BYTES

64 BYTES
—

—
o

8 BYTES/ENTRY

12 ENTRIES

B
64 BYTES/ENTRY

800
MK-2495

Figure 6-5

6.2.4

Data Memory Organization

LSI-11 Bus Interface

The circuitry in this block interfaces the LSI-11 bus to the 6502 mlcroprocessor (see Flgure 6- 6) It
consists primarily of the:

LSI-11 bus DAL 1nterface
CSR controller,
~ Interrupt controller, and
NPR controller.

6.2.4.1

LSI-11 Bus DAL Interface — This circuit consists of four DC0O0S5 chips which interface the

[.SI-11 data and address lines to the DMV11. The device address of the DMV11, which may be anywhere in the 1/0 page, and its associated vector address, is selected by switches on the input to the
DCO005 chips.

The DCO005s are initially in the receive mode (receiving from the LSI-11 bus) in anticipation of an
‘address match from the LSI-11 bus. When a match occurs, the collector-ored match lines go high and
enable the CSR control circuit. In addition to detecting an address match, the DC005s in the receive

mode pass address or data to the DAL lines. In the transmit mode the DCO005s pass data from the
DAL lines to the LSI-11 bus.

6-7

INTERRUPT

LOGICAND }——

/
67

CONTROL
L]
DCOO5
|

LSI-11 BUS

CONTROLLER

DCOO05’s

CONTROL

INTERFACE

\pr

INTERNAL DATA/ADDRESS BUS (16)
RPLY

LSI-11 BUS

Lsl-11 BUS

CSR

)

| conTrOLLER L EE
~

|ouTHB

MK-2498

Figure 6-6

LSI-11 Bus Interface

CSR COntroller _ This circuit consists of a DC004 which:

6.2.4.2

Supphes outputs Wthh indicate when CSRs are selected. CSRs are decoded when the enable

pin is asserted high and BYSNC strobes on its negative edge.

Supplies outputs mdlcatmg when a word of datais to be placed on the L.SI-11 bus, or when a
byte or word of datais to be read from the bus. Byte operations are controlled by the OUT
HB and OUT LB signals.

Asserts BRPLY approximately 30 ns after information is placed on the LSI-11 bus. BRPLY
is then apphed toa sequencer which halts the 6502 and accesses the partlcular CSR mapped

in the RAM.

Applies the BRPLY from the sequencer to the LSI-11 bus by the LSI-11 control l1ne interface

6.2.4.3 Interrupt Controller — This ClI‘CUlt consists of a DCOO3 1nterrupt controller ch1p and two
74LS74 flip-flops. The DC003is usedin a typical manner for interrupt serv1c1ng of the L.SI-11 bus. The
two flip-flops hold the interrupt request (one for A and one for B) until it is serv1ced When either
|
o
request 1s serviced, both flip-flops are cleared by the vector signal.
6-8

IN WORD

eT—

6.2.4.4
NPR Controller — The DC010is usedin this application for doing direct memory accesses to
the LSI-11 memory. Only single NPRs are allowed; HOG modeis not 1mp1emented A description of
the NPR operation follows.

The microcode sets up the NPR current address and data-out registers, and then sets the A flip-flop by

‘writing to the NPR register with bit 6 equal to zero. Once the NPRis initiated, the DC010 handles the
sequence of enabling appropriate registers to transmit or receive data from the LSI-11 bus. This operation 1s sequenced by the 5 MHz clock input to the DCO10.

When the NPRis honored, the leading edge of RPLY releases the NPR so that a second request does
not occur. The trailing edge of RPLY sets NPR busy H to zero to indicate that the NPR transfer is

complete. However, when doing data-out transfers, the NPR data-out register must not be updated for
100 ns after the trailing edge of RPLY in order to comply with LSI-11 bus specifications. The microcode can immediately service the data
-in reglster when NPR busy His cleared durmg a data
-in transfer.

are exceeded, the transfer is aborted. The timer is set each time an NPR is initiated.

The NPR abort timer is used to ensure thatan NPR transfer does not take more than 16 ,uS. If 16 us
6.2.5

Memory and Reset Control

CSR access control allows the LSI-11 bus to access the CSRs. The operatron is as follows.
1.

The CSR controller (Section 6.2.4.2) asserts the CSR L signal when a CSR is selected. This

signal 1s used by the processor halt circuit to halt the microprocessor on the next Phase 2

cycle if a write is not in progress. When the processor has halted, the 741.S164 shift register

1s enabled.

When the first output of the shift register is true, the appropriate CSR address is selected
and the direction of transfer is determined by the state of the WRT RAM L signal from the
PAL. WRT RAM L sets the direction of the RAM transceivers (Section‘ 6.2.3.2).

During the next Phase 2 cycle, the CSRis either read and its contents placed onto the LSI-

11 bus, or the data on the LSI-11 busis written into the CSR

The next Phase 2 cycle terminates the write cycle ifin the write mode, and asserts BRPLY to

the LSI-11 bus.

After RPLY from the DC004 drops, the microprocessor address and data bus are again con-

trolled by the 6502. When CSR selectis dropped the microprocessor resumes operation.
Memory and reset control also generates signals for master resetting the DMV11 and haltmg the microprocessor.

6.2.6

Modem Interface

The modem interface consists of line receivers and drivers for all modem data and control signals. The
interface supports RS-232-C, RS-423-A, V.35, and the integral modem. Circuitry to accommodateinternal loopback for test purposes is also provided. Because the DMV11 supports RS-423-A for category

1 signals (except test mode and ring), dummy generators are used for the following signals.

Seclect frequeney,
Terminalin service,
New signal,

SRTS,
Remote loop,

Local loop, and
Select standby.

6-9

Only one interface can be enabled at a time. The modem interface select circuit enables an interface as
selected by the interface select switch. On power-up and during any reset operation, the selected interfaceis disabled and loopbackis selected until deselected by the microcode (Section 6.2.2.3).

Interface to the outside worldis 1mplemented with two 40-pin Berg connectors J2 1sused for RS-232-C
andRS-423-A with a BC55H-type cable. J1 is used for V.35 w1th the BCO5Z cable or for the integral
modem with the BC55F cable.

The modem interfaces all have a null clock thatis switch selectable for speeds of 56K b / sor 19.2K b/s
627

Integral Modem

and controlled by the PIA.

The integral modem is used for local commun1cat1ons and 1S transformer coupled to twinax or triax
cables for common mode reJectlon and common mode Voltages up to 500 V. For discussion purposes
the mtegral modem1S descr1bedin two sect1ons receive (Frgure 6 7) and transmrt (Figure 6- 8).

6.2.7.1 Receive — The recerved data enters the modem through an isolation transformer whose output

is directed to a differential amplifier to eliminate common mode noise. The amplifier’s second stage
uses an active Buterworth filter with an added passive filter for high and low cutoff. The filter’s complimentary outputs are input to a comparator which detects zero crossover. Positive and negative transitions from the comparator clock the UP and DOWN flip-flops. All clockingis done at a clock rate 20
times the b/s rate and the UP and DOWN fhpflops latch until cleared by the transitions (TRANS)
o
|
o
|
flip-flop.|

When either the UP or DOWN flip-flopis set, the next clock pulse loads the transitions (TRANS) flip-

flop which then clears the UP or DOWN flip-flop and holds it clear for one clock time. The clock input
to the TRANS fhp-flop and rece1ve counter (REC CNTR)is 20 times the data rate clock time.

The REC CNTRis clocked at half clock time (or inverted 20X clock), and counts 16 clock times and
sets the 3 /4 time flip-flop. The counter is loaded if the TRANS flip-flop is set and 16 clocks have
occurred since the last load. The counter enableis true except in an overflow condition or when operating in half-duplex mode with the transmitter active.

When 3/ 4 T is set, and the time between transrtrons 1S greater than 16 clock times, the IDATA flip-flop

is clocked to one. If the transition time is less than 16 clock times when 3/4 T is set, the IDATA flip-flopis clocked to a zero. The minimum time between transitions 1s .05 to . 10 bit times as determined by
the TRANS flip-flop clearing the UP and DOWN flip- flops. The next 3 /4 T clock loads IDATA into
the RI DATA flip-flop. The next 3/4 T clock ANDs RI DATA with IDATA, and if IDATA is zero,
sets the I CARRIER flip-flop. IDATA is input to the receive FIFO and I CARRIER is gated with
LINE UNIT STEP to become GRX CLK.

The overflow flip-flopis set when no transition occurs within one and one-half bit times. Overflow then
sets IDATA which allows the next clock pulse to clear the RI DATA and I CARRIER flip-flops until
the next sequence of one followed by two Zeros occurs.

6.2.7.2

Transmlt — The transmlt c1rcu1t encodes the data into d1phase space; in a square wave se-

quence. The output is 6 V peak-to-peak into a 50 ohms load and does not exceed 15 V peak-to-peak into
|

an open circuit.

RTS allows TI CLK to set the ICS flip-flop. When the ICS flip- flopis set, the encoder flip-flop (ENC)
is allowed to toggle with each data or TI CLK. The encoded output feeds a bipolar line driver that
generates an ac signal with zero crossover points. The line driver output is connected to the protection
»

transformer.

6-10

The +5 V low circuit turns off the transistors on low logic power to keep the transmitter from generating noise or from loading the line. During power-up, this circuit keeps the modemin the disabled state
for several milliseconds to prevent the transmission of nonsense characters that would interfere with
transmission in progress on a multipoint line.

The transmitter is disabled when line units are not transferrlng data. The transmltter does not load the
line when power is off.
BERG

DIFFERENTIAL

‘

“

—

AMPLIFIER
!

gigA “ TRANSFORMER
ISOLATION
IN

FF's

CONN.

DIFF EFILTER
,
T
,
l
L

—

gg?gfl

" CROSS- |
'
OVER

‘

—| DowN

TRANS

COMPARATOR

H
20X CLK

REC
CNTR

[

]
'

OVF
I

|
—CLK

Rl DATA
SYNC

UP
&

DATA
co

ox CLKH

| CARRIER

——\

____//

MK-2500

Figure 6-7

Integral Modem Receive

6-11

'RTS

ICS

Tl CLK

TX DATA

_-l

LOOPBACK

Bl POLAR

LINE

DRIVER

XFORMER
+5V LOW &

DATA
OouT

T
MK-2503

Figure 6-8 Integral Modem Tr'anvs'mit

/"‘"‘-\\\

+5V

//'dm\\_.

ENABLE

CHAPTER 7
SERVICE

7.1

SCOPE

This chapter provides information for servicing the DMV11. It includes the maintenance philosophy,
troubleshooting techniques in a multipoint environment, maintenance functions, preventive maintenance, and corrective maintenance. The section on troubleshooting techniquesin a mu1t1p01nt envrronment 1ncludes
e

The general overall approach to multipoint troubleshooting.

Some common problems associated with different multipoint network configurations

The use of error counters and other 1nformat1on for 1solating problems to a spe01f10 portion of

the physmal link.

\v/

The corrective maintenance section contains brief descrrptrons of the diagnostics a33001ated with the
DMVI11.

7.2
MAINTENANCE PHILOSOPHY
The field replaceable unit (FRU) for the DMVI11 is either a module (M8053 or M8064 microcontroller/line unit) or cable. Training of field service personnel is directed to functional and apphcation troubleshootlng, using diagnostics, for fault isolation to the FRU. Spare parts for module repair
are not stockedin the field. Typical applications of the DMV11 do not permit lengthy troubleshooting
sessions, and component troubleshooting/repair requires at least a 16-channel logrc analyzer
|

7.3

TROUBLESHOOTING TECHNIQUES FOR MULTIPOINT

Because of the complexity of some multipoint network configurations, there is a potential for using

valuable time in trying to isolate a problem. For this reason, troubleshootlng technlques for mult1p01nt

networks differ from those for point- to-pomt networks

The following sections discuss these troubleshooting techniques in terms of approach and error
counters.

7.3.1

Approach

Before attempting any corrective measures, it is 1mportant to get some basic 1nformatron about the
network configuration and the nature of the problem This information can be obtained by querying the
user and by referrlng to the topology diagram for the network. The topology diagramis generated at
installation time andis maintained by the field service representatrve The flow chart (Figure 7-1) illus-

trates a typical approach to troubleshooting from the time a service callis placed, until corrective maintenance is begun. This procedure should be followed to help isolate a failing tributary before anyone is

dispatched to a site.

CALL RECEIVED

FROM CONTROL
STATION

¢ OTHER STATIONS p———3{

\\WORKING
Tves

GATHER THE FOLLOWING

R—rd

FAILING TRIBUTARY

" ANY TRIB TM\

€

AT THIS SITE

CALL MODEM

SERVICE REP.

N\ WORKING

RELOAD TRIB SOFTWARE
AND INITIALIZE POLLING
FROM CENTRAL STATION

* RELOAD TRIB SOFTWARE |
“AND INITIALIZE POLLING |
FROM CENTRAL STATION |

#" THE TRIB

DTRON

WORK
\\_ NOW _~

(MODEM)

| YES

/DOE
N\N_
S
THETRIB

\\_WORK NOW

{

N _NO

/~ RX PROBLEM
OR
IN MODEM

\. LINE.CALL SERVICE _

GO TO FAILING

C ”J

STATION

MK-2398

N .-Figure'7w-_1 Ex,amplebf a Typiclallsola'tion._ Flow.vDiagram (Sheet*l of 5)
7-2

)

The transmitter is disabled when line units are not transferrlng data. The transmltter does not load the
line when power is off.

BERG

CONN.

DIFFERENTIAL

AMPLIFIER

DIFF EFILTER

TRANSFORMER

”

SETSCT

crossaverm

‘

|
|

FF's

up
1

——

TRANS

—

DOWN

COMPARATOR

gigA “ ISOLATION

20X CLK H

REC

)

CNTR

[

OvVF

SYNC
UP

|
e

RI DATA

DATA
coj—

»ox CLK H

| CARRIER

———\

MK-2500

Figure 6-7

Integral Modem Receive

6-11

‘

RTS

ICS

Tl CLK

TX DATA

-_l
CL\R

)

LOOPBACK

- Bl POLAR
-

LINE

" DRIVER

—:j XFORMER

+5V LOW &

—

DATA

ouT

ENABLE

+5V

1
MK-2503 .

Figure 6-8

Integral Modem Transmit

RELOAD SOFTWARE AND
RESTART AT CONTROL

AND TRIB STATIONS

‘

ARE
ALL TRIBS

WORKING

/CALL MODEM
SERVICE REP.

WORKING

| YES

ARE

ANY TRIBS

GO TO CONTROL

~NO

STATION

CALL MODEM

'SERVICE REP.

” DATA LAMP ON

AT MODEM

ANY TRIBS
RECEIVING

\\POLLING

TX PROBLEM IN
MODEM OR LINE.

CALL SERVICE

MK-2399

Figure 7-1

Example of a Typical Isolation Flow Diagram (Sheet 2 of 5)

7-3

.~ BLINKING AT TM\, _
\CONTROL SITE,

RX PROBLEM IN
MODEM OR LINE

CALL SERVICE

_~"CD AND TX "\

DATA STEADY
N\

ON?

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

Example of aTy’pvi,cal Isolation Flow Diagram (Sheet 3 of 5)
N

~ Figure 7-1

O
IS

RTS COMING

_NO

>
|

YES

GO TO FAILING

STATION

" CTS COMING

Y
CALL MODEM

YES

SERVICE REP.

DATA COMING
ON

YES

TX PROBLEM IN
MODEM OR LINE
CALL SERVICE
MK-2401

Figure 7-1

Example of a Typical Isolation Flow Diagram‘ (Sheet 4 of 5)

¢ RELOADED AND >

“\\RESTARTED,”

{
RELOAD SOFTWARE AND

| vES

RESTART CONTROL

“OR MULTIPORT > N0, ¢

“ OBVIOUS

\\_NO

MODEM

DOES TM\
"

FAULT

THETRIB

_WORK NOW

YES:’
BYPASS SPLITTER OR

CALL MODEM

USE DIFFERENT PORT

SERVICE REP.J

/”DOES
TM\,
THE TRIB

\\_ WORK NOW

OR
HAVE SPLITTER

MODEM REPLACED

" INITIATE CORRECTIVE
MAINTENANCE

" PROBLEM
\\_

FOUND

f cAL

>\_ SUPPORT

)
MK-2402

Figure 7-1

Example of a Typical Isolation Flow Diagram (Sheet 5 of 5)

7.3.2

Error Counters
.
o
|
:
|
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 DMV11

uses

error counters. Every DMV11 in the network (the control station and each tributary) uses error

counters to record errors. This allows user programs at any DMV
11 in the network to determine overall
error rates and to detect a malfunctioning link.
3
|
|
B
:

The main way in which errors are indicated to the DMV11 is by DDCMP negative acknowledge messages (NAKSs). Each NAK contains an address field and a reason code that identifies the source and
reason for the NAK. In general, when an error is detected in an incoming message, the station that

receives the message sends a NAK to the station that sent the message. By recording NAKs sent and
NAKSs received, each point or tributary
in the network is able to compile statistics on the condition of

the link established 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 DMV11; 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 located in the tributary status slots
(TSS) of the data memory (Figure 7-2). Station counters are maintained for the physical link as a
whole, and are located in the global status slots (GSS) of the data memory (Figure 7-3). The information gained by checking error counters may be helpful in pinpointing a problem area.
7.3.2.1
Data Link Error Counters — Data link counters are of two types; cumulative and background.
The cumulative counters are 8-bit counters which latch at 255. The background counters are 16-bit
counters which latch at 65535. The cumulative data link counters record and total all occurrences of an
error and group them into the following categories.

Data errors outbound,
Data errors inbound,

Local reply timeouts,

Remote reply timeouts,
Local buffer errors,
Remote buffer errors,

Selection timeouts.

Background data link counters
are used to provide a statistical base for the cumulative error counters

and therefore record:

o
e

The number of data messages transmitted,

The number of data messages received, and

The number of selection intervals.

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 established tributary. Data link
counters are cleared by:

e
e

A master clear of the DMV11,
A control command to establish the tributary, or

o
A user-issued control command to read and clear the TSS error counters.

Data Errors Outbound - This 8-bit group counter 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
(ODBCC), 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 re-

)

ceived for a header block-check error for either data or control messages.

ODBCC(outbound data field blockcheck)is set when a NAK with a reason code of two 1s
received for a data field block-check error.

- OREP (outbound reply response)is set when a NAK with a reason code of three1s recelved
for a reply message response.

TRIBUTARY STATUS SLOT (TSS)
ADDRESS (OCTAL)

RESERVED
RECEIVE THRESHOLD ERRORS
>

' TRANSMIT THRESHOLD ERRORS

SELECTION THRESHOLD ERRORS

—

DATA MESSAGES TRANSMITTED

—

DATA MESSAGES RECEIVED

—_—

—

SELECTION INTERVALS

DATA ERRORS OUTBOUND

| OREP I ODBCC I OHBCC

RESERVED

'DATA ERRORS INBOUND
RESERVED
I»IREPI IDBCCI IHBCC |

LOCAL BUFFER ERRORS

RESERVED
15

)

SELECTION TIMEOUTS

RESERVED
17

I RBTS I RBTU
|

I IRTS I NRTS

REMOTE REPLY T»lME.O_UTS |
MK-1960

Data Link and Threshold Error Counters

7-8

)

LOCAL REPLY TIMEOUTS

Figure 7-2

I LBTS i LBTU

REMOTE BUFFER ERRORS

RESERVED

GLOBAL STATUS SLOT (GSS)
ADDRESS (OCTAL)

REMOTE STATION ERRORS

l RSTR I RSEL iRMHFEIROVRN
16

LOCAL STATION ERRORS

lLOVR ILUNDR[LMHFE I LOVRN
17

GLOBAL HEADER BLOCK CHECK ERRORS

MAINT, DATA BLOCK CHECK ERRORS
MK-18589

Figure 7-3

Station Error Counters

)

Data Errors Inbound
— This 8-bit group counter records occurrences which normally result from data
errors on the communications channel inbound to thls station. Three separate bits 1ndlcate speerflc error

types associated with this counter.
e

[HBCC (inbound header blockcheck)is set when messages havrng header block-check 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 multlpornt control station records this error for the
selected tributary regardless of the address fieldin the received message. A multipoint tributary records this error only if the address field matches 1ts station address
¢

IDBCC (inbound data field blockcheck)is set when NAKS with a reason code of two are to
be sent for data field bloek-eheck errors.
IREP (inbound reply response)is set when NAKs wrtha reason eode of three are to be sent

for a reply response.
Local Reply Timeouts - This 8-bit counter records oceurrences Wthh result from:
e

The loss of communications between two stations while the one reeordrng thrs error has data
to transmit, or

The choice of an inappropriate value for the reply timer.

Specifically, this error counter records the sending of a REP message.
‘Remote Reply Timeouts — This 8-bit counter records occurrences which result from:
e

The loss of commumeatrons between two stations wh11e the remote statron has data to transmrt or

The eh01ce of an 1nappropr1ate Value for the remote station reply timer.

Specifically, th1s counter records ACKs sent in response to a REP The remote statron 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.
o

7-9

Local Buffer Errors — This 8-bit counter records the fact that the user program at the station recording
the error failed to properly allocate receive buffers to data messages from the remote station. Two separate bits indicate the specific errors associated with this counter.

e
e

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 t0o small)is set when a local bufferis too small for the incoming mes-

sage. This condition indicates that a NAK with a reason code of 16 is to be sent.

Remote Buffer Errors — This 8-'biteounter records the fact that the user program at the remote station
failed to properly allocate receive buffers to data messages from the station recording the error. Two

separate bits indicate the specific errors associated with this counter.

RBTU (remote receive buffer temporarily unavailable) is set when a NAK with a reason
: .’

code of eight is received

4\\

e
e

RBTS (remote receive buffer too small)is set when a NAK w1th a reason code of 16 is received.

Selection Timeouts — This 8-bit counter recerds the occurrences which result from:
°

Loss of Commumcatlons with a remote station,

The choice of an inappropriate value for this station’s select timer.

Data errors on the communications channel to or from the remote statron and

This counter is used only by half-duplex point-to-point ormultipoint control stations. Two separate bits
indicate the specific errors associated with this counter.
e

NRTS (no reply to select) is used to record selection intervals in which no transmission 1s

IRTS (incomplete reply to select) is used to record selection intervals which were not properly terminated. Specifically, it records the expiration of the select timer preceded by receipt
of a valid control message, recelpt of a valid header, or detection of an attempted transmission. An attempted transmission is indicated by:
—

The presence of a carrier signal,

—

The receipt of an SOH, ENQ, or DLE.

—

The receipt of a DDCMP synchronization sequence, and

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,
a result of retransmission are not inlocal reply timeouts, and remote buffer errors. Messages sent as
cluded in this count.
|

Data Messages Received— This 16- bit counter records messages recelved 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. Messages received out of sequence or in error are not includedin this
count.

%aa

7-10

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 half-duplex point-to-point and multipoint control stations. It
can be used as a statistical base when evaluating the number of selection timeouts.

7.3.2.2

Station Error Counters — Station counters are 8-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, or

A data error on the communications channel undetected by the header block-check field.

A single set of these counters is used for all tributaries on a multipoint link.

There are four types of station counters:
Remote station errors,
Local station errors,

Global header block-check errors, and
Maintenance data field block-check errors.

Station counters are cleared by:
A master clear of the DMV1I1, or
|
A user-issued control command to read and clear the GSS error counters.

Remote Station Errors — This 8-bit counter records occurrences caused by a fault in a remote station or

by an undetected data error on the channel inbound to this station. Four separate bits indicate the specific errors associated with this error counter.
|
ROVRN (remote receive overrun) is set when a NAK with a reason code of nine is received
for a receive overrun.

RMHFE (remote message header format errors) is set when a message is received which has

a header 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 containing an address field which does not match the address of the currently selected tributary. This error is recorded only by multipoint control stations.

RSTR (remote streaming tributary) is set by either one of two events: 1) an implementationdependent maximum transmission interval is exceeded without releasing the channel (babbling tributary), or 2) the channel is not released following the end of a selection interval
(streaming tributary).

Local Station Errors — This 8-bit counter records occurrences caused by a fault in a local station or by
‘an undetected data error on the channel outbound from this station. Four separate bits indicate the
specific errors associated with this error counter.

LOVRN (local receive overrun, NAK sent) is set for local station receive overruns: This con-

dition indicates a NAK with a reason code of nine is to be sent.

7-11

LOVR (local receive overrun, NAK not sent)is set by a receive overrun when a NAKis not

- sent. For a multipoint tributary, this happens 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 with a header format error was sent by this station

Global Header Block-Check Errors — This 8-bit counter records the occurrence of header block—check

errors that are not recorded on a per tributary basis. Specifically, it counts header block-check 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 8-bit counter records the occurrence of data field
block-check errors for maintenance messages.

7.3.2.3

Threshold Error Counters — Threshold error counters are used to determine if a persistent fault

exists. A pers1stent faultis one which occurs seven consecutive times. Whenever a threshold counter

reaches its maximum value (7), the user program is notified by a control response.
In the DDCMP run state, threshold counters are cleared when the user is notified. In this way the user
is continually informed of a persistent fault. In the DDCMP ISTRT and ASTRT states, threshold
counters are not cleared when the user is notified. In this way the user 1S not continually 1nformed of an
inoperative remote station
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 addressesin which case it maintains a smgle set for each estabhshed tributary address.

There are three types of threshold error counters: transmit, receive, and selection.
Transmit Threshold Errors - This 3-bit counter is incremented (if less than 7) in the following instances.

The DMVI11 is in the ISTRT state when a STRT message is sent,

The DMVll is in the ASTRT state when a STACK message 1s sent, or

The DMVI1 is in the Tun state and a NAK with a reason code other than three (REP re-

sponse)is received, or when this station sends a REP message.

The transmit threshold error counter 18 cleared
'

Upon entering the ISTRT, ASTRT or run states.

Whilein the run state one of the following occurs.

~ A transmit threshold-fi.,error»is r.eport:ed N
—

A NAK, ACK or data message 1S received acknowledging a new message or

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

7-12

Receive Threshold Errors — This 3-bit counter is incremented (if less than seven) when a NAK with one

Reason Code

Description

NI =

Header block-check error.
Data field block-check error.

OO0 W

of the following reason codes is sent.

REP response.

Buffer temporarily unavailable.
Receive overrun.

— A\

)

Message header format error.

This counter is cleared when:
e

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, or

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

Selection Threshold Errors — This 3-bit counter is only used by multipoint control stations and halfduplex point-to-point stations. It is incremented (if less than seven) when a selection timeout occurs.
It is cleared upon receipt of a message with the select bit set, or while in the run state and a selection
threshold error is reported.

17.3.3 Error Counter Analysis

In most applications the software operating system records all the error counters, but does not attempt
to analyze or take any particular action of its own as a result of any specific errors.

The system manager or another operator instructs the software to retrieve the counters. The counters
are analyzed by the operator or system manager and then the software is instructed to perform a specific function relative to the counter indications.

For information on retrieving the counters, refer to the system-specific DECnet System Manager’s
Guide or consult the network manager.

The following example serves to illustrate how the counters can be used in diagnosing a system failure
(refer to Figure 7-4). A user should keep in mind that DCLT also allows access to error counters. See
Section 7.6.5.

Assume that a full-duplex multipoint network is made up of seven tributaries and a control station as
shown in Figure 7-4. The type of electrical interface (EIA RS-232/RS-449, V.35, integral modem) i1s
not important in this example. The system manager at the control station notices that data transfers
over the network appear to be “sluggish.” Some standard file transfers are taking a longer time than
usual to complete. (Note that in some cases this could be caused by a sudden increase in traffic on the
network.) The problem appears to be intermittent in nature since no threshold errors have occurred and
data is still being transferred between all tributaries. The system manager examines the error counters
at the control station for tributary seven. They are as follows.

7-13

|-_AJ"_§_LAXb=HO4)I1dINVX3
-2InSig4-L—xo[dnq-|fJiYngOMuLoAaNdsV1Avraein-qui],jutodnjyy|IomioN

T|BAiLw1x|To|]yiaw1yxl|A|o€|S|‘lO|HL2nI—.SNoOgw€4 DLyflu9mx[|ASvE|r7oLu[jNuonx||twgxyg2 nlou(V|NOHoI[O|LIDH3I'NJNONxDVLS|ISaT1I4]H.1Ao||Juwa1xL9 ylLrod
A

| —sL—Ls =S—=L¥—= LI|

7-14

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:
e
e
e

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 acknowledg-

ements for outstanding 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 gathered so far, there appears to be a problem on the link
between the tributary’s transmitter and the control station’s receiver. This is reenforced by the fact that
there are no data errors outbound indicating that NAKS have not been received from the tributary. The
problem could be in the control station’s receiver hardware (DMV11, modem, and so forth), the cable
to the tributary, or the tributary’s transmitter hardware (DMV11, modem, and so forth). The next logical step is to examine the error counters at the control station for the other tributaries. Upon doing this
the system manager notices 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 could be that both tributaries six and seven have transmitter problems, but this
is unlikely.) Examination of the error counters for the remaining tributaries reveals 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 cases all stations are
affected.

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 could also 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 gained 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 TIMEOUTS = 40
LOCAL BUFFER ERRORS =0
REMOTE BUFFER ERRORS = 0
SELECTION TIMEOUTS = N/A

DATA MESSAGES TRANSMITTED = 310
DATA MESSAGES RECEIVED = 420
SELECTION TIMEOUTS = N/A

Notice that these error counters provide the same type of information as the control station error

counters. The errors all seem to be in one direction, namely outbound. Notice also that the number of
data messages transmitted and received match those at the control station. Since data transferis occurring, the likelihood of a hard error in the hardwareis 1ow

NOTE

Hard errors at the control or tributary statlons generally are revealed during the start-up sequence.

Examination of the remainder of the tributary’s error counters reveals that tributary six has errors similar to tributary seven, and the other tributaries have very few errors. The “sluggishness” of the network
is due to the:
|

Timeouts that intermittently occur when pOHing tributaries six and seven, and

Numerous retransmissions that occur in order to get the data vtransferred. |

These timeouts and retransmissions of data are the result of a cable fault between tributary five and six
|

as indicated in Figure 7-4.

7 4 MAINTENANCE
B
For mamtenance purposes the DMVII may be operatedin two bas1c modes

7. 4 1

Maintenance mode, or -

Standard operating mode.
Maintenance Mode

\\\.._.//

Mamtenance modeis used to test the DMV11] by causmg the mlcrocodeto respond to certam command

functions 1ssued by a diagnostic program. Section 4.8 discusses the maintenance modein detail.
7.4.2

Standard Operatmg Mode

The DMV11

is testedin the standard operating mode by termlnatmg the cable at the BC55F or BC55H

panel or at the modem end of the external eable and runmng dlagnostles

The DMV11 options are conflgured as follows

1. DMVI1I-AA (forRS-232-C; RS-423-A)
° _' The modem must be dlsconnected and the test connector must be attached to the cable.
This may be done at the BC55H panel or at the modem end of the external cable

Test Connector h i Interface Cable |

RS-232-C

H325

RS-423-A

H3251

Modem Cable -

BCO05D-25

BCSSH

BC55D-33

BC55H

The clock signal is looped back in the test connector to simulate modem transmit and
-

EIA

receive clocks. The data rate for this application must not exceed 19.2K b/s.

Modem control signals are tested for proper level conversion and cable paths. These

signals are looped backin the test connector as shownin the signal flow of Figure 2-8
View E.

)

DMVI11-AB (CCITT V.35/DDS)
e

The modem must be disconnected and the H3250 test connector must be attached to

the BC05Z-25 cable.
e

The clock signal is looped back in H3250 to simulate modem transmit and receive

clocks.
e¢

DMVI11-AC (for integral modem local use)
e

The local link connections of the BC55F connector panel are disconnected at the local

panel, and the FDX switch on the BC55F connector panel is switched to half-duplex to
accomplish the external loopback.

]

Modem control signals are tested for proper level conversion and cable paths. These
signals are looped in the H3250 as shown in the signal flow of Figure 2-8 View B.

CAUTION
If DMV11is connected to another running DMV11

disconnect the cable at the BC55F connector panel
during diagnostic execution.

The data is looped back through the BC55F connector panel to test, transmit, and receive data. The data rate for this application must be 56K b/s.

7.5
PREVENTIVE MAINTENANCE (PM)
There 1s no specific DMV11 PM schedule. A general check of voltages and connections should be done

when system PM is performed. After handling DM V11 modules or cables, a complete checkout of the
device, by running all diagnostics and, if possible, the interprocessor test, is required.

7.6

CORRECTIVE MAINTENANCE

Since the FRU is either a module or cable, all corrective diagnosis should be directed towards isolating
the failing FRU. DMV11 diagnostics are designed to aid in the isolation process and should be run
starting with the DMV11 static logic test and continuing to the DEC/XII program. The proper se-

quence of diagnostics is shown in Table 7-1.

7.6.1

DMV11 Static Logic Tests Parts 1 and 2

These diagnostics test the DMV11 microcontroller circuits except for the USYRT. Through dlalogue
with the operator and by using the diagnostic supervisor (DS), the program allows modification of device parameters, such as the LSI-11 bus address, vector address, and processor type.
These programs are compatible with the stand-alone diagnostic supervisor and do not exceed 16K of
memory. The total time required to run DMV11 static tests is approx1mately from 30 seconds to 2
minutes per pass, depending on the CPU type.
DMVI11 static logic tests part 1 and part 2 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. A summary of the tests performed are listedin Tables 7-2 and 7-3.

71-17

Table 7-1
- Diagnostic

Description

DMYV11 Diagnostics

DMV 11 static logic test part 1
- DMV11 static logic test part 2
DMV11 static logic test part 3
DMV11 static logic test part 4
DMV
11 static logic test part 5
DMV11 functional diagnostic
DMV11 DCLT program
DMV11 DEC/XII master module
DMV11 DEC/XII slave module

Indicates
the revision level

Table 7-2

DMV11 Static LOgic" Test Part 1 Diagnostic Summary

Test Number

| .Descrnptlon

DMVI1l1 avallablhty
Master clear, run mlcrodlagnostws
CSR addressmg
CSR registers data read/write
Basic master clear
Bus reset

CSR, maintenance mlcrocode 1nteract10n |
- Run fhpflop

10
11
12

VIA register addressing
VIA’s DDRB data read /write

14
15

16
17
18
19
20

21
22
23
24
25
26

Low RAM (00- OF) scratchpad )
|
Data RAM moving inversions (LOC’s 0018-01FF hex)

VIA’s DDRA data read /write

VIA’s ORB data read/write
VIA’s timer #1 data read/write

VIA’s shift register data read/write
VIA’s ACR data read/write
VIA’s PCR data read/write
VIA’s IER data read/write
VIA’s ORB/DDRB master clear test

VIA’s DDRB master clear test
VIA’s DDRA master clear test
- VIA’s shift register master clear test
VIA’s ACR master clear test
. VIA’s PCR master clear test
VIA’s IER master clear test

7-18

\\_‘/’, 3

DMV11 Static Logic Test Part 2 Diagnostic Summary

Test Number

Description

—

Table 7-3

VIA timer 2 one shot mode

VIA’s SR input (MODE 2) — system clock mode
NPR control register — master clear
NPR data-out

00 ~JON L

NPR data-in
NPR XFFR abort

NPR extended address bit test

Special MFG extended bit test

Q-Bus interrupt “A” & “B” selection
Bus reset with disable init set
Master clear with disable init spt
DCOK H LO bit
Halt mode verification

10
11
12
13

7.6.2

DMYV11 Static Logic Tests Parts 3, 4, and 5

These diagnostics perform static tests of USYRT read/write logic; basic transmitter functions; receiver
sequencing and data buffering; and static operations in character and bit-stuffing modes. In addition,
data messages are sent 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. Additional tests provide fault isolation to facilitate 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 LSI-11 bus address, vector addresses, and device priority. The
operator can specify particular tests to be run and a variety of looping, running, and reporting modes.
Device errors are reported as they occur. The report includes the test number and error description,

good and bad test data, and applicable device register contents.

A summary of the tests performed are listed in Tables 7-4, 7-5, and 7-6. For greater detail, refer to the
diagnostics listings.

Table 7-4 DMV11 Static Logic Test Part 3 Diagnostic Summary
Test Number

1
2
3
4
5

Description

- TBMT microcode interrupt test
Switch setting test
USYRT master clear test
USYRT program reset test
USYRT register addressing test
R /W bit test of PCSAR high point

7-19

Table 7-4 . ‘DMVII Static Logic Test Part 3 Diagnostic Summary (Cont)
Test

Number

Descrlptlon

R/W bit test of S/AR register

R/W bit test of PCR register

10
11

R /W bit test of TDSR register’s high byte

R/W bit test of TXDB register
Pseudo R/W bit test
of RXDB

Pseudo R /W bit test of RDSR’s high byte

Null clock test

BCP TXreset w/IDLE
= 0

BCPTXreset w/IDLE
=1

16
17
18

|
- BCP TX underrun w/TSOM termination
|
BCP TX underrun w/RESET termination

BCP TX disable test

19
20

-BCP character length test

21
22

- FIFO stacking characters test

BOP TX TABORT/(IDLE
= 0) test
|

BOP TX TABORT/(IDLE
= 1) test

BOP TX TXGA (transmit go-ahead) test

25
26
|
27
|

BOP TX message without CRC

BOP RX character length test
TX “spacing sequence”
FIFO overrun integrity test

BCP PX overrun set and clear test

BCP RX sync-character recognition

- BCP RXstrip-sync test

Table 7 S

Test Number

BCP RX lost RXE test

DMVI] Statlc Loglc Test Part 4 Dlagnostlc Summary

Descrlptmn |

—_—O

— e \D OO0 ~J O\ L

£ W N

- VRC parity generation test
VRC error detection test

BCP CRC generation/detection test
BOP RX basic receive/flag recognition test
BOP RX secondary station addressing
BOP RX all parties address test
- BOP RX bit stuffing test
BOP RX underrun idle aborts/ flags
BOP RX lost RXE test
BOP RX GA (go-ahead) recognition
- BOP RX “ABC” test

7-20

\-'—‘_—//’

DMYV11 Static Logic Test Part 5 Diagnostic Summary
Description
RX data flushing test
Integral modem interface test

O 00 JO\ L A W

Test Number
—

Table 7-6

7.6.3

Data test — BCP external loopback (XLB) CRC-16
Data test — BCP XLB odd VRC
Data test — BCP XLB even VRC
Data test — BOP XLB CRC-CCITT-1
Data test — BOP XLB CRC-CCITT-0
- Modem control signal loopback test
DDCMP message test

DMV11 Functional Diagnostic

This diagnostic performs testing on the DMV11 option in a functional manner to verify its proper operation under microcode controlled use of the DDCMP. This includes a ROM CRC/ CCITT check, microdiagnostic, command utilization, and error generatlon

This functional test provides troubleshooting capabilities such as tight-scope loops, sw1tch options, and
the ability to lock on intermittent errors. Additionally, this program conforms to the stand-alone version
of the diagnostic supervisor andis compatible with APT, ACT, XXDP+, and SLIDE.

Through dralogue with the operator, the program permits modrflcatlon of device parameters such as the
L.SI-11 bus address, vector addresses, and device priority. The operator can specify particular tests to
be run and a variety of looping, running, and reporting modes.

A summary of the tests performed are listedin Table 7-7. For greater detall refer to the diagnostics
listings.

- 7.6.4

DMV11 Microdiagnostic Error Reporting

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 normal interaction with the CPU can take place.

The user program is notified of the results by way of the CSRs. Table 7-8 is a summary of the possible
results.

7.6.5 Data Communications Link Test Program (DCLT)
DCLT is a communications equipment maintenance tool designed to verify DMV11 to DMV11 communication links. The DCLT program provides the coverage necessary to isolate the following faults:
Communications interface program functlonahty,

“Communication modem,

Communication cabling and installation, and

Physical link/network.

7-21

DCLT programs allow testing between modes with different hardware interfaces implementing the
same or compatible protocol. The DCLT program can be exercised under normal maintenance loop-

back tests:

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)

DCLT’s main goalis to test the communications link. DCLTassumes that the CPUs and DMV1l1s at
each end of the link have previously been tested and found to bein proper workmg order.
Prior to analyzing any data, the user must have a thorough understandmg of the protocol formats apphcable to the system under test.
o
|

DCLT may be used to access DMVll error counters or other 1nformat1on by uslng the print command.

The print command 1nvokes a DCLT level called REPORT w1th1n Wl’llCh the followmg commands are
ava1lable

Command

‘Description

HELP OR ?

Prints help information for RPT.

TSS NNN/SW

Shows tributary status slot information where NNNis the decimal trlbutary address and SWis one of the followrng switches.
|

ERROR

Indrcates that only error slots are to be prmted,

FULL

Indicates that all tributary status slots are to be printed.

OFFSET = NN

Indicates that the tributary status slot whose offset is NN is to be printed.

GSS/SW

Print the global status information.

~ Switches are the same as for TSS.

LOG
EXIT

Dumps the event log

Exits back to the command level that the user entered from. IDCLT> or
DP>].

DCLT is XXDP + or APT compatible and runs under control of the d1agnost1c superwsor (DS). It
requires 24K of memory. For more information on DCLT refer to the (C)ZCLM** document.

7-22

Table 7-7 DMV11 Functional Diagnostic Summary

}

Test Number

Description

2-7

Address test

DMP ROM verification tests

8-9

DMV ROM verification tests

Initialization test

DMP interface diagnostics

RDI remains set test

Test for RDO setting

Check for procedure error 100

15
16

|
3
|

Check for procedure error 104
u

Test mode change of duplex portion of mode

17
18
19-20

Test for max tribs to be established
Read/write tributary status slots test
Tests for procedure error 132

Test for read /clear command

Tests for global status slots

Halt trib command tests

- Kill trib command tests

Check for procedure error of 102

Check for procedure error of 110

Check for procedure error of 120

Check for procedure error of 134

29
30
31
32

Latch/unlatch poll check
Short message sending test
Check for procedure error 122
Check for procedure error 124

)

Check for procedure error 126

Check for procedure error 130

Transmit/receive 256 bytes, MTP, DDCMP

DMYV Q22 mode TX/RX 256 bytes, MTP, DDCMP

Transmit/receive 255 bytes, MTP, DDCMP

39-41

Test of mem extension bits

DMP read/write modem register tests

Test for TX/RX 257 byte

Test for TX/RX 1 byte

Polling state tests

7-23

~ Table 7-8

Microdiagnostic Error Codes

BSELG6

BSEL4

Descnptmn

101

N/A

Branch test has falled and the mlcrocode1S spinning in a loop

102

N/A

6502 1nterna1 register test has fa1led and the microcodeis sp1nn1ng in a loop.

103

N/A

Load and store 1nstruct10ns test has falled and the microcodeis spinning in a

104

N/A

Compare 1nstruct10ns test has falled and the microcodeis sp1nn1ng in a loop.

105

N/A

~ Increment and decrement 1nstruct1ons test has falled and the microcode1s spinning in a loop
|

106

N/A

Shift and rotate instructions test has fa11ed and the microcodeis sp1nn1ng in a

loop.

107

N/A

Logic 1nstruct10ns test has falled and the microcodeis spinning in a loop.

110

N/A

Add with carry, subtract w1th carry, set and clear decimal mode instructions
test has falled and the m1crocodeis sp1nn1ng in a loop.

111

N/A

Stack push and pull 1nstruct10ns test has failed and the microcodeis spinning in

a loop.

112

N/A

Subroutine instrlictions test 'has failed and the microcode is spinning in a loop.

113

N/A

114

N/A

Ram scratchpad CSR and NPR address resisters data test has failed and the
-mlcrocode1s sp1nn1ng 1n aloop.

115

N/A

True 1nterrupt test has failed and the mlcrocode1s spinning in a loop.

116

N/A

Ram data and addressmg test has falled and the microcodeis sp1nn1ng in a loop.

117

N/A

120

N/A

Indexed indirect addressing mode instruction test has failed and the microcode
is spinning in a loop.

121

N/A

Line unit message test has failed and the microcode is spinning in a loop.

305

The microdiagnostics have completed without errors.

Ram scratchpad CSR, and NPR address resisters addressmg test has failed

and the mlcrocode1S sp1nn1ng in a loop

Ram alternating data test has fa11ed and the microcodeis spinning in a loop.

7-24

DEC/X11 DMV11 Modules

7.6.6

There are two DEC/X11 modules for the DMV11; DMD* and DME*. Together these two modules
can operate:

Upto 16 DMVI1 devices in point-to-point links.

A single DMVI11 configured as a multlpomt control station communicating with up to 12

tributaries.

Up to 16 devices configur\ed as multipoint tributaries on the same LSI-11 bus.

These modules transmit, receive, and check 32 data messages of 1024 bytes each on a given physical

link. By default, this involves a single LSI-11 system with one or more devices operating in internal or
external loopback mode. However, by operator selection of nondefault modes, actual point-to-point or
multipoint operatlon is possible.

7.6.6.1 DMD* DMD*is the master module. It can operate up to 16 DMV11 devicesin looped-back
and point-to-point modes, or a single devicein 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 required for each group of looped-back DMV11 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 reglsters for that
module. The DMD#* module uses switch registers SR1-SR4 as follows:
SR4 has three allowable values: 0, 1, or 2.

- SR4=0:IF TESTING DMPI11

SR4=1 :IF TESTING DMV11

SR4=2 :IF TESTING DMV11 (AND Q22 SOFTWARE MODE IS DESIRED,).

SR1 has three allowable values; O, 1, or 2.
When SR1=0:

All selected DMV11s run in point-to-point full-duplex mode with internal or external loop-

SR2 has the following meaning:

back on all devices.

If SR2=0,'internalrloopback is provided by the program. This is accomplished using TTL-

level loopback on the line unit. SR2=0 is the default mode of operation.

If SR2=1, external loopback is provided by H3254 or H3255 test connnectors on each de|

vice.

If SR2=2, cable loopback is provided by H3250 or H3251 test connectors.
When SR1=1:

All selected DMVI11s run in point-to-point full- or half-duplex mode without loopback.

The DMD* module communicates with DME* (slave) modules on the same or other LSI-11
systems.

7-25

SRZ and SR3 software swrtch regrsters are not used

When SR1=2:
e

Only one DMVI1I is selected.

The selected DMV11 runs in multipoint control full- or ha‘l.f-dupleX mode without loopback.

The DMD* module communicates with DME* (slave) modules on the same or other LSI 11
systems

e SR2
=- The total number of trrbutarres on thrs multrpomt link. The allowable range is from 1

{

(377 to 1) if necessary

SR3= ‘The starting tributary address. The program uses this startmg address to compute the
other addresses. The allowable address range is from 1 to 3778 and they may wraparound

e
-

to 14g.

.6.6. 2 DME* — DME*is the slave module It can operate up to 16 DMVll devices i1n pornt-to- pornt
slave or mult1p01nt tributary modes.

A separate DME* moduleis requrred for each group of point-to-point slaves
or multipoint tributaries

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 swrtch regrsters SR1 SR4 as follows:

SR4 has three allowable values: 0, 1, or2.
SR4=0:IF TESTING DMPI11
SR4=1:IF TESTING DMV11
SR4=2:IF TESTING DMV11 (AND Q22 SOFTWARE MODE IS DESIRED)

SR1 has two allowable values 0 and 1

When SRl =0:
~ o

All selected DMVI1 devices run inpoint-to-point slave, full- or half-duplex mode without

loopback.

e
-

The DME* module commumcates with the DMD* (master) modules on the same or other
LSI-11 systems.

SR2 and SR3 are unused.

When SR1=1:

All selected DMVII devices run.in multipoint tributary full- or half-duplex mode without

loopback.

The DME* module commumcates w1th a DMD* (master) module on the same or other LSI11 systems.

7-26

on a system.

SR2 = The total number of tributaries on the multipoint link on this CPU. The allowable
range is from 1 to 14sg.
|
SR3 = The starting tributary address. The program uses this starting address to compute the

other addresses. The allowable address range is from 1 to 377g and they may wraparound
(377 to 1) if necessary.
NOTE
If the DMV11 DEC/X11 modules are configured to
run in linkmode, it is recommended that the exerciser be started in run lock mode. If this is not done, the
exerciser may hang.

7.6.7 Soft Error Reports Under DEC/X11
Soft errors indicate errors which occurred 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 errors. The soft error report may be used in the isolation of certain DMV11 failures from UNIBUS
loading or data late problems.

"The DMV11 has no data late bit or capabilities for detecting the fact that it did not obtain bus mastership in time to service the synchronous line. The DMV11 interprets 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 occurrence causes incrementation of the cumulative error

counters in DMV11 RAM memory.

A process of elimination must be used to determine whether soft errors (BCC) are caused by bus latency or failing DMV11 hardware.
Typically, the DMV
11 should show no errors when running in a local loopback mode. This is normally a
noise-free circuit. Therefore, any soft error reports should be examined and the cause isolated.

If soft errors are reported while running a DMV11 on a fully loaded system (other devices being exercised simultaneously), they may be due to bus latency. This may be verified by running only the
DMD*DEC/X11 module with only one DMV11 enabled. If the soft errors cease, a latency condition is
indicated.

If soft errors persist while running only the DMD*DEC/X11 module, the DMV11 device diagnostics

should be run. The problem could be a faulty DMV11 or cable.

SR1 and SR2 (bit 0) may be used in the isolation process. If SR1=0 and SR2=1, DEC/X11 does not
set line unit loopback but it uses an external turnaround. By running with SR1=0 and SR2=0,a TTL
loopback is performed, eliminating the possibility of the cable/turnaround connector being faulty.
TTL loopback eliminates the level converters and the integral modem. The bit rate selected is 56K b/s
using the internal clock.

7-27

APPENDIX A

DDCMP IN A NUTSHELL

DDCMP

A.1

The Digital Data Communications Message Protocol (DDCMP) provides a data hnk 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.

Controlling Data Transfers

A.1.1

The DDCMP message format is shown in Figure A-1. Three control characters are provided in
DDCMP to differentiate between the three possible types of messages
‘e

SOH - Data message follows,

ENQ - Control message follows,

DLE — Maintenance message follows.

“

Note that the use of a fixed-length header and message size declaratlon obviates the requirement for
extensive message and header delimiter codes.

SOH

DATA

(ANY NUMBER OF 8-BIT

CRC-2

CHARACTERS UP TO 2'4) 16 BITS
MK-2248

Figure A-1
A.1.2

DDCMP Data Message Format

Error Checking and Recovery

DDCMP uses a 16-bit cycle redundancy 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 example, 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 NAKis simply added to the sequence of messages
for the transmitter.

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

A-1

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 does 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’'m still looking for message 5.”
A.1.3

Character Coding

DDCMP uses ASCII control characters for SYN, SOH, ENQ and DLE. The remainder of the mes‘sage, including the header, is transparent.
A.1.4

Data Transparency
DDCMP 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 includedin 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 freld Thus, character stuffmgis

avoided.
A.1.5

Data Channel Utilization

)

DDCMP uses either full-duplex or half-duplex circuits at opt1mum efflcrency In the full- duplex mode

DDCMP 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, therefore, reducrng the control overhead Acknowledgements
are simply placedin the response field of the next data message for the opposite direction. If several

data messages are received correctly before the terminalis able to send a message, all of them can be
acknowledged by one response. Only when a transmission error occurs, or when trafficin the opposite
direction is light (no data message to send), is it necessary to send a spec1al NAK or ACK message,
respectively.
In summary, DDCMP data channel utilization features 1nc1ude

The ability to run on full-duplex or half-duplex data channelfacrhtres

2. Low .control character overhead,

3 No character stuffivng,»‘ | N
4.

No separate ACKs when traffic is heavy; this saves on extra sync-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 transmissions in a serial synchronous mode; that operatmg environment is
emphasizedin the followmg description.
|

The headeris the most. 1mportant part of the message because it contarns the message sequence num-

bering.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

CRC-1. 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 (indicatedin
Figure A-2 as CRC-2).
|
v
-

NOTE 1
A

INFORMATION

SYN

CLASS COUNT | FLAG
|

14 BITS|

2 BITS|

8 BITS

|16 BITS

OF 8-BIT

XOOOOONK

XOIOXIXXIXXINK

OO XXX

SOH
- DATA MESSAGES
ENQ {ACKNOWLEDGEMENT
~
NEGATIVE ACKNOWLEDGE

10000001
00000101
00000101

CHARACTER COUNT
00000001000000
00000010---------

QS
QS
QS

RESP
#
RESP
#
RESP
#

MESSAGE#
00000000
00000000

REASONS:

BCC HEADER ERROR

000001

BCC DATA ERROR

000010

REP RESPONSE

000011

BUFFER UNAVAILABLE

001000

RECEIVER OVERRUN

MESSAGE TOO LONG
HEADER FORMAT ERROR

|
ENQ
DLE

REPLY MESSAGE
START MESSAGE

OO
ADDRESS
ADDRESS
- ADDRESS

001001

010000
010001

0000001 1000000
00000110000000

QS
11

00000000
00000000

LSTMESS#
00000000

START ACKNOWLEDGEMENT 00000101

00000111000000

00000000

00000000@

ADDRESS

MAINTENANCE MESSAGE

CHARACTER COUNT

00000000

ADDRESS

00000101
00000101

16 BITS

CHARACTERS

10010000

ADDRESS
ADDRESS

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

\W/

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 "1 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

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 palr of stations that exchange messages with each other number those messages se-

quentially starting with message number one. Each successive data message is numbered using the next
numberin sequence, modulo 256. Thus, a long sequence of messages would be numbered 1, 2, 3,...254,
255, 0, 1,...255. The first message sequence always starts with a number 1, and every sequence thereafter begins with a 0. 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 sendmg its messages 5, 6, and 7 to
station A. Thus, in a multipoint configuration where a control station is engagedin two-way commu-

nication with ten tributary stations, there are 20 different message number sequences involved — one

Whenever a station transmits a message to another station, it assigns its next sequential message number to that message and places that numberin 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 includedin 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 transmittmg station. DDCMP does not require an acknowledgement for

each message, as the numberin the response field of a normal header (orin 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 REP (reply) 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 specifiedin 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 DMV11 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,
representedin 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, respectively. The next
two characters of the message are broken into a 14-bit field and a 2-bit field. The 14-bit fieldis usedin
data and maintenance messages to indicate the number of characters that follow the header CRC field
and form the information part of the message. In control messages, the first eight bits of the 14-bit field
~are used to desrgnate what type of control message it is; the last six 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 2-bit field contains the quick-sync and select flags.

A-4

sequence for messages from each of the ten tributaries to the control station, and one sequence for
messages from the control station to each of the ten tributaries.

The quick-sync flag is used to inform the receiving station that the message will be followed by synccharacters; the receiver may wish to set its associated synchronous receiver hardware into sync-search

mode and sync-strip mode. This reestablishes synchronization and syncs are 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 is true of half-duplex point-to-point links, as well as full- and half-duplex multipoint links. On halfduplex 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 ceases 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 messages. 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 messages, 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.

" APPENDIX B

FLOATING DEVICE
AND VECTOR ADDRESSES

B.1

FLOATING DEVICE ADDRESSES

UNIBUS and LSI-11 addresses, starting at 760010 and contmumg through 763776, are designated as
floating device addresses (see Figure B-1). These are used as register addresses for communications
-

(and other) devices interfacing with the PDP-11 (refer to Table B- 1)

|
NOTE
Some devices are not supported by LSI-11; however,
the same scheme applies. That is, gaps are provided

as appropriate. The convention for assigning these
addresses is as follows:

A gap of 10g 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 10g boundary. The gap of
10g must also be left for devices that are not installed but are skipped over in the priority ranking list.
Multlple devices of the same type must be assigned contiguous addresses. Reass1gnment of device types
alreadyin the system may be requlred to make room for addltlonal 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-11 and VAX-11.
Multiple devices of the same type would be assigned vectors sequentially (refer to Table B-2).
NOTE

Some devices are not supported by LSI-11; however,
the same scheme applies. Vector size is determined
by the device type.

There are no gaps in floatmg 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
43 boundary).

B-1

777 777

2K
WORDS

DIGITAL EQUIPMENT

CORPORATION
(FIXED ADDRESSES)

770 000

767 777

DR11-C

v
o

WORDS

USER ADDRESSES

764 000
763 777

WORDS

i
FLOATING ADDRESSES

760 010
DIGITAL EQUIP CORP (DIAGNOSTICS)

760 006
'

760 000

757 777

001 000

000 777

VECTORS

'FLOATING VECTORS

000 300
000 277

TRAP & INTERRUPT

VECTORS

000 000
MK-2190

Figure B-1

UNIBUS and LSI-11 Address Map

Table B-1

Floating CSR Address Devices

Rank

‘Decimal

- Option

Octal

Size

Modulus

DJ11

DHI11

3
4

DQ11
DU11

5
6
7
8
9
10
11

VMV31

207

4
4

DUPI11
LKI11A
DMCI11/DMRI11

10
10

4
4
4

10
10
10

DZ11* and DZV11/DZ32
KMCIl11
,
LPP11
|

4
4
4

VMV21

DWR70

RL11 and RLVI11

LPA11-K
KWI11-C

8
4

Reserved

18
19

RX11
DR11-W

4
4

DR11-B

21
22
23
24

DMPI11-AD
DPV11
ISB11
DMVI11-AD

10 (after second) -

4
4
4

10
10
10

10
10

201

10 (extra only)
20 (extra only)
10
10

- 10 (extra only)
10

*DZ11E and DZ11F are dual DZ11s and are treated by the algorithm
as two DZ11s.
TStarting CSR address must be an even multiple of 20 (octal).

Rank

Table B-2

Floating Interrupt Vector Devices
Decimal
Size

- Option

QOctal
- Modulus

DCIl11

KL11 (extra)

10*

DLI11-A (extra)
. DL11-B (extra)
DP11
DMI11-A

4
4
4
4

10*
10
10
10

2
2
3
4

5
6
7

DNI11
"DM11-BB
DH11 modem control

DRI11-C

2
2
2
4

4
4
4
10*

*The vector for the device of this type must always be on a 10g boundary.

Table B-2 Floating Interrupt Vector Devices (Cont)

Rank

|
Option

Decimal
Size

Octal
Modulus

10
11
12
13
14
14
14
15
16

PA611 (reader & punch)
LPDI11
DTI11
DX11
DL11-C
DL11-D
DLI11-E
DJ11
DHI11
|

8
4
4
4

4
4
4
4
4

10*
10
10*
10*
10*
10*
10*
10*
107

GT40/VSV11

18
19

LPS11
DQ11

12
4

10*
10t

KWI11-W

21
22

DU11
DUPI11

4
4

10*
10*

DV and modem control

24
25

LKI11-A
DWUN

4
4

10
10

27
28
29
30

DZ11/DZ32/DZV11
KMCl11
LPP11
VMV21

4
4
4
4

32
33
34
35

VTVO0l1
DWR70
RL11/RLVI1I
TS11

4
4
2
2

36
37
38
39
40
41
42
43
44
45
46

LPA11-K
IP11/1P300
KWI11-C
RX11/RX211
DRI11-W
DR11-B
DMPI11-AD
DPV11
MLI11
ISB11
DMV11-AD

4
2
4
2
2
2
4
4
2
4
4

DMCI11/DMRI11

VYMV31

- 10%*

10*
10
10
10

10
10*
4
|
4 (after the first)
10
4
10
- 4 (after the first)
4
4 (after the frrst)
10
10
4 (MASSBUS dev1ce)
10
10

*The vector for the device of this type must always be on a 10g boundary
TThese devices can have either a M7820 or M7821 interrupt control module However it should always
be on a 108 boundary

B-4

B.3 EXAMPLES OF DEVICE AND VECTOR ADDRESS ASSIGNMENT

This example
has three devices that require device and vector address assignment in the floating ad-

dress space. The devices are:

Device
(Option)

Device
Address

Vector
Address

760020
760030
760040
760050
760060
760070
760100

760110
760120
760130
760140
760150

760160
- 760170

300

760200

760210
760220
760230
760240

DPVI1I
DPVI1I

DMVI11

Comment

Gap left for DJ11
Gap left for DH11
Gap left for DQ11
Gap left for DUI11
Gap left for DUPI1

760010

RLVII

760250
760260
760270
760300
760310

310

760320
760340
760360

330

320

Gap left for LK11A
Gap left for DMCI11/DMRI11
Gap left for DZ11/DZV11
Gap left for KMC11
Gap left for LPP11
Gap left for YMV21
Gap left for VMV31
Gap left for DWR70
First and only RLV11
Gap left between RLV11 and next
device
Gap left for LPA11-K
Gap left for KW11-C
Reserved
Gap left for RX11
Gap left for DR11-W
Gap left for DR11-B
Gap left for DMPI11
First DPV11
Second DPV11
Gap left between DPV 11 and next
device
- Gap left for ISB11

First and only DMV11
Gap left after last device, in this
case the DMV11, to indicate that no
other devices follow

APPENDIX C
MODEM CONTROL REGISTER FORMATS

C.1 MODEM CONTROL REGISTER FORMATS
The modem signals made available by the DMV11 can be examined or modified by the user program if
- needed. This supplies the flexibility needed to meet the varlous modem interface requirements of different countries.

READ MODEM STATUS
BSEL4:

Bit

Name

CARRIER

Description

Received line signal detector, commonly referred to as carrier detect, indicates that there is an appropriate audio tone being received from the remote modem. Typically,in full- and half-duplex
applications,
carrier
detect
is
on
whenever
the
communications line is intact and the remote modem has the signal request to send asserted (the modem is transmitting). ThlS
s1gna1is also applicable to the DMV11 1ntegra1 modem.

NOT USED

ALWAYS READ AS ZERO.

- CLEAR TO SEND

This 51gna1is generated by the local modem to indicate whether
or not it is ready to transmit data. Clear to send is the local
modem’s response to the asserting of request to send. This signal
has a slightly different meaning with different modems. With
some modems it indicates that the carrier is being received from
the remote modem, and, therefore, is an indication that a suitable
communications channel exists.

MODEM READY

This signal indicates 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 exchange further control signals
with the DMV11. The OFF condition indicates that the local
modem is not ready to operate. This signal, when implemented by
the modem, is used by the DMV11 to detect either a power-off
|
condition or a cable-related modem malfunction.

HALF-DUPLEX

This signal, when asserted, indicates that the DMV11 is in the
half-duplex mode. This means that the DMV11 is connected to a
communications line designed for transmission in either direction, but not in both directions simultaneously. When cleared, it
implies full-duplex operation which is two-way independent transmission in both directions.

C-1

Description

REQUEST TO SEND

This signal serves to control 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 non-

transmit mode. On a half-duplex channel, the ON condition

maintains the modem in the transmit mode and inhibits 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 ac- tion as may be necessary and indicates completion of such actions
by asserting clear to send, thereby, indicating to the DMV11 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 nontransmit or receive mode, whichever is appropriate. The modem responds to this instruction by
turning OFF the signal clear to send when it is again prepared to
respond to a subsequent ON condition of request to send.

This signal controls the switching of the local modem to and from

DATA TERMINAL

READY

the communications line. When asserted, this signal serves to in-

B ~ form the local modem that the DMV11 is ready to operate. This

signal also prepares the modem for connection to the commu-

‘nications line and maintains this connection as long as it is ON.
When turned OFF, this signal causes the local modem to dis- connect 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 DMVII or a cable-related
modem malfunction.

RING|

ThlS s1gna1 1ndlcates whether an mcommg call 51gna1is being re-

- ceived by the local modem. When ON, this signal indicates that
-

an incoming »call v(yrmgmg) 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 rmgs)
on the communications line. The OFF condition must be maintained during the OFF segment of the rmgmg cycle (between
rings) and at all other times that ringing is not being received.
This signalis not affected by the state of data terminal ready.

READ MODEM STATUS
BSEL5
Bit

Name

Description

MODE

This bit indicates the operational mode of the line unit. A one in‘dicates character-oriented protocol operation, and a zero in-

dicates bit-oriented protocol operation. The DMV11 initializes
this bit to one.

C-2

\»//

”

Name

Bit

}

Bit

Name

Description

NOT USED

ALWAYS READ AS ZERO.

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
DMV11 that the local modem has been placed in a test condition.
The ON condition can also be in response to either local or reby means of any other modem test condition. Acmote activation
tivation 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 modemis not in the test mode andis available for

normal operatlon

WRITE

NOT USED

.ALWAYS READ AS ZERO

NOT USED

ALWAYS READ AS ZERO

' NOT USED

ALWAYS READ AS ZERO

NOT USED

ALWAYS READ AS ZERO

NOT USED

ALWAYS READ AS ZERO

MODEM CONTROL

Bit

Name

NOT USED

[GR—

\\""—///'

BSEL4

Description

SELECT STANDBY

Defaulted to O by DMVI 1 hardware.

MAINTENANCE

| Defaulted to 0 by DMV11 hardware.

MAINTENANCE

Defaulted to 0 by DMV11 hardware.

MODE 2

MODE 1

HALF-DUPLEX

The DMV11 uses this bit to place the line unit into the half-duplex mode. The user program cannot set or clear this bit. The
DMVI11 can change line characteristics only through the mode
definition command. The DMV11 is equipped with a software interlock that prevents simultaneous transmission and reception
when in the half—duplex mode. While the transmitter is transmitting, the receiver is disabled from rece1vmg data via a hard
ware interlock.

SELECT
FREQUENCY

This signal is used to select the transmit and receive frequency

bands of a modem. In the ON condition, the higher frequency
C-3

Bit

Description

Name

“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 como 'munications chan"nel.
|

NOTE
|
The modem, if it supports select frequency, must be
set up to 1gnorethis s1gnal from DMVH

DATA TERMINAL

| Thls srgnal controls sw1tch1ng of the local modem to and from the

READY

communications line. When asserted, this signal serves to inform

the local modem that the DMV11 is ready to operate. This signal

also prepares
the modem for connection
to the communications

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 DMVII1 or a cable related modem
malfunction.

NEW SIGNAL

This signal determines whether or not the local modem will rapid-

ly respond to new data on the communications line. This signal 1s
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 permit rapid accommodation to signals from several tributaries appearing in quick succession, the
control station informs the local modem when a new signal is
about to begin by asserting polling for a brief 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 mes-

sages 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
more quickly to the line signal present after polling has been
turned OFF. Thrs s1gnal applres only to modems that support pol
lmg

C.2 RS- 449 VERSUS RS232 C

The most common interface standard in use durmg recent years is RS-232-C. However, when usedin
modern commumcatlons systems it has crltrcal llmltatrons the most serious being speed and distance.

For thrs reason, the mterface standard RS449 was developed to replace RS-232-C. This standard main-

tains a degree of compatibility with RS-232-C to accommodate an upward transition to RS-449.

The most significant difference between RS-449 and RS-232-C is the electrical characteristics of signals used between the data communications equlpment (DCE) and the data terminal equipment

C-4

(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
contained in EIA standards RS-422-A for balanced circuits and RS-423-A for unbalanced circuits.
These new standards permit greater transmission speeds and allow greater 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 normal transmission speed limits are 20K b/s for RS-423-A at 61 m (200
feet) and 2M b/s for RS-422-A also at 61 m (200 feet). These normal transmission speeds can be varied
by tradeoffs between speed and distance.
Another major difference between RS-449 and RS-232-C is the specification of two new connectors to
accommodate the leads required to support additional circuit functions and the balanced interface circuits. One connector is a 37-pin cinch used to accommodate the majority of data communications applications. The other is a 9-pin cinch for applications requiring 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 require connection between RS-232-C and RS-449 interfaces must adhere to the limitations of RS-232-C,
which specifies a normal transmission speed of 20K b/s at a maximum distance of 15.2 m (50 feet).
DMVII does not support RS-422-A balanced circuits.

APPENDIX D

MODEM CONTROL

D.1
MODEM CONTROL
There are two levels of modem control available to the DMV11. The first levelis provided by the hardware, and the second by the DMV11 mlcrocode
D.1.1
Hardware Modem Control
The DMV11 provides the following modem control function:
e

Prevention of simultaneous transmlssmn and receptlon in half-duplex mode.

Half-Duplex Mode- When set, HALF-DUPLEX spe01fles that the DMV11is in the half—duplex mode.
In half-duplex mode, a hardware 1nterlock prevents the DMV11 from transmlttmg and receiving simul-

taneously.

NOTE
This hardware lockout prevents the DMV11 from
being usedin the half-duplex mode on a full-duplex

modem with the continuous carrier option installed.

D.1.2 Modem Control Implemented by the DMV11 Microcode
The modem control signals implemented by the DMV11 are:

Modem ready (data set ready),
Request to send/clear to send,
Carrier,
Data terminal ready, and
Auto answer.

Each of these signals are outlined in Table D-1.
Once modem ready goes ON, the DMV11 reports any transition from ON to OFF to the user program
by issuing a control response containing the code for the system-event modem disconnect. The microcode tests that modem ready is OFF for 10 ms. Transmission is initially inhibited by the microcode by
interlocking the signals modem ready and request to send.
Whenever the signal carrier detect is dropped by the modem for greater than 1.28 seconds, the user
program is notified by a control response containing the code for the system event modem carrier loss.
Diagrams are used in the discussion of modem control functions. Refer to Figure D-1 as an aidin inter-

preting these diagrams. The flow depicted by the diagrams (Figures D-2 through D-8) describes the processing of EIA modem control signals by the DMV11. Each diagram represents a serial flow for a specific modem control function. However, the functions performed, as represented by each diagram, are
performedin parallel. The readable and writeable modem 51gna1s listed on the diagram for modem status can be read and written through the control command usmg the request keys read modem status
and write mode control.
»

Table D-1

DMV11 Modem Control Functions

Signal

Description

Data Set Ready-

Software interlock prevents the DMV11 from transmitting if DSR is not

Modem Ready:

returned. If DSR drops (meaning that it once was asserted) for a period

of 10 ms, the transmitter and receiver are resynchronized, the transmitter
and receiver sections of the microcode are reset to the idle state to allow
the user to return buffers, DTR is then dropped to clear the line (see
of the DSR
DTR for reasserting conditions), and the user is then notified
|

for disconnect.
drop via a control-out

Software interlock preventing the DMV11 from transmitting if DSR is
by the user to start up
not returned: If the DMV11 has been instructed
the communications line, and DSR is not asserted, the DMV 11 does not
transmit. There is no timer started for the first assertion of DSR. It is the
responsibility of the user to ensure that the modem is plugged in. Also, if

the modem is a dial-up modem, the user should make sure that the num-

ber is dialed. In most cases the start command is issued with the intent of

- waiting for an incoming call. In this case, the timer value is arbitrary so
that it has been left up to the user software to determine this timeout. If
data terminal ready (DTR) is not asserted because of a past error condition that caused the dropping of DTR (that is, disconnect), the user program may assert DTR via the write modem command to enable transmis-

osion.

Request to Send/
Clear to Send:

For all applications: Before RTS is 'dsserted (i‘f élready asserted thisis

bypassed) CTS is checked for the “ON” condition. If CTS 1s “ON”, a
10-20 ms timer is started while waiting for CTS to drop. If CTS does not
drop within the timer period, constant CTS is assumed and RTS is set.

For all applications: Software interlock prevents transmission if CTS i1s
not returned. IF CTS is not returned within 30 seconds (plus or minus 10
ms), a disconnect control-out is queued with a CTS failure code in
BSEL7. The transmitter and receiver are resynchronized, the transmitter
"and receiver sections of the microcode are reset to the idle state to allow
- the user to return buffers, and DTR is then dropped to clear the line. (See
DTR for reasserting conditions).

For all applications: During the time that RTS' is set, evéry 10 ms CTS is

for
checked for the “ON”’ condition. If CTS stays in the “OFF” condition

30 seconds (plus or minus 10 ms), a disconnect control out is queued with

a CTS failure code in BSEL7. The transmitter and receiver are res~ ynchronized, the transmitter and receiver sections of the microcode are
reseto the idle state to allow the user to return buffers, and DTR is then
dropped to clear the line. (See DTR for reasserting conditions).

- For all half-duplex applicatiokns}: T'he sett‘in‘g of requést_to send is “AN-

- DED?” with the half-duplex bit in the hardware to “blind” the receiver
when transmitting.

Table D-1 DMV11 Modem Control Functions (Cont)
Signal

Carrier:

-~ Description

Software interlocks prevent transmission in half-duplex if carrier is in the
“ON” condition. This prevents the DMV11 from running half-duplex on
four-wire constant carrier modems.
For all applications: Hardware interlock of carrier and the receiver clock

- stop the USYRT from receiving if carrier were to drop in the middle of a
message.

For all applications: If carrier “drops” while the DMV11 is in the process
of receiving the carrier, the loss timer is started. If the carrier loss timer
expires (1.28 second interval), the user is notified via a control-out for
carrier loss. The receiver is then resynchronized and the receiver microcode is reset to the waiting state for the next message. If the carrier loss is

less than 1.28 seconds (carrier is reasserted before the timeout), the message being received is allowed to finish. If CRC errors are detected (normal case), the protocol recovers from the failure.
Data terminal ready:

)

DMV11 clears DTR on a power-up bus initialization, and a master clear.
This
is a hardware function. DTR is not gated from the interface drivers
when the DMV11 is placed in loopback mode. DTR is monitored by diagnostics running in internal loopback to ensure that the microcode does not
set it.

When DTR is dropped because of errors, it is only reasserted if any of the
following conditions exist: auto answer is enabled or remote load detect is
enabled. The code is in the process of power-on boot or request boot.
Auto Answer:
\\)

This option is switch selectable. If enabled, the DMV11 asserts DTR and
waits for modem ready (DSR). Because of the difference between
modems in the U.S. and other countries, ring is not used as an indication
that an incoming call has been established. As it stands, DSR is the indication that the call has been established. If a valid DDCMP message is
not received within 30 seconds (plus or minus 10 ms) after a connection is
established, DTR is dropped (hang up the phone). The connection is considered to be established on assertion of carrier or clear to send. The
transmitter and receiver are resynchronized, the transmitter and receiver
sections of the microcode are reset to the idle state, and DTR is then reasserted after DSR drops (or in 10 seconds whichever comes first). In this
case the user is not notified of the cancelled call. An internal counter is
incremented tec log the incoming calls (latches at 256) and is available for
reading by the user program.

D-3

CONDITIONS TO BE SATISFIED. IF CONDITIONS ARE NOT

MET SERIAL FLOW DOES NOT CONTINUE.

~ ACTIONS PERFORMED BY DMV11.

(

)

ENTRY
OR EXIT SYMBOLS

Figure D-1

Flow Diagram Symbology

D-4

MK-2657

POWER ON

SYSTEM

[2.4]

INITIALIZATION

DEVICE

(%4

2.4

MASTER CLEAR [2-4]

POWER ON

L CLEAR DTR

BOOT ENABLED

(SWITCH)

][2.5]

{

USER ISSUES

MODE DEFINED

MODE DEFN

BOOT

IN SWITCHES

s sgT

COMMAND

[2.2]

SET DTR

SET CALTMR

= ON

| seT p/moP 23]
A

USER REQUEST

CALTMR

MODEM STATUS

= ON

READ/WRITE

U
2.2]

}

DSR = ON

[

CD=ON

L DSR = ON

SCAN EOR

‘

( MODSTAT

)

2.1]

( CALTMR

REMOTE

)

LOAD

MESSAGE
|

VALID
MESSAGE

I 7~
CB[2.3]
H SET P/MOP
MESSAGE TO

DSR = ON TO

BE SENT

OFF

(

FOR

10mSEC

RECEIVE )
‘

|
NOTIFY USER OF

P/MOP = ON

P/MQP = OFF

DSR DROP

DISCONNECT

—-—>®

(CONTROL RESPONSE)

)

(

TRANSMITZ)

{ TRANSM@

NOTES:
[2.1] REMOTE LOAD DETECT IS A MAINT. DDCMP MESSAGE INITIATING A DOWN LINE LOAD.
[2.2] CALTMR

- RUNNING.

- (CALL TIMER): USED TO DETERMINE IF VALID MESSAGE IS RECEIVED. “ON" INDICATES TIMER

[2.3] P/MOP- PRIMARY MAINTENANCE OPERATION PROTOCOL, REQUESTING REMOTE LOAD
[2.4] RUN COMPLETE MICRODIAGNOSTICS (MICROPROCESSOR AND LINE UNIT)
[2.5] HARDWARE FUNCTION

[2.6] INVOKE PRIMARY MOP BOOT (BIT FIVE, BSEL 1)
MK-2705

“Figure D-2

Modem Control (Start)

D-5

RECEIVER NOT
ACTIVE

.
|

CD = OFF

NOTIFY USER OF
STREAMING

ENCOUNTERED

,
|

ASSERTRTS

CTS=0
|

{

CTS TIMR=0

CTS=ON

CD = OFF
\

NOTIFY USER

OF CTS FAILURE

RTS = ON

RTS=OFF

.'w o

'(SHUT DOWN)

CTS TIMER 20ms
:
CTS=OFF

"
CTS TMR=0

TRANSMIT

END OF TRANS

CTS=OFF

\
A

FDX

HDX

)

CTS TIMR <—

10 SEC

CLEAR RTS

TIMR=0

CTS=ON

CANCEL CTS TIMR

NOTIFY USER OF

CTS FAILURE

CSHUT Dow@

NOTES:

(3.1] CWTIMR
-- CARRIER WAIT TIMER, IT DETECTS THE CONDITION WHERE THE LINK WAS NOT RELINQUISHED IN
TIME BY THE 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.

Figure D-3

Modem Control (Transmit)
D-6

| CTS TIMER
1 SEC |

30 SEC

CLEARRTS
r

CTS TIMR <«—

‘L

STATION

LINE ERRORS

ENCOUNTERED

‘

CWTIMR

CARRIER WAIT |

[3.1]

NO LINE ERRORS.

TIMR =1 SEC

FULL-DUPLEX

CTSTMR =0
}

\\_,.J‘L/’

HALF-DUPLEX

‘

CTRAN]SMIT )

‘

"TRANSMIT 2 )

LESS THAN 24 REQUEST
REMOTE PROGRAM LOAD

24 REQUEST REMOTE PROGRAM
LOAD MESSAGES SENT
)

MESSAGES SENT

rT|MER<-3 sed

| CLEAR
DTR ’ ]
.

TIMER = 0

j .

‘

C TRANSMIT) »

| TIMER=10 SEC |
TIMER = 0

r SEJDTR

MK-2703

Figure D-4

Modem Control (Transmit 2)

( RECTIVE )
3

;

HALF-DUPLEX

- FULL-DUPLEX

Y
TRANSMITTER NOT
ACTIVE

IN HALF DUPLEX, LINE UNIT 1

ENABLE RCV CLOCK ONLY IF

CD = ON

SYNC'ED ON NEW
FRAME

;

START TO RECEIVE J—l

CD DROPPED
IN THE |

{

(RTS =OFF)
|

MIDDLE OF DATA

CD = OFF

STREAM

)
RESET CDTIMR
TO 1.28 SEC

CDTIMR =0

= ON
CD

/‘/

TRANSMITTER IS IDLE,

'PROVIDES INTERLOCK TO

NOTIFY USER OF
/

END OF
RECEIVE

CARRIER LOSS
(CONTROL RESPONSE)

| RESET RECEIVER]|

]

MK-2704

Figure D-5

Modem Control (Receive)

(

]

!
6.1]

MOD STAT )

v
WRITE MODEM

READ MODEM

STATUS

)

- STATUS

o A )

\_/

[6.1]

[6.2]

READABLE MODEM SIGNALS:

WRITEABLE MODEM SIGNALS:

CARRIER
CLEAR TO SEND

DATA TERMINAL READY

MODEM READY (DSR)
HALF DUPLEX
REQUEST TO SEND
DATA TERMINAL READY
RING

TEST MODE
MK-2702

Figure D-6

Modem Control (Modem Status)

—

( CALTMR ) o I START 65 SEC CALL TlME‘R]

CALL TIMER = 0

VALID MESSAGE

'RECEIVED FOR

;

" THIS STATION

r CLEAR DTR

SET CALTMR

= OFF

]

RESET RECEIVER
-~

AND

TRANSMITTER

| TIMER=-10 SEC |
TIMER =0

SETDTR

NOTE: CALTMR = ON
MK-1967

Modem Control (Call Timer)

Figure D-7

(SHUT DOWN)

CLEAR DTR
AND RTS

RESET RECEIVER

AND TRANSMITTER

[

BOOTING

REMOTE LOAD

CALL TIMR=ON

DETECT

TIMR=<—10 SEC

TIMR=0
[

SET DTR

MK-2866

Figure D-8

Modem Control (Shutdown)

D-11

DMV11 Synchronous Controller Technical Manual |

Reader’s Comments

EK-DMV11-TM-001

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?
well written, etc?

In your judgement is it complete, accurate, well organized,

Is it easy to use?

What features are most useful? ‘

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 Cata/og which contains information
on the remainder of DIGITAL's technical documentation.

Name

Company

. Street .

Department

SRS

State/Country
Zip

Additional Cdpi‘e‘s of this document are available from:

| :D.iAgil"'ftéI‘,--fE»'cw.,ifipment Corporation
- 444 Whitney Street

" Ndfthboto;"MA 01532
o -Attentlon Prmtmg and Circulation Serwces (NR2/M15)

' Customer Services Section

Order No. _EK-DMV11-TM-001

n o —— ———— ---————---—--—Do Not Tear — Fold Here and Staple - — —— ——

ol [

No Postage

Necessary
if Mailed in the

United States

BUSINESS REPLY MAIL
FIRST CLASS

PERMIT NO. 33

MERRIMACK, NH

POSTAGE WILL BE PAID BY ADDRESSEE

Digital Equipment Corporation

Educational Services Development & Publishing

Continental Blvd. (MK1/2M26)
Merrimack, N.H. 03054

Digital Equipment Corporation « Bedford, MA 01730