Digital PDFs

EK-DEBET-TM-003

May 2000

224 pages

Original

67MB

view download

OCR Version

31MB

view download

Document:	LAN Bridge 100 Technical Manual
Order Number:	EK-DEBET-TM
Revision:	003
Pages:	224
Original Filename:

OCR Text

Networks.Communications

—

LAN Bridge 100
Technical Manual
EK-DEBET-TM-003

dijgliltlall}

NOTICE - Class A Computing Device:
This equipment generates, uses, and may emit radio frequency energy. The
equipment has been type tested and found to comply with the limits for a Class A

computing device pursuant to Subpart J of Part 15 of FCC Rules, which are

designed to provide reasonable protection against such radio frequency interference when operated in a commerical environment. Operation of this equipment
in a residential area may cause interference; in which case, measures taken to

correct the interference are at the user’s expense.

LAN Bridge 100

Technical Manual
Order No. EK-DEBET-TM-003

This manual describes hardware logic, diagnostic firmware and software, and general
operating procedures. The manual is intended for use in training, in field service, and in

manufacturing. The level of technical information assumes previous training or experience

with Ethernet networks.

1 st Edition, August 1986

2nd Edition, July 1987
3rd Edition, March 1989

The information in this document is subject to change without notice and should

not be construed as a commitment by Digital Equipment Corporation. Digital

Equipment Corporation assumes no responsibility for any errors that
may
appear in this document,

SRR,

Printed in U.S.A.
.

This manual was produced originally by Networks and Communications
Publications and revised by Educational Services Development and
Publishing in Merrimack, N.H.

The following are trademarks of Digital Equipment Corporation:

mngnmm

DECUS

RSTS

DECwriter

RSX

DEC

DECmate
DECset
DECspell
DECsystem-10
DECSYSTEM-20

DIBOL

MASSBUS
PDP
P/OS
Professional
Rainbow

BRI

Scholar

ULTRIX
UNIBUS
VAX
VMS
VT
Work Processor

Amphenol 906 is a trademark of Amphenol.
AT&T is a trademark of American Telephone and Telegraph Co.
Corning is a trademark of Corning Glass Works.
FOTEC T302D is a trademark of FOTEC, Inc.
Motorola 68000 is a trademark of Motorola, Inc.
Xerox and XNS are registered trademarks of Xerox Corporation

Contents

—
Preface

—
-

Introduction
1.1

Introduction to the LAN Bridge 100 Unit ... 1-1

1.1.2

Transparent OPETALION .........ooooiuiieiiiiiiiiii et 1-3

1.2
1.2.1

LAN Bridge 100 Functional DeSCription ................iiiiineiniiciinnne. 1-4
Product Versions and Designations ................cooovieiiiiiiiiiin. 1-5

1.1.1

—

1.1.3

e
"

1.2.2
1.2.3

1.2.4

1.2.5

1.2.6
1.3
1.3.1

|
-

1.3.2
1.3.3

1.3.4
1.3.4.1

—
5

1.3.4.2

1.4

e
=
|

—

S—

N
—

UP SR 1-18
Loop Considerations .................cccoccoiieiininnnOO
Local LAN Bridge 100 Considerations ................o.coooiiimiiiiene 1-18

Remote LAN Bridge 100 Considerations ................c....cooeeeiiiininnn 1-20
Fiber-Optic Cable Between LAN Bridge 100 Units ................c....... 1-21

LAN Bridge 100 to Repeater Considerations ...................cccoooeeeeeeen 1-22

LAN Traffic Monitor Configurations .............cccccceeemmiiiiiieieeeenn ore. 1224

Dual Port Connections Between Two Ethernets ............................. 1-25

Bridge/Router Considerations ..............ccooivrouieneeiiirainannnn. e 1-27

1.5.2

Modifying Router Parameters for Extended LANS ............ccccccoeeeen 1-29

1.6.3
2

Remote Bridge Management Software (RBMS) ... 1-15
Configuration Considerations .................ccccooioiiiiiiins 1-16.
Performance Considerations .............cccoviimiiiiiiiiiiiiiiis 1-18

1.5

1.6.1
1.6.2

LAN Traffic Monitor Software ................coooeiiiiiiciennen. e 1-15

Dual Port Connections with Bridged Ethernets .............................. 1-26

1.6

LAN Traffic Monitor (LTM) OPtion ............ocooiiiiiiiiiini.. 1-14

1.4.3

153

LAN Bridge 100 Packaging ..............ccooooiiiiiii e 1-7
LAN Bridge 100 Controls, Status LEDs, and Connectors .............cee.-.. 1-8

Single Port Configuration with Loopback Connector Installed ....... 1-24

1.5.1

LAN Bridge 100 to Network CONNECHiONS ........ocoovviiiiiiiirniiiiinnnin. 1-4

1.4.1

1.4.2

Extended Networks ..........ocooiiiiiiiiiiiiiiiiin e 1-2

LAN Bridge 100 Unit in Parallel with a Router .................cccoooeeenn 1-28
e et et e e b e 1-30
e e et e e ett
e o
Setting Up Multiple ALCAS o

SPECIICALIONS ....oiiiiiiiiiieiiiiii e 1-31

Physical Specifications .................ccoociiin e 1-31
Environmental Specifications .....................ooone 1-32

Electrical SPeCifiCations ...............coooveiiiiiiiiiiiiiiiii 1-33

Installation and Verification
2.1

2.2

2.3
2.4

2:4.1

e . 2-1
eeeenerntas
Overview ................... et etteteeereeeaieeneneeeeteanenaeeeaaraeen

UNPACKING < oevevninieiiiiiiie e 2-2

Arranging for Software Installation ... e 2-4
e et e 2-6
INSEALIALIOM oo eeenee ittt et etettt

Setting the Voltage SwitCh ... 2-6
Contents-1

2.4.2

Veritying Configuration Switch Settings ................

................ 2-7
Rack-Mount Installation (Optional) ...........................
............ 2-9
Wall-Mount Installation (Optional) ........ TR
T 2-10
Connecting the Transceiver Cables ................................
.._ 2-12
Connecting the Fiber-Optic Cable ........................................
.. 2-13
Connecting the Power Cord ............ccccoooo
cooiii 2-15

2.4.3

2.4.4

2.4.5
2.4.6

2.4.7
2.5

Verifying the Installation ........................................
........._.___ 2-15
Verifying the Bridge Installation ...........................
............... 2-16
Verifying LAN Traffic Monitor (LTM) Installation ...........
............._. 2-18

2.5.1

2.5.2

Operation
3.1
3.2
3.2.1

3.2.2

3.2.3
3.2.3.1
3.2.3.2

3.2.4
3.2.5
3.2.6

3.3
3.3.1

3.3.2
3.3.2.1
3.3.2.2

3.4
3.4.1

3.4.2
3.4.3

3.4.4

3.4.5
3.4.6
3.4.6.1
3.4.6.2
3.4.6.3

3.4.7
3.4.7.1
3.4.7.2
3.4.7.3
3.5
3.5.1
3.5.2

353

3.5.4
3.6

3.6.1

3.6.2
3.7

Contents-2

Chapter Overview ...
3-1
Operational States .............................c
oo 3-1
SELF-TEST State ...........ccccoooooiiiiiiiiii
iiiiaiiiee 3-5
BROKEN State ...tt
t 3-5

INITIALIZATION Stat€ ..............oooooiieeiiiiiiieseee
3-5

Bridge Initialization ................................ e
3-5
Down-Line Loading ...........................e e
3-6

PREFORWARDING Stat€ ...................coooovviimmmeeeiiii 3-6

FORWARDING State ..............cocoooviiiiiiiiiiiiisieiioee oo 3-7

BACKUP State ...
3-7
Packet Memory Data Structures ....................................
.... 3-7
Initialization Blocks ........................................
. e 3-7
Descriptor Rings ...
3-9
Receive Descriptor Ring Entry ................................
.____ 3-11
Transmit Descriptor Ring Entry ....................................
____ 3-14

Learning and Forwarding ...
Table Initialization ......................................................
.
Address Table Entries ..................o.ooooooviimiimin
neei
Binary Search ...
Forwarding ...

3-18
3-20

3-21
3-21

3-23
Writing the Ethernet Address Table ..........................................
3-24
Special Cases ...
3-24
Stale Packets ......................oooeemiiii
3-24

Swapped Sides ................ccoooiiiiiii
Address AGINg ..................c.ccoiiiimii
i
Ethernet Address Table Maintenance ....................................__
Inserting New AddresSes ................coccoiiviiiiii
Single-Entry COMPAre .............c.ccccooeivvveiii
i
G4-Bit MOVE ......ocooooiiiiiiiiiiieii
e

3-24
3-25

3-25
3-25
3-26

3-27
LOOp Detection .............cooooiiiiiiiiiiiii i
3-27
The Spanning Tree Principle ............................
.oooii 3-28
The Spanning Tree Computation Process ........................
............._ 3-28
Spanning Tree Parameters ........................cc.ooooi
iiiii 3-30

ATy

Examples of the Spanning Tree Algorithm .................................... 3-31

Foreground and Background Operations ...................................
... 3-38
Foreground Operations ............................cccoooo
3-39
Background Operations ...
3-40
Remote Software and Bridge ACCeSS ..............cvii
iiiii 3-41

3.8
3.9
3.9.1
3.9.1.1
3.9.1.2
3.9.1.3
3.9.1.4

3.9.1.5
3.9.1.6
3.9.1.7
3.9.1.8
3.9.1.9
3.9.2
3.0.2.1
3.9.2.2
3.9.2.3
393
3.0.3.1
3.9.3.2
3.9.3.3

Maintenance Operation Protocol (MOP) ...............SURUUURUTROIRRPPRS 3-42
QEIE-TESE .« ovnenenieeen et e et ettt e 3-42
PPEP 3-43
BasiC TeStS .......ccceeviiiinininennnn. TSP
3—44
PPPPPPR
I
R
Program ROM Test ............ST O
3-44
i
Program RAM TESt .........cooiiiiiiimiiiiiiiiiiiii
i 3-44
e
NVRAM Checksum TeSt ..........cooimiiiiimiiiiiiiiiii
3-44
e
iiiiii
e
et
iiiaii
te
uniiin
..ooue
NVRAM WIte TEST

Ethernet Address ROM Checksum Test ................c...ooeeni. TR 3-44
Ethernet Address RAM TeSt ...........ccoooiiiiiiiiiiiiiiiiiiiiiiiies 3-44
t 3-44
et e
THMEE TEST ..orniiteiie ett
... 3-44
e
cooiiiiiiiiinnn.
Packet Memory Test ...............c
3-44
.
Packet Memory Refresh Test ..
3-45
LANCE Tests .......c.cccceennnn.e
iaiiiie 3-45
e
niiiie
et
LANCE RESEE TESt ....couiinei
t 3-45
e
iniiiiieiie
Internal LOOP TESES ....c..oouuiiiii
3-46
i
External LOOP TESLS .........ooouiiiiimiiiimieiiieiie
3-47
Table Lookup (TLU) Tests ...........ccooeeeeeinesPP
Status RAM TESE ...oiuniiniiiiiiiiiiiaiiiie ettt 3-47
Basic Binary Search Test ............c.coooiiiiiiiiiiiii s 3-47
Binary Search Engine Test ..o 3-47

Technical Description

4.1
4.2
4.2.1
4.2.2
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.4.1
4.3.4.2
4.3.4.3
4.3.4.4
4.3.45
4.3.4.6
4.3.5
4.3.5.1
4.3.5.2
4.3.5.3
4.3.5.4
4.3.55
4.3.5.6
4.3.5.7
4.35.8
4.3.5.9
4.3.5.10
4.35.11

Introduction ................. e eenn 4-1
ee 4-1
s s e eeneaaes
it tiae et e iteiie
et et eieieii
HArdWAre OVEIVIEW .....ovn
4-2
iiniiciian
....coovniiiii
Overview of LAN Bridge 100 Functxcmal BIOCKS
e,e e 4-3
The Data Path .........coooiiiiiiiiiiiie
t 4-5
e
iiiiiiniae
.....cooouuiuiiiiii
ProcCessOr SUDSYSTEIM
eiiiniieeaeaaans. 4-5
reereeeiiiiiiieeea
Processor Circuit DESCLIPLIONS ..........coeereiu
ae s s 4-7
e
e e e e e tetreriaia
The MICTOPIOCESSOL .....euuuiuieieieriiiiiiii
Interrupt Controller ... e 4-7
e e e e 4-9
e e e etaaiite
iieiii
e et
PrOCESSOT MEIMOTY ....unieinei
PROM oottt 4-9
Program RAM ...OO 4-9
Address Table Status RAM ..., 4-9
Physical Address PROM ...........cccooiiii, e 4-9
NVRAM ... TP PSP 4-9
Miscellaneous Control REGIStErs .............ocooviiiiiiiiiiiiiiii.. 4-10
Processor Memory Map ..........ccooieiminiinieiiiicnininnn. e 4-10
i 4-12
Program ROM ..........ccooooiiiiiiiiiiii
iiiiimiiiiiimi 4-12
.....oooo
Miscellaneous COoNtrol .........
i 4-13
ii
Nonvolatile MEMOLY ........coooiiuiiiiiiiiiaiiiie
i, 4-13
LANCE Descriptor Rin@s ..........ccocoooiimiiiiiiiii
LANCE CSR ReGiSters .........cccooeivuemniniiiinncnnnn.U OTPRTPRRS 4-13
Program RAM ...t 4-13
e 4-13
t
niiieii
e
niimiii
et
PaCKet MEMOIY .....ccouueimi
4-13
e
Ethernet Address Table ...,
4-13
Ethernet Address Table Status RAM ..., e
Compare-and-Move Registers ......................... e 4-13
Ethernet Address Table 64-Bit MOVe ..............ccooiiiiiiiiiiiiiiiiinn, 4-14
Contents-3

4.3.6

4.3.7

4.4

4.4.1
4.4.2
4.4.2.1

4.4.2.2

4.4.2.3

4.4.2.4
4.4.2.5

4.4.3
4.4.4
4.4.5

4.4.5.1
4.4.5.2

4.4.5.3
4.4.5.4
4.4.5.5

4.5
4.5.1

4.5.2
4.5.3
4.5.4
4.5.4.1
4.5.4.2

4.5.4.3
4.5.4.4

4.5.4.5
4.5.4.6

4.5.4.7

4.6

System Timers ....................cooeenen. etetierenetuareseebrarararaenrerraesnenses 4-14
ReSet CirCUItIy .......ociiiiiiiiii
el 4-14

Network Interconnect Subsystem ....................ccoooviiiiiiiiiiiniiniaiin... 4-15
NI Subsystem CirCUits .............coooviiiiiiiiiiiiii
e 4-15
NI Ports .............. e 4-16
STA CRP coooviiiiee oo 4-17
LANCE CRpP .ooooooiiiie oo, 4-17
LANCE CSR BUS REISTEIS .........ouuiiiiiiiiiiiiiiiieee e 4-19

LANCE CSR Control CirCUit ...............ccccooeiiieeiimiiiiaaeeiieieaeeeiiiii, 4-19
Fiber-Optic Module .................cccoiiiiiiiiiiiiiiiiiiie
e 4-19
LANCE Address LatCh ... 4-19
OWN Interrupt GeNErator ..............cveuiiniiiriiniiiiiiieei
e 4-20
Packet MEMOTY ..ot 4-20

Address Multiplexer ................cccoooooiiiiiiiieeian....SRR 4-20
Memory Control CifCUit ............ccoeiiimiiiiiiiiiiiiiiee e, 4-21
Dynamic Packet MEMOIY .............cooiiiiiiuneiiiiiieee e 4-22
Refresh COUNTET .........ccoeiiiiniiiiiii
e 4-23
DataBus LatChes ....................ocooiiiiiiiiiiiiiiieaeiann P 4-23
Table Lookup Subsystem ................e e 4-23
Overview of the TLU ..., 4-23

Circuit DesCriptiOns ............coiiiiiiiiiii
e, 4-26
Search Address MultipleXer ............coocooiiiiiiiiiii
el 4-26
48-Bit Address COMPATAtOr ............cc..eviiiniiiiiiiiie e 4-27
Functional DeSCriptiOn ................ccoouiiiiiiiiiiii e 4-27
Address Comparator Circuit DesCriptions ...............cc.cooevveeeeeneeni.... 4-30
Ethernet Address Table MEMOLY .............cccoeeiiiiiiuiuieeeeeiiieeeeeeeeeinnns 4-30
BUS TrANSCEIVEIS .....cccovuniiiiiiie it 4-30
Ethernet Search Address COmMParators ................c.ccceeeeeuruereeeeennnn..., 4-30

RGN

Search Control and Status PALS

....................c.cooiiiimiieiiiiiieaeiieeeaana, 4-31
Search Results REGISIEr .............cocoiiiiiiiiiiiiiiee e, 4-31

Power Supply ... 4-32

Maintenance
5.1

S0P e
e 5-1

5.2

Maintenance Philosophy ... 5-1

5.2.1

Required EQUipment ... 5-1

5.2.2

Optional EQuUipment ............ccoiiiiiiiiiiiiiiiii

5.3

Preventive MainteénancCe

5.4
5.4.1
5.4.2

Corrective MaintenancCe ...............cccooiiiiiiiiiiieiiiiiei
e 5-4
Troubleshooting TipPs ..........ccooiiiiiiiiiiiii
e 5-4
Fault Diagnosis ... 5-5

................cccoooiiiiiiiiiiiiiiiii

e 5-2
e 5-3

5.5

LAN Bridge 100 Replacement Procedures ..................cocoeevininennnn.... 5-14

5.6
5.6.1
5.6.2

Bridge Disassembly ............ccooiiiiiiiiii R 5-15
Plastic Enclosure Removal .................ooooiiiiiiiii i 5-16
Removing the Chassis Cover ...................... et
ineaaarea. D=18

5.6.3

Removing the Fiber-Optic Interface .......................cooooviiiiienienn... 5-21
Removing the Power SUPPLY ............ooeiiiiiiiiiiiiiiiiiiiie
e 5-23

5.6.4

5.6.5

—

Removing the Logic Module .....................ooiiiiiii
. e 5-29
S

Contents-4

Glossary

Fiber-Optic Link Analysis
B.1

B.2
B.2.1

B.2.2
B.3
B.3.1
B.3.1.1

B.3.2

B.3.2.1
B.3.2.2
B.3.2.3

B.3.2.4
B.3.2.5
B.3.3

B.4

B.4.1
B.5

B.6

B.6.1

B.6.1.1
B.6.1.2
B.6.1.3

B.6.1.4
B.6.1.5

B.6.1.6
B.6.2

B.6.3

B.6.3.1
B.6.3.2
B.7

B.7.1

B.7.2
B.8

B.8.1
B.8.1.1

B.8.2
B.9

B.9.1

B.9.2
B.9.3

B.10

Introduction ................ll fetsedeeiesesaesshinvecaracoiiossasnirrrersnesnanraracns B-1
System CONVENTIOMNS ........uiiiiniriiiiiiiiiiii it B-2
Absolute and Relative Measuréments .............oocoeiveiiiniinincnainnceen. B-2
Worst-Case DefinitiOn .........ocooviiiiiiiiiiiiiiiii B-3
Coupling Considerations ................ccoooiiiiiiiiiiii B-3

LED (Emitter)-to-Fiber ... B-4

LED-to-Fiber Coupling Derating Factors ..., B-5
i B-6
s
Fiber-to-Fiber Coupling ..............coooiiiiiiiiiiiiiii

Splice and Barrel Connector LOSSES ...........c.cccovviiiiiiiiiiiniiiiies e B=7
Guidelines for Connectors and Splices on Ethernet Products ............ B-8
Typical Cable Attenuation Values ... B-10
Single-Window and Dual-Window Fiber ....................nnn B-10
Single-Mode Fiber ............oooiiiiiiiiiii B-10
Fiber-to-Pin Diode (RECEIVer) ..........coooiiiiiiiiiiiiin, B-11
Numerical Aperture and Coupling Dxameters ................................ B-11
Bending LOSSES ........cccoiiiiiiiiiiiiiiiiii i B-13

PT PP B-14
System Losses and Varlablllty ............ O
Attenuation Measurement Procedures ... B-14
Field Measurement Procedure for Relative Power Loss e B-15

Equipment Required ..o B-15
PTPTOTPTOP B-15
P
PO
Preparation ............. P
ettt e et et et et et e et s et e e esneaennes B-15
PLOCEAUIE ...ttt
Test Source Adjustment Procedure ... B-16
Measurement Correction Calculation Procedure ............................. B-17
System Certification Prodedure ... B-19
Optical Test Set (Single-Cable) Method ..o B-20
Optical Time Domain RefleCtometry ................ooeiiiiiiiiiiie B-22
teaeteieet ettt eeaaaaans B-23
Equipment Setup ..................oooeis ett
ees B-23
eree
st aaaaneanansaeanaaa
aieaaet
aeeiaia
et
....ouiniteninte
Link MEASUTCIMIEIIT
B-25
...connn.
Bandwidth Analysis and Derating Factors ..................c
Fiber Bandwidth Rating System ...............cccooiiiiiiiiiniiii. B-25
Distance Derating Factors for Ethernet Products ............................ B-25

Suggested System Design Methodology .......................... e B-26

en B-27
aaanaraan
et et
Example Worksheets ......... et eteeeaeeenetee
Sample Loss Calculation .......................... e B-28
Worst-Case and Statistical Results .......................e B-30
Attenuator Installation ...SR B-30
When to Install the Attenuator ..............cooiiiiiiiiiiiiiiiiiia, B-31
Where to Install the Attenuator ..o B-33

How to Install the Attenuator.......................oeeee. PP B-34
Reference Material ............ooiiiiiiiiiii i B-36

Index

Contents-5

Figures
LAN Extended by Means of a LAN Bridge 100 Unit ......................... 1-2
Several LANs Connected with LAN Bridge 100 Units ....................... 1-3
Local and Remote Versions of the LAN Bridge 100 Unit .................. 1-6
LAN Bridge 100 Unit with Plastic Enclosure ................................... 1-7

LAN Bridge 100 Unit with Plastic Enclosure Removed ..................... 1-8
Local LAN Bridge 100 Controls, Status LEDs, and Connectors

(Switches Shown in the Default Position) ....................................... 1-9
Remote LAN Bridge 100 Controls, Status LEDs, and Connectors ..... 1-10
LAN Bridge 100 Configurations .....................ccoccooveeiiiiiiiiiiiiiiiii.. 1-17

Typical Extended LAN with a Local LAN Bridge 100 Unit ............. 1-19
Example Showing a Fiber-Optic Link as a LAN .............................. 1-20
Extended LANs Using Remote Bridges ...................cccooovivniveiiiiin
... 1-21
Remote Bridge to Repeater Configuration ...................................... 1-23

LTM Single Port Configuration

.......................cccccooiiiiiiiiiiiiiiiiii. 1-24

LTM Connected to Two Separate Ethernets

.................................... 1-25

o’

[

NI QN

R W

|
NNNNI}JNMNN
R

LTM on Two Connected Ethernets ...................ccc..cooooiiii... TR 1-26
Bridge in Parallel with a Router ........................c...coocooiiiii
. 1-28

2-11

2-12

2-14

Checking the LAN Bridge 100 Contents .....................ccccoevevveieeeiiii., 2-3
Location of Serial Number and Ethernet Address .............................. 2-5
Setting the Voltage Select Switch ...................o.oooiiiiiiiii 2-6
Verifying Configuration Switch Settings ..................c....cc.o....o.oi.. 2-7
Removing the Plastic Enclosure from the LAN Bridge 100 Unit ....... 2-8
Installing the LAN Bridge 100 Unit in a Rack ................................. 2-9
Fastening the Wall-Mounting Brackets to the LAN Bridge 100 Unit ... 2—-10
Positioning the LAN Bridge 100 Unit for Wall Mounting ............... 2-11
Connecting the Transceiver Cable .....................cc..ccooviviiiiiiiiiiiii.. 2-12
Removing the Protective Caps from the Fiber-Optic Connectors .... 2—13
Connecting the Fiber-Optic Cable ....................ccoocoeviiiiiiiiiiiiii 2-14

Connecting the Power Cord ...............ccooooiiiiiiiiiiiiiiiiiii
i 2-15
Bridge Hardware Verification .............................. e 2-16
LTM Hardware Verification ........................cooooiiiiiiiiiiiiiiiaiiias
i, 2-18
Operating State Transition

.......................... e 3-2

RN
QN

Bit Format of Word 1 in the Receive Descriptor Ring Entry ............ 3-12
Bit Format of Word 2 in the Receive Descriptor Ring Entry ............ 3-13
Bit Format of Word 3 in the Receive Descriptor Ring Entry ............ 3-13
Bit Format of Word O in the Transmit Descriptor Ring Entry .......... 3-14

= NO

.............................veenas 3-8
Descriptor Ring StruCtures ................c.ooviiiiiiiiiiieeeeie
e 3-10
Bit Format of Word 0 in the Receive Descriptor Ring Entry ............ 3-11

-11

LANCE Initialization Block Field Functions

WWWWW%WWW@W

2-15

An Extended LAN with ROUters ....................cooooiviiiiiiiiiniiininniin.. 1-29
Unpacking the LAN Bridge 100 Unit .............ccocooveunniiieinainane 2-2

Bit Format of Word 1 in the Transmit Descriptor Ring Entry .......... 3-15
Bit Format of Word 2 in the Transmit Descriptor Ring Entry .......... 3-16
Bit Format of Word 3 in the Transmit Descriptor Ring Entry .......... 3-17

3-12

Table LOOKUDP PrOoCEeSS

........c...oiivuiiiiiiiiiiiiii

3-13

oo 3-19

Address Table ENtry .........oooooiiiiiiiiiiiiii

e 3-21

AR,

Contents-6

\o Q0

~J O\ W R W N e

|
%%%Af\%»&x%%w
|
[
T
|
| et
lf\fi\!
ARA
NN
R W
e D QO

N
(V)

Binary Decision TIEE€ ...........coooiiiiiiiiiiiiiii 3-22
3-23
Binary Search PrOCESS ..........cccooiiiiiiiiiiiiiiiiiniii
Inserting a New Address into the Ordered Table e 3-26
Result of the Spanning Tree Computation Algorithm ...................... 3-29
Hello Message Propagation ...................coccooiiniaiinn. e 3-30
Hypothetical Extended LAN Using Bmdgcs .......... e 3-32
Logical Extended LAN Formed by Learning Algorithm ................... 3-33
Reconfigured Logical Extended Network ... eeees 3-34
Reconfigured Logical LAN with a New Root Bmdge e 3-35
Extended LAN Showing Physical Loops ....................ooees ORI 3-36

5-10

5-11
5-12
5-13

5-14
B-1

B-2
B-3

B-4
B-5

B-6
B-8

Logical Extended LAN with Backup Bridges ...............cc.cccooooeeiinnn 3-37
Logical Network Topology for New Root Bridge ............................ 3-38
LAN Bridge 100 Subsystems ............ ertreaninirieekeentsaieesstnsnnteaareareraans 4-2

LAN Bridge 100 Functional Block Diagram ....................ccciieieiiee 4-3
ProCessor SUDSYSTEM ........c..uiiiuiiiiiiimiiieriie i e 4-6
Hardware Interrupt Configuration ................ooooiiiiiiiiin. 4-8
Address Mapping Used During Power Up .............ccoooiiiiiiin 4-11
Address Mapping Used for Normal LAN Bridge 100 Operation ....... 4-12
NT SUDSYSTEIM ...ooiiiiiniiiiiiiiiieee ettt ... 4=16
ettt e e eaeeees 4-18
CSR CONLLOL ..ot
oooeeiieeene 4-22
...............cc
Circuit
Memory Control
Table Lookup SUDSYStEM ........coouiiiiiiiiiieiiiiicicii i 4-24
e et e st 4-28
e s e e e e e eiiiiniss
48-Bit COMPATATOT ....ceeeiiiiiiiiiiiiiiii
16-Bit COmparator ................c..ce.ens PRSP 4-29
LAN Bridge 100 Status LEDs ........................... OO 5-3

Troubleshooting Flowchart .................ooooii e 5-6
Removing the Top Plastic Enclosure and Side ches ...................... 5-16

Removing the Bottom Plastic Enclosure ... 5-17
Removing the Chassis Cover for Revision FO8 and Below ............... 5-19
Various Removal Procedures for Revision FO9 and Above ............... 5-20
Fiber-Optic Interface Retaining Hardware for Revision FO8

et 5-21
ANA BELOW .ottt
F08 and Below .... 5-22
Revision
for
Imerface
c
Removing the Fiber-Opti
Removing the Power Supply Connectors and Retaining Screws

PP UPPPPP PPN 5-24
for Revision FO8 and Below ......... RO

Removing the Power Supply for Revision FO8 and Below ............... 5-25

Removing the Power Supply for Revision FO9 and Above ................ 5-26
Power Supply Assembly Removed from the Chassis

(Revision FO8 and BelOW) .........oooiiiiiiiii 5-27

Power Supply Assembly Removed from the Chassis

(Revision FO9 and ABOVE) ... ... 5-28
Locations of Logic Module Retaining Hardware .............................. 5-30
Fiber-Optic System Responsibilities ... B-2
Area of Network Consultant’s Concern-Allowable System Loss .......... B-3
LED Coupling into a Fiber ... B-4
Types of Fiber Misalignment ... e B-9
Fiber-to-Receiver Coupling .............coooiiiiiiiiie B-11
Near-End and Far-End Coupling LOSS€S ...........cocoiiiiiiiiiiii. B-12
B-13
Minignum Bend Radius ...........oooii

End-to-End Cable Loss Calculation Worksheet ............................ ... B=17
Contents-7

B-9

B-10

Calibrating the Optical Test S€t ..............cccooooiimiiiiiiiiiaiieiieaiieiii, B-21

B-12

Measuring the Cable Under Test .............PP B-22
OTDR Test Configuration ..................oooooiiiiiiiiiiiiiiiiiie i B-23
OTDR TraCE ....ooiiiiiiiiiii e B-24

B-13

Sample Worksheet for Budget Loss Calculation

B-11

B-14
B-15

.............................. B-26
Sample Worksheet for DEBET-RH Loss Calculation ......... e, B-28
Sample Worksheet for DEBET-RC Loss Calculation ......................... B-29

B-16

ALLEIMIUALOL . ...oiitiiiiit

B-17

Installing the Attenuator on Both Sides of the Link ........................ B-32
Installing the Attenuator on -RH/-RJ Side Only .............................. B-32

B-18
B-19

B-20

B-21

et e, B-30

Removing the Protective Caps from the LAN Bridge 100 ............... B-34
Connecting the Fiber-Optic Cables to the LAN Bridge 100 ............ B-35
Fiber-Optic Cable Installed to the LAN Bridge 100 ....................... B-36

——— o

Q0 ~J O\ W

R WV DN

—
O

9 N =

N W

mfiu#v*fixfi\fixusuauaua?susuausueuaush*ha_a_apaha
[
T
R
|
|
U
e e 0N i et ek \O Q0
N

Tables

Contents-8

Versions of the LAN Bridge 100 Unit

................coccoooviiiinininien. 1-5
LAN Bridge 100 Product Designations .....................cccoeeuveuniunineinn.... 1-7
LAN Bridge 100 Controls ............ccoooiiiiiiiiiiiii
e, 1-11
LAN Bridge 100 Switches

...t 1-11

LAN Bridge 100 Status LEDS

..........ccooiiiiiiiiiiiiiiii

e, 1-13

LAN Bridge 100 CONNECLOrS .......ccuvuuiiniiniiniieiteiiie

e 1-14

Bit Descriptions of Word O in the Receive Descriptor Ring Entry ..... 3-11

Bit Descriptions of Word 1 in the Receive Descriptor Ring Entry ..... 3-12
Bit Descriptions of Word 2 in the Receive Descriptor Ring Entry ..... 3-13

Bit Descriptions of Word 3 in the Receive Descriptor Ring Entry ..... 3-13
Bit Descriptions of Word 0 in the Transmit Descriptor Ring Entry ... 3-14

Bit Descriptions of Word 1 in the Transmit Descriptor Ring Entry ... 3-15

Bit Descriptions of Word 2 in the Transmit Descriptor Ring Entry ... 3—16
Bit Descriptions of Word 3 in the Transmit Descriptor Ring Entry ... 3—-17

Background Operations ....................cooiiiiiiiiiiiiii i 3-40
Bridge Access SWitChes ..o 3-41
MOP Request and Response MeSSages ................cccoeuviuiinenninainann..n, 3-42
Hardware INterrupts .........ooooiiiiiiiii
e, 4-8
Device Prioriti€s ...........ocoiiiiiiiiiiiiiiiii
e, 4-21
Search Results Register Bit Descriptions ......................coccoeveienn.. 4-31
Controlled Distribution Kits for the LAN Bridge 100 Unit ............... 5=2

Types of Optical Coupling ............coooiiiiiiiiiiiiiiii

e B-4
LED-to-Fiber Coupling Derating Factors for -RC/-RD Variations ........ B-5
LED-to-Fiber Coupling Derating Factors for -RH/-RJ Variations ......... B-6
Splice and Barrel Connector LOSSES ..............cccoeiiiiiniiniiiiiiiiieinn, B-7

Typical Cable Attenuation ................ccoooiiiiiiiiiiiiiiiiiiiiie e B-10
Fiber Size Correction Settings for the Power Meter Reading ........... B-16
Wavelength Correction Values

.................ooooooiiiiiiiiiiiiiiiiiiiiiiaean.
.. B-18
Cable Certification Values ........................ e B-19

Preface

The LAN Bridge 100 Technical Manual is a reference document that provides technical information on the LAN Bridge 100 network bridge. This document provides an overview, installation and

operational information, block-diagram level functional descriptions, and maintenance information on the LAN Bridge 100.

This manual is organized as follows:

Chapter 1

Provides a general overview of the LAN Bridge 100. Configuration considerations

and LAN Bridge 100 specifications are also included.

Chapter 2

Provides installation information for the LAN Bridge 100.

Chapter 3

Discusses LAN Bridge 100 operating states and data structures. Also described are

various features provided by the LAN Bridge 100 such as loop detection and message filtering.

Chapter 4

Provides block-level functional descriptions of the LAN Bridge 100 hardware circuits.

Chapter 5

Describes the maintenance procedures for the LAN Bridge 100.

Appendix A

Defines some terms used in the LAN Bridge 100 Technical Manual.

Appendix B

Describes differences in transmission characteristics among the various types of

fiber used in fiber-optic cables.

Related Documents

Additional information about the LAN Bridge 100 can be found in the following documents:

Communications Options Minireference Manual, Vol. 4 (Order no. EK-CMIV4-RM)

DECconnect System General Description (Order no. EK-DECSY-GD)

DECconnect System Planning and Configuration Guide (Order no. EK-DECSY-CG)
DNA Ethernet Node Product Architectural Specification (Order no. AA-X440A-TK)

Preface-1

DNA Maintenance Operations Functional Specification (Order no. AA-X436

A-TK)

DECnet Router Installation/Operations Manual (Order no. AA-X01

9-TK)

DNA Routing Layer Functional Specification (Order no. AA-X435A-TK)

Ethernet Communications Server DECnet Router Software Installation Guide

(Order no. AA-X019B-TK)

Ethernet Transceiver Tester User’s Guide (Order no. EK-ETHTT-UG)
Field Maintenance Print Set (Order no. MP-01785-01)
Guide to Networking on VAX/VMS (Order no. AA-Y51 2A-TE)
A

LAN Bridge 100 Installation/User’s Guide (Order no. EK-DEBET-UG)

LAN Traffic Monitor Installation Guide (Order no. AA-JP1 5A-TE)
LAN Traffic Monitor User’s Guide (Order no. AA-JP16A-TE)

LAN Traffic Monitor Identification Card (Order no. EK-LANTM-IC)
Remote Bridge Management Software Guide (Order no. AA-FY9 3A-TE)
Attenuator Installation and Configuration Reference Card (Order no. EK-DEFOE-

RC)

ST,

Preface-2

1
Introduction

1.1

Introduction to the LAN Bridge 100 Unit

The introduction of Ethernet and IEEE 802.3 local area networks (LANs) has reduced the cost and
increased the capability of networking. This improvement has resulted in an increased demand for
networking. To accommodate the increased demand, larger and denser networks are being created. However, as networks approach the design limitations of single-LAN technology, their performance may be degraded by the limitations.

The LAN Bridge 100 unit minimizes many single-LAN limitations by creating a high-speed, logical
link between two LANs. Networks that are joined by bridges are called extended LANs. Note that
each individual LAN can be of maximal configuration in terms of length, number of stations, and
other specifications.

The LANs that make up this extended network can be either IEEE 802.3-type or Ethernet-type
baseband or broadband. The LAN Bridge 100 unit provides this logical extension without creating
a bottleneck in the network.

The LAN Bridge 100 unit dynamically learns the locations and station addresses of nodes for each
of the networks that are connected to the bridge. This learning capability enables the LAN Bridge
100 unit to forward packets selectively, based on their destination addresses. In this way, the

bridge can create an extended LAN that has the following advantages:

More stations — Each LAN can still support its maximum number of stations including the
bridge. However, the bridge is transparent to the stations; so the two LANs appear as one larger,
extended LAN to all users.

Larger network — Each LAN can still support the maximum length for a LAN, but the extended
LAN can be much longer.

Reduced traffic — The bridge forwards only nonlocal traffic. Thus, if a large LAN is broken into
several smaller LANs, the traffic on any one of these smaller LANs may be greatly reduced.

1.1.1

Extended Networks

Figure 1-1 shows an example of an extended network that uses a LAN Bridge 100 unit. If these two

networks were joined with a repeater instead of a bridge, all packets that originate on one of the
networks would also appear on the other network. The LAN Bridge 100 unit filters packets so that

only those packets destined for a station on the opposite network are forwarded. Packets destined
for stations on the same network are filtered. This does not mean that all repeaters should or can be

replaced by bridges. The two devices are functionally different, and a careful evaluation of the network requirements is needed to determine which device is more appropriate in a given situation.

ETHERNET OR

802.3 LAN

["LAN 100
BRIDGE |

ETHERNET OR 802.3 LAN
LKG-0726

Figure 1-1:

LAN Extended by Means of a LAN Bridge 100 Unit

The Ethernet-type and/or IEEE 802.3-type networks that a bridge can connect can be two
maximum-length broadband, two baseband (each with the maximum number of stations), or one
of each. The LAN Bridge 100 unit can connect the two Ethernet-type LANs so that they appear as a
single extended LAN. This allows a logical extension of an Ethernet LAN beyond the normal limita-

tions of 2800 meters for baseband or 3800 meters for broadband Ethernet (see Figure 1-2). Broadband and baseband Ethernet networks are each limited to 1024 stations; however, LAN Bridge 100
units can overcome this limitation by joining two or more LANs together, thereby creating an

extended LAN.

1-2

LAN Bridge 100 Technical Manual

1900 METERS MAXIMUM

3800 METERS MAXIMUM

|LAN BRIDGE |

2800 METERS MAXIMUM*

'H4000

DELNI

[LAN BRIDGE |

LA
1

|LAN BRIDGE|

/
185 METERS MAXIMUM

*The 2800-meter (9194-foot) maximum distance between any two nodes is the
sum of two 50-meter (164-foot) transceiver cables, three 500-meter (1640-foot)

coaxial cable segments, four 50-meter (164-foot) transceiver cables connected
to repeaters, and 1000 meters (3280-foot) of point-to-point link.
LKG-0727

Figure 1-2:

1.1.2

Several LANs Connected with LAN Bridge 100 Units

Transparent Operation

Bridge operation is transparent to other stations on the LAN, and no special software is required on
any station unless the LAN Bridge 100 unit is configured to operate as a LAN Traffic Monitor (more

information on the LTM feature is provided in the following sections). Remote Bridge Management
-

Software (RBMS) is available for VMS hosts. RBMS allows you to observe and control
LAN Bridge 100 unit in the network.

any

Introduction

1-3

1.1.3

LAN Bridge 100 to Network Connections

The LAN Bridge 100 unit connects to the network through a transceiver cable and one of the fol-

lowing devices:

H4000 Ethernet transceiver (baseband)

DELNI local network interconnect

DECOM broadband modem

DESTA Ethernet/IEEE 802.3 transceiver

The LAN Bridge 100 unit can be either a local bridge or a remote bridge. The local bridge is connected-to two LANs through transceiver cables. The remote bridge is connected to one LAN

through a transceiver cable and to either another remote bridge or a remote repeater over a fiber-

optic link. Remote bridges can be used when the distance between LANSs is more than 100 meters or
where adverse environmental conditions exist.

The LAN Bridge 100 unit operates at the Data Link layer of the ISO model and therefore is transparent to protocols above this layer. There are no jumper or switch settings required to make the

bridge compatible with either 802.3-type or Ethernet-type LANS.

1.2

LAN Bridge 100 Functional Description

The LAN Bridge 100 unit provides a logical link between two LANs and extends the range of the
LAN. The LAN Bridge 100 unit actually minimizes many of the limitations of conventional LAN Sys-

tems and provides the following advantages:

The LAN Bridge 100 unit selectively forwards or filters (disregards) packets based on the desti-

nation address contained in each packet.

A dynamic address-learning capability enables the LAN Bridge 100 unit to acquire a working

knowledge of the network configuration. By storing source addresses from received packets,

the bridge learns which port (port A or B of the bridge) is associated with each active station on

the extended network.

Once the LAN Bridge 100 unit has built a database of station locations, the bridge selectively

forwards packets based on the destination address. This minimizes network congestion by
keeping local traffic local and allows literally thousands of stations to be connected to the

extended network.

|
e

The LAN Bridge 100 unit functions at the Data Link layer and is protocol independent. This

allows non-Digital Ethernet LANs to be included in the extended LANs. Typical protocols
include DECnet, Xerox Network System (XNS), Transmission Control Protocol (TCP/Internet
Protocol), Local Area Transport (LAT), or any protocols based on Ethernet or on IEEE 802.3

standards.

The LAN Bridge 100 unit has an automatic backup feature that is based on its ability to learn the

locations of other bridges in the extended network. When bridges are configured in a loop, one
of the bridges automatically enters a BACKUP state. Thus the data link loop is broken and the
BACKUP LAN Bridge 100 unit serves as a warm standby. This enhances network availability.

1-4

LAN Bridge 100 Technical Manual

The LAN Traffic Monitor function is optional software available for use with LAN Bridge 100
units at ROM ECO Revision Level E, or later. This optional feature allows the LAN Bridge 100
unit to be used as a base from which to gather traffic data in the form of counters, and periodically forward them to a VAX/VMS system for compilation and analysis. More information on
this optional feature is provided in Section 1.2.4.

Optional Remote Bridge Management Software (RBMS) allows network operators to monitor
and control individual bridges in the extended network. By using RBMS to change bridge
parameters or to load a forwarding database over the network, operators can control the
extended network better.

1.2.1

Product Versions and Designations

There are two versions of the LAN Bridge 100 unit: the local bridge version and the remote bridge
version. Table 1-1 describes the two versions. Both versions are shown in Figure 1-3.
Table 1-1:

Versions of the LAN Bridge 100 Unit

Version

Description

Local Bridge

The local bridge connects LANs separated by 100 meters (328 feet) or less. The

distance from the bridge to either LAN cannot exceed the maximum allowable
transceiver cable length of 50 meters (164 feet).

Remote Bridge

The remote bridge connects LANs separated by more than 100 meters (328 feet)
or where fiber-optic cable capabilities are needed.

A fiber-optic cable is used either to connect two remote bridges or to connect a
remote bridge and a remote repeater.

The fiber-optic cable can be up to 3000 meters (9840 feet) in length when
connecting two remote bridges or up to 1500 meters (4921 feet) when connecting a remote bridge and a remote repeater.

Introduction

1-5

RECEIVE
SYMBOL

TRANSMIT

SYMBOL

FIBER-OPTIC

OUTPUT

CONTROLS, CONNECTORS, AND
INDICATORS ARE SIMILAR TO
THOSE OF THE LOCAL BRIDGE

EXCEPT THAT PORTAIS A
FIBER-OPTIC LINK INTERFACE.
FIBER-OPTIC
INPUT

LAN BRIDGE 100
(REMOTE BRIDGE)

LAN BRIDGE 100
(LOCAL BRIDGE)

LKG-0950
[

Figure 1-3:

Local and Remote Versions of the LAN Bridge 100 Unit

LAN Bridge 100 Technical Manual

There are four model designations of the LAN Bridge 100 unit, as shown in Table 1-2. Note that the
only difference between U.S. and European versions is the product labeling and the voltage selection switch setting.

Table 1-2: LAN Bridge 100 Product Designations
Product

Model

Version

DEBET-AA

Local Bridge

DEBET-AB

Local Bridge

DEBET-RC

Remote Bridge

DEBET-RH

Power Requirements

Labeling

120 Vac Nominal

U.S.

(88 Vac to 132 Vac)

240 Vac Nominal
120 Vac Nominal

U.S.

(88 Vac to 132 Vac)

Extended Remote

120 Vac Nominal

Bridge

(88 Vac to 132 Vac)

DEBET-RD

Remote Bridge

240 Vac Nominal

DEBET-R]

Extended Remote
Bridge

240 Vac Nominal
(176 Vac to 256 Vac)

1.2.2

Non-U.S.

(176 Vac to 264 Vac)

U.S.
Non-U.S.

(176 Vac to 264 Vac)

Non-U.S.

LAN Bridge 100 Packaging

The LAN Bridge 100 unit is housed in a plastic enclosure and is ready for table-top installation. The
plastic enclosure is easily removed to allow for rack or wall mounting. An optional wall-mounting
kit (part number H039) is available that allows mounting the bridge to a wall or partition while it is
still in its plastic enclosure. The LAN Bridge 100 unit contains its own power supply and cooling
fans.

Figures 1-4 and 1-5 show the LAN Bridge 100 unit with and without the plastic enclosure.

LAN BRIDGE 100
(LOCAL BRIDGE)

LKG-~0467

Figure 1-4:

Introduction

LAN Bridge 100 Unit with Plastic Enclosure

1-7

LAN BRIDGE 100
(REMOTE BRIDGE)

LKG-0468

Figure 1-5:
1.2.3

LAN Bridge 100 Unit with Plastic Enclosure Removed

LAN Bridge 100 Controls, Status LEDs, and Connectors

All the controls, status LEDs, and connectors are located on the I/O panel of the LAN Bridge 100
unit. The I/O panel of a local bridge is shown in Figure 1-6, the I/O panel of a remote bridge, in
Figure 1-7. Tables 1-3 through 1-6 describe the controls, status LEDs, and connectors available on

the local and remote bridges.

[
—

1-8

LAN Bridge 100 Technical Manual

SWITCHES

STATUS LEDS

ON LINE

SELF-TEST OK

g g g g gg

DC OK/ PORT B
ACTIVITY \
PORT A\

ACTIVITY

-1

l OOOOO l

.SW!TCHfis

1= LOOP SELF-TEST

2= NVRAM RESET
3= PORT A ACCESS

4 = PORT B ACCESS

p—

5= DOWN-LINE LOAD ENABLE
6=NOT USED

a UP = OFF
j DOWN = ON

Miree

i I( W”( ((W ‘

1 —VOLTAGE SELECT
—AC INPUT

CIRCUIT BREAKER

PORT
B CONNECTOR
PORT
A CONNECTOR
LKG-0960

Figure 1-6:

Local LAN Bridge 100 Controls, Status LEDs, and Connectors (Switches Shown
in the Default Position)

Introduction

1-9

RECEIVE
SYMBOL

TRANSMIT
SYMBOL

FIBER-OPTIC
OUTPUT

FIBER-OPTIC

INPUT

==
CONTROLS, CONNECTORS, AND
1 | INDICATORS ARE SIMILAR TO

» [
E)l|

THOSE OF THE LOCAL BRIDGE
|

(SEE FIGURE A-1), EXCEPT

| THAT PORT A IS A FIBER-OPTIC
LINK INTERFACE.

LKG-0937
AR

Figure 1-7:

1-10

Remote LAN Bridge 100 Controls, Status LEDs, and Connectors

LAN Bridge 100 Technical Manual

—

LAN Bridge 100 Controls

Table 1-3:

Description

Control

Voltage Select Switch

The voltage select switch is used to set the bridge input voltage to the range

required for operation in your country. This switch was factory set for the correct power source for your country. Do not change this switch setting unless you

are sure that the switch setting is incorrect (see your electrician if you are not
sure). Chapter 2, Section 2.4 provides information for changing the bridge input

pr—
b

voltage setting, if necessary.

Circuit Breaker

The circuit breaker provides overcurrent protection for the bridge. If an
overcurrent condition causes the circuit breaker to trip, the white center portion
of the circuit breaker pops out as a visual indication, and the power is cut off from
the bridge. The circuit breaker can be reset by pressing in the white center portion of the circuit breaker.

Bridge Switches

Table 1-4:
-

Switch

_—

Number

These switches control the LAN Bridge 100 functions. Each switch is described in

Table 1-4.

LAN Bridge 100 Switches

Name
Loop

Self-Test

ON (Down)

OFF (Up)

The bridge loops self-test

Bridge runs self-test once on powerup or

continuously after powerup.

reset.

Loopback terminators must

be installed for this test.
This switch position is for

_—
Lo

manufacturing and Field
Service use only.

NVRAM
Reset

NVRAM resets to factory

Prevents NVRAM from resetting to factory

default settings when the
bridge is powered up.
NVRAM Reset removes all
bridge management

default settings when the bridge is powered up. This setting should be used to prevent the loss of parameters stored by
RBMS, during a power failure.

configuration changes.

Port A

Access

Stations on the LAN

Stations on the LAN connected to port A

connected to port A that

that have bridge management capabilities

capabilities are allowed
to read and write

ment parameters.

have bridge management

can read but cannot write bridge manage-

(modify) bridge management parameters.

If a load host resides on port

A, the switch must be on for
down-line loading
software.

Introduction

1-11

Table 1-4 (Cont.):
Switch
Number
4

LAN Bridge 100 Switches

Name

ON (Down)

- OFF (Up)

Port B

Stations on the LAN

Access

connected to port B that have

Stations on the LAN connected to port B
that have bridge management capabilities

bridge management

are allowed to read but cannot write

capabilities are allowed

bridge management parameters.

to read and write (modify)

bridge management
parameters.

If a load host resides on
port B, the switch must

be on for down-line

loading software.
5

Down-Line

Configures the unit to

Load Enable

operate as a LAN Traffic
Monitor. Enables the unit

to down-line load software

Configures unit to operate as a bridge.
RBMS can override the switch causing the

bridge to operate as a LAN Traffic Monitor.

(such as LTM Listener software)
from a load host. When the
switch is on, the bridge does not

forward packets.
[

——

Applicable Port A Access
or Port B Access switch must be

on so that the load host can write

IR,

to bridge memory.

Not Used

NOTES
1.

You can change the switch settings for port access while the bridge is operating.
However, the switch settings for Loop Self-Test, NVRAM Reset and Down-Line
Load Enable are read only during power up. Changing either of these switches

while the bridge is operating has no effect on bridge operation. To change either

of these switches, unplug the unit, change the setting, then plug the unit back in.

Port A and Port B Access switches can prevent bridge management software
from changing any of the bridge’s internal parameters. If security is a concern at

the site, set the bridge’s parameters with RBMS and then disable one or both
ports by putting one or both switches in the up position. Bridge management
software can still read the bridge’s counters and other parameters. Placing either
switch in the down position enables bridge management software write access

from stations on the LAN connected to that port of the bridge. Normally, both
switches are placed in the down position to enable bridge management software

write access from stations on either LAN.

1-12

LAN Bridge 100 Technical Manual

Table 1-5:

LAN Bridge 100 Status LEDs

Name

ON Steady

Port A

Activity

On-line

OFF

Blinking

A message is

No message traffic on

Infrequent messages are being received or

being received

port A.

transmitted on port A or the bridge is checking

or transmitted

for loops (sending Hello messages) about once

on port A.

a second.

Unit is con-

When configured as a

Flashing twice every 2 seconds indicates that

figured as a

bridge, the unit is in

the load host successfully down-line loaded

bridge, is fully

the INITIALIZE,

the LTM Listener software image.

operational,

PREFORWARDING,

and is forward-

BACKUP, or BROKEN

ing messages.

state.

Flashing once each second indicates that the

load host has started the LTM Listener function.

When configured as

an LTM Listener, the

load host has not completed down-line loading the LTM Listener

software image.
(Down-line loading

typically takes up to

two minutes depending on network

traffic.)
DC OK

Internal power

Internal power supply

Internal

supply is

is not functioning

properly.

functioning

properly.

power

supply is not functioning

properly.
Self-Test

Passed self-test.

Running self-test.

NVRAM failed, which requires replacement.
This failure does not affect normal operation. *

OK
Port B

A message is

No message traffic on

Infrequent messages are being received or

Activity

being received

port B.

transmitted on port B or the bridge is checking

or transmitted

for loops (sending Hello messages) about once

on port B.

a second.

NVRAM stores network pointers, parameters, and addresses set by RBMS so that they are not lost in the
event of a power failure. If the Self-test OK LED is blinking, the fault may be bypassed by setting the NVRAM
*«

Reset switch (switch 2 on the I/O panel) to the down (ON) position, and then turning the bridge power off
and on. (See Chapter 5 for more information on troubleshooting.) Note that this causes the bridge to use
default parameters.

Introduction

1-13

Table 1-6:

LAN Bridge 100 Connectors

Connector

Description

AC Input

This connector accepts ac input voltages of 120 or 240 Vac, depending on the setting

of the voltage selection switch (refer to Table A-1).
Port A

For local bridges, this 15-pin, female, D-type connector accepts a transceiver cable. A

slide latch is provided for locking the transceiver cable in place. The pins have the fol-

lowing definitions:

Chassis ground

Collision presence +

Transmit +

Ground

Receive +

+12voltreturn

No connection

Ground

Collision presence —

10.

Transmit -

11.

Ground

12.

Receive -

13.

+12volts

14.

Ground

15.

Ground

OIS

For remote bridges, port A has two fiber-optic connectors. The left-hand connector

(marked—w) is for receiving optical data. The right-hand connector (marked —&) is
for transmitting optical data.
Port B

For local and remote bridges, this 15-pin, female, D-type connector accepts a trans-

ceiver cable. A slide latch is provided for locking the transceiver cable in place. The

pins have the same definitions as for port A (described in this table).

1.2.4

LAN Traffic Monitor (LTM) Option

The LAN Traffic Monitor (LTM) is an Ethernet/IEEE 802.3 LAN monitor that uses the LAN Bridge

100 unit as a hardware base. Either version of the LAN Bridge 100 unit (local or remote) supports
the LTM option. When the LAN Bridge 100 unit is configured to operate as a LAN Traffic Monitor,
bridge

operations

are

suspended

until

the unit is reconfigured for bridge operation. The

LAN Bridge 100 hardware unit processes 48-bit Ethernet addresses, and the LTM software calcu-

lates the statistics. The statistics are periodically reported to a host system that performs additional
data reduction, such as averaging and peak traffic analysis. LTM has two components:
®

The LTM Listener — a LAN Bridge 100 hardware unit that is down-line loaded with LTM monitoring software. The LAN Bridge 100 firmware must be ROM ECO Revision Level E or later.

The LTM User Interface (UI) — remote application software that is installed on any DECnet

VAX/VMS system with an Ethernet controller and associated driver.
For more information about the LAN Traffic Monitor, refer to the LAN Traffic Monitor User’s

Guide.

1-14

LAN Bridge 100 Technical Manual

1.2.5

LAN Traffic Monitor Software

The basic software required for installing and operating the LTM follows:

LAN Traffic Monitor distribution software — installed on each LTM load host.

DECnet Phase IV software, running on VMS Version 4.4 or later — installed on each LTM load
host.

The distribution software must be installed on a load host that runs DECnet Phase IV software and
that is connected to the same extended LAN as the LTM Listener. Digital recommends installing the
LTM Listener software on a load host that is on the same LAN as the LTM Listener. Doing so avoids
the possibility of segmenting the load host from the LTM Listener due to a bridge failure. The distribution software includes an LTM Listener software image file that is down-line loaded to the LTM
Listener. All software must be installed and verified before operating the LTM.
1.2.6

Remote Bridge Management Software (RBMS)

Remote Bridge Management Software (RBMS) is an optional product available for VMS hosts. RBMS
significantly enhances the network’s operation by allowing you to observe and control bridges in
the network. RBMS allows you to:

Understand and modify your network topology by displaying and modifying the bridge for-

warding database.

Evaluate network performance by displaying bridge counters, status, and characteristics.

Troubleshoot network problems by understanding your network topology, disabling selected
bridges to segment your network, and signaling selected bridges to run their built-in self-test
diagnostics.

Save your configuration data in the bridge’s nonvolatile RAM (NVRAM) so that it is not lost dur-

ing a power failure. Changes are saved by bridge hardware.

Remotely switch the LAN Bridge 100 unit between bridge usage and LTM Listener.

Remotely determine whether the LAN Bridge 100 unit is operating as an LTM Listener or a
bridge.

For more information on RBMS, refer to the Remote Bridge Management Software Guide.

Introduction

1-15

1.3

Configuration Considerations

This section describes some configuration considerations that apply to implementing extended

LANs with LAN Bridge 100 units. Refer to the DECconnect Planning and Configuration Guide
(EK-DECSY-CG) for additional information on LAN configuration.

LANSs connected by bridges appear as one extended LAN as far as data traffic is concerned.

Individual LANs that are connected by bridges can each be configured for up to the normal maximum for length, number of stations, and other specifications. For example, each Ethernet
baseband LAN can be up to 2800 meters (9194 feet) in length and have 1024 stations.
Extended LANs may consist of combinations of any of the following LANs joined by bridges:

Ethernet baseband

Ethernet broadband

IEEE 802.3 baseband (10base5)

ThinWire Ethernet (10base2)

Within these configurations, bridges may be connected to the network through transceivers. Alter-

natively, remote bridges may be connected directly to other remote bridges or remote repeaters.
Also, a bridge may be connected to a DELNI that may or may not be connected to a remote net-

work. Figure 1-8 shows some of the possible configurations involving bridges.

1-16

LAN Bridge 100 Technical Manual

S0mmed BROADBAND TAPs
IEEE 802.3 BASEBAND

] 802.3 MAUet

|LAN BRIDGE 100|

L]
DELNI

DECOM

[LAN BRIDGE 100

ETHERNET BASEBAND

[CAN BRIDGE 100]

| FIBER—OPTIC

[LAN BRIDGE 100]

!. FIBER—OPTIC

LINK

REMOTE REPEATER]

| LINK

LAN BRIDGE 100}

B My §02.3 MAU et
ETHERNET BASEBAND

IEEE 802.3 BASEBAND
LKG—-0857-87

Figure 1-8:

Introduction

LAN Bridge 100 Configurations

1-17

1.3.1

Performance Considerations

A packet may have to travel through a number of bridges before reaching its destination. Note that
increasing the number of bridges in the data path causes corresponding increases in the data path

delay. This delay could have a negative impact on network performance, especially with timecritical protocols or with interactive tasks such as character echoing for users on terminal servers.
A general rule for networks with typical traffic loading is that performance may start to degrade if a
packet must travel through more than seven bridges to get from its source station to its destination
station.

1.3.2

Loop Considerations

When the LAN Bridge 100 unit is turned on, it executes an internal self-test. This test takes about 15
seconds. The bridge then spends about 30 seconds learning station addresses and communicating
with other bridges in the network to determine whether there are any loops (multiple paths

between two or more LANS).
When bridges in an extended LAN form a loop, a loop detection process determines that one or
more of the looped bridges enters the BACKUP state, so that only one path exists between any two
LANSs (the loop detection process is described in Chapter 3).

If an on-line bridge fails, a backup bridge takes over and begins forwarding packets."With RBMS
software, bridges can be selectively placed in the BACKUP state. In this way the most direct path

with the fewest number of bridges can be provided for the heaviest network traffic.
If the bridge is in a loop with a repeater, the bridge enters the BACKUP state. The bridge continues

to check the loop through the repeater about once a second. If the repeater fails, the bridge automatically takes over and begins forwarding packets.
AR

1.3.3

Local LAN Bridge 100 Considerations

The local LAN Bridge 100 unit (DEBET-AA or DEBET-AB) connects two LANs that are separated

by less than 100 meters (328 feet). This distance is made up of the combined length of two transceiver cables, each 50 meters (164 feet) in length. Figure 1-9 shows a typical extended LAN configuration using a local LAN Bridge 100 unit.

1-18

LAN Bridge 100 Technical Manual

l STATION l

STANDARD ETHERNET
CABLE

_TRANSCEIVER CABLES

[oEBET—AA/BB]

(H4000 |

EACH UP TO 50 m (164 ft)

{ H4000 |

| STATION l
LKG—-0855—87

Figure 1-9:

Introduction

Typical Extended LAN with a Local LAN Bridge 100 Unit

1-19

1.3.4

Remote LAN Bridge 100 Considerations

Two remote LAN Bridge 100 units (DEBET-RC/RD) connect two LANs that are separated by up to

2000 meters (6560 feet) or two Extended Remote LAN Bridge 100 units (DEBET-RH/RJ) connect

two LANs separated by up to 3000 meters (9840 feet) in length. The length of the fiber-optic link
joining the two remote bridges can be up to 3000 meters (9840 feet) in length.

The fiber-optic link has all of the characteristics of a LAN except that it has no stations and its

P—

length cannot exceed 3000 meters (9840 feet). As such, the fiber-optic link must be included in

the path cost of the extended LAN (more information on path cost calculation is provided in

Chapter 3). The extended LAN example shown in Figure 1-10 is drawn to show that the fiberoptic link between the remote bridges is equivalent to a LAN.
RO

EQUIVALENT DIAGRAM

EXTENDED LAN EXAMPLE

STATION

| DEBET—RC/RD |
-

| DEBET-RC/RD |

| DEREP—RA
/RSB |
SRR

FIBER—OPTIC LINK

STATION

| pEBET—RC/RD |

LKG-0731—-87

Figure

1-10:

Example Showing a Fiber-Optic Link as a LAN
A,

T
—

1-20

LAN Bridge 100 Technical Manual

Figure 1-11 shows some of the possible extended LAN configurations using remote bridges.

UP TO 50 m

(164 ft)

| oEBET-RC/RD |

l STATION

FIBER—OPTIC LINK

UP TO 2000 m (6562 ft)

| bEBET—RC/RO |
UP TO 50 m

(164 ft)

{ H4000 “

UP TO 50 m
(164 m)
-

STATION

| DEREP—RA/RB |
i-'

FIBER—OPTIC LINK

UP TO 1500 m (4922 ft)

|, DEBET-RC/RD |

STATION

UP TO 50 m

(164 m)

LKG~0858—87

Figure 1-11:

Extended LANs Using Remote Bridges

1.3.4.1 Fiber-Optic Cable Between LAN Bridge 100 Units — In a bridge-to-bridge configuration, the dual-cable fiber-optic link that connects the bridges does not affect the cable configuration guidelines of either of the LANs connected to the bridges. The end-to-end light loss of the
cable must not exceed 12.5 decibels (dB) for the DEBET-RC/RD and 16 decibels (dB) for the
DEBET-RH/R] versions, regardless of the length. The type of optical fiber used affects the length of
the cable. The loss budget is based on Corning 1508, 100/140 optical fiber with a bandwidth of
300 MHz/km measured at 820 nanometers.

CAUTION

Exceeding the loss limit may cause the bridge configuration to fail (see Appendix B
for information on attenuation budget allocation).

Introduction

1-21

To achieve long distances, particularly those more than 1000 meters (3280
feet), installation of the
fiber-optic cable must be carefully planned. The type and quality of the
cable’s optical fiber, the
cable repair strategy, and the cable’s total end-to-end light loss are of great
importance to successful
bridge installation.
The end-to-end cable light loss depends on the quality of the fiber, the

number and quality of
splices required for installation, and the number and quality of the connectors used.

The cable repair strategy affects the budget in that cable repair typically consists of replacing
section of cable. This requires two splices. The repaired link must remain

under the end-to-end

light loss budget. If the initial installation uses the entire loss budget,
a repair would not be
possible. Therefore, plan for a minimum of two splices (about 0.5
dB for each splice).

For longer cable runs or for installations requiring more splices, request a lower-loss
fiber-optic
cable from your cable vendor. For information on measuring light loss through
a fiber-optic link,
refer to Appendix B.

1.3.4.2

LAN Bridge 100 to Repeater Considerations — When a remote bridge is connected

a remote repeater, the fiber-optic cable can be up to 1500 meters (4920

feet) long. When a remote

bridge is connected to a remote repeater, the length of the fiber-optic cable is considered
part of the
length of that LAN (see Figure 1-12). Therefore, the distance between the bridge
and the farthest
end station on the LAN cannot exceed 2800 meters (9184 feet).
The 2800-meter (9184-foot) distance between any two stations on a typical Ethernet

up of the following:

LAN is made

®*

Two 50-meter (164-foot) transceiver cables (connecting the farthest end stations)

Three 500-meter (1640 foot) coaxial cable segments

Four 50-meter (164 foot) transceiver cables (connected to repeaters)

1000 meters (3280 feet) of point-to-point fiber-optic link

1-22

LAN Bridge 100 Technical Manual

BROADBAND TAP

|[EEE 802.3 BASEBAND

oecom

LAN BRIDGE 100}

DELNI

‘

ETHERNET BASEBAND

TAN BRIDGE_100]

[CAN_BRIDGE 100}

| FIBER—OPTIC

p—

| LINK

l FIBER—OPTIC

LINK

REMOTE REPEATER

[CAN BRIDGE 100]

ETHERNET BASEBAND

IEEE 802.3 BASEBAND
PRELIMINARY

Figure 1-12:

Introduction

Remote Bridge to Repeater Configuration

1-23

R——

1.4

LAN Traffic Monitor Configurations

The following sections describe several ways for configuring the LAN Bridge 100
an LTM Listener unit.

hardware unit as

For more information about the LAN Traffic Monitor, refer to the LAN

Traffic Monitor User’s Guide.

1.4.1

Single Port Configuration with Loopback Connector Installed

As shown in Figure 1-13, the LTM Listener always monitors Ethernet 2 and sends

statistics to the
user interface on Ethernet 2. The LTM Listener can send statistics to a user interface
on Ethernet 1
also, as long as the LAN Bridge 100 unit connects the two Ethernets. Note that port
B has aloopback
connector installed and is not in operation.

NOTE

The LAN Bridge 100 unit fails self-test if an unused port is left disconnected unless

the unused port is the fiber-optic port used with the remote version (DEBET-RC/
RD). An Ethernet loopback connector (shipped with the unit) must be connected to
the unused transceiver port. Do NOT install a fiber-optic loopback connector to an

unused fiber-optic port. A fiber-optic loopback causes the bridge to fail the power-

up self-test.

ETHERNET 1

wryw

1H4000 |

=
AR

PORT A

[LAN BRIDGE 100

PORT B

ETHERNET 2

| H4000|
|

/J PORT B

LOOPBACK
CONNECTOR

1-24

{ H4000

PORT A

| LT LISTENER |

Figure 1-13:

{ H4000 |

VAX /VMS

(LTM USER

INTERFACE)
LKG—0858—87

LTM Single Port Configuration

LAN Bridge 100 Technical Manual

1.4.2 Dual Port Connections Between Two Ethernets
As shown in Figure 1-14, the LTM Listener is connected to two completely separate Ethernets. In
this case, the LAN Traffic Monitor can monitor either Ethernet 1 or 2 but must report to the LTM
user interface on Ethernet 1.

VAX /VMS
(LTM USER |
INTERFACE)

‘?

;

}
{ H4000

ETHERNET 1
'

PORT A

| LTTM UISTENER |
'

PORT B

ETHERNET 2
LKG—0859—-87

Figure 1-14:

Introduction

LTM Connected to Two Separate Ethernets

1-25

1.4.3

Dual Port Connections With Bridged Ethernets

The configuration shown in Figure 1-15 describes two Ethernet LANs bridged together, forming a
single extended LAN. The LTM Listener can monitor either Ethernet 1 or 2 and can report to either

port.

VAX /VMS

(LTM USER

INTERFACE)
TSR

H4000

ETHERNET 1
H4000

:
PR

PORT A
L_LTTMTM LISTENER |

| PORT A
|LAN BRIDGE 100
| PORT B

' PORT B
w

ETHERNET 2

LKG—-0860—-87

Figure 1-15:

LTM on Two Connected Ethernets

NOTE
Digital recommends that you configure the LTM Listener to report on the port that
has the least number of intervening bridges between it and the LTM user interface

host(s). Doing so minimizes the impact of a possible bridge failure.
A,

1-26

LAN Bridge 100 Technical Manual

1.5

Bridge/Router Considerations

This section discusses some considerations for configuring extended LANs that include LAN Bridge
100 units and routers.

NOTE

The information in this section applies specifically to equipment manufactured by
Digital Equipment Corporation. However, all routers and bridges designed for layered architecture networks adhere to similar operating principles and may have similar parameters that should be considered when implementing networks with
bridges and routers.

The configuration shown in Figure 1-16 involves a LAN Bridge 100 unit that is installed in parallel
with a router. This configuration creates many problems and therefore is not recommended (see
Section 1.5.1). The configuration shown in Figure 1-17 is a general case in which two local area
networks are connected by a LAN Bridge 100 unit. This configuration does not have associated
problems, provided that routing parameters are set properly in all routers.

The information contained in this section assumes that the reader is familiar with the concept of
routing and how routers operate on LANs. For more information on the subject of routing, refer to
the following documents:

DECnet Digital Network Architecture Phase IV: Routing Layer Functional Specification
(Order no. AA-X435A-TK)

Guide to Networking on VAX/VMS (Order no. AA-Y512A-TE)

FEthernet Communications Server DECnet Router Software Installation Guide
(Order no. AA-X019B-TK)

The LAN Bridge 100 unit operates at the Data Link layer and is transparent to higher protocol layers

in the Digital Network Architecture (DNA) model (such as Routing, End Communication, Session,
and so on).

Since the LAN Bridge 100 unit connects two local area networks to form an extended LAN, higher
protocol layers effectively see a single LAN. For the Routing layer, the creation of an extended LAN
means either the addition of nodes (stations) and/or routers to the existing network or the actual
merging of two distinct networks.

NOTE

An Ethernet repeater and LAN Bridge 100 unit are similar since both devices provide
a channel through which data can pass (a repeater connects segments together to
form a LAN, while a bridge connects LANs together to form an extended LAN). In
addition, both devices are transparent to the higher protocol layers in the DNA
model. Thus both devices introduce similar symptoms if implemented incorrectly.

Introduction

1-27

1.5.1

LAN Bridge 100 Unit in Parallel with a Router

The configuration in which a LAN Bridge 100 unit is installed in parallel with a router (see
Figure 1-16) does not enhance network performance. Since the LAN Bridge 100 unit can handle all

of the traffic on both LANS, there is no need to have a router in parallel with the bridge to ‘‘assist’’

it.

In fact, in this configuration, both router ports are connected to the same extended LAN. This

causes problems for the router since the router may receive two copies of the packet. Consequently, if the router is a designated router, it will hear its own Hello message on the other link. In

response, the router defers to the ‘“‘echo’’ Hello message and ceases to be the designated router for
the LAN. This has a major impact on the network because end nodes on the LAN will not be able to
communicate with nodes that are not on the LAN.

END

NODE

ROUTER
END

END

NODE

1
|

LKG-0734

Figure 1-16:

Bridge in Parallel with a Router

There are additional problems associated with having routers receive multiple copies of packets on

a LAN. However, it is not necessary to describe them all here. The network manager should adhere

to the following guideline:
®

A router should never have two ports connected to the same LAN or extended LAN. That is, a
router should never be in parallel with a bridge or a repeater.

P
—

1-28

LAN Bridge 100 Technical Manual
A

1.5.2

Modifying Router Parameters for Extended LANs

This section identifies router parameters that must be set to reflect the Extended LAN. The parameters may need to be modified when two LANs are joined with a LAN Bridge 100 unit (see
Figure 1-17).

Each of the two LANs shown in Figure 1-17 has many nodes, some of which are broadcast routers
(levels 1 and 2), while the remaining nodes are broadcast nonrouters (end nodes). (For a brief
description of level 1 and level 2 routing, see Section 1.5.3.) When the bridge is installed, the network grows larger. Consequently, the following parameters may have to be modified in all routers
in the extended LAN:

TM
o

MAXIMUM BROADCAST NONROUTERS

MAXIMUM BROADCAST ROUTERS

{ ROUTER]

ROUTER '

[ROUTER}
|

[eEn0

| NODE|

| enD [
- NODE

END

BRIDGE

ROUTER

{

NODE |

NODE
|

|
END

ROUTER

LKG-0735

Figure 1-17:

Introduction

An Extended LAN with Routers

1-29

The MAXIMUM BROADCAST NONROUTERS parameter specifies the maximum
number of
Ethernet nonrouting nodes (end nodes) that the router database can contain
at any one time. This

parameter can range from O to 1022 and has a typical default value of 64.

The MAXIMUM BROADCAST

NONROUTERS parameter must be greater than or equal to the actual

number of nonrouting nodes on the LAN. The larger this number is, the greater

the memory over-

head for the router.However, if this number is less than the number of active end nodes, some

nodes will be unreachable because of limits on the size of the router’s database.

end

NOTE
When nodes on a LAN are configured into multiple DECnet areas, this parameter is
se€t on a per-area basis. For more information about areas, see Section 1.5.3.
Also

refer to the Ethernet Communications Server DECnet Router Software Installation

Guide.

If the MAXIMUM BROADCAST NONROUTERS parameter is set too low, it is difficult to

diagnose

R—

on networks with many nodes (both routing nodes and end nodes). One of the symptoms
of this
parameter being set too low is partial partitioning of the network. Under this
condition, some

nodes may lose their ability to initiate conversations with other nodes even though they
same LAN.

are on the

SO,

Therefore, whenever network changes are made, the network manager should take steps
to verify
(and, if necessary, modify) this parameter in all routers so that it is either equal to or greater
than
the number of nonrouting nodes in the extended LAN.
O

The MAXIMUM BROADCAST ROUTERS parameter specifies the maximum number of Ethernet
routing nodes that the router database can contain at any one time (this applies to routing

nodes on

the same Ethernet). Routing nodes include level 1 and level 2 routers (see Section 1.5.3). In

a multiple area network, this parameter must include all the level 1 routers in a particular area plus
all the
level 2 routers in the entire network. This parameter can range from 0 to 32 and has
a typical
default value of 10.
The larger the value for this parameter, the greater the memory overhead for the router.

more routers there are on the Ethernet, the greater is the control traffic associated with

rary looping property of the routing algorithm.

A5,

Also the

the tempo-

If this parameter value is too small, some routers will be unreachable because of limits on the

size of
That is, when the MAXIMUM BROADCAST ROUTERS parameter is
exceeded, routers with the lowest priority are dropped from the router’s database.

the router’s database.

Therefore, whenever network changes are made, the network manager should take steps

to verify

PSS

(and, if necessary, modify) this parameter so that it is either equal to or greater than the
number of
level 1 routers in the DECnet area plus the number of level 2 routers in the extended LAN.

1.5.3

Setting Up Multiple Areas

When the total number of routers and/or nonrouters on the extended LAN approach the maximum

values, the network manager can set up multiple areas on the Ethernet.

1-30

LAN Bridge 100 Technical Manual

SRR

In general, each area is a group of nodes. Nodes are grouped together in areas for purposes of hierarchical routing. Hierarchical routing involves the addition of a second level of routing to the net-2
work. Routing within an area is referred to as level 1 routing; routing between areas is called level
routing. When creating multiple areas on the Ethernet, the number of routers and nonrouters are
considered separately for each area.

For more information on configuring multiple area LANs with routers, refer to DECnet Digital

Network Architecture Pbase IV: Routing Layer Functional Specification.

1.6

Specifications

Physical specifications

Environmental specifications

Specifications for the LAN Bridge 100 unit are divided into the following categories:

e Electrical specifications

1.6.1

—

Physical Specifications

The LAN Bridge 100 unit is housed in a plastic table-top enclosure that has the following
dimensions and weight.

With plastic enclosure:

Revision FO8 and below

Revision FO9 and above

Height 16.2 cm (6.4 in)
Width 49.4 cm (19.4 in)

Depth 31.3 cm (12.3 in)
Weight 9.5 kg (21 1b)

Depth 31.3 cm (12.3 in)
Weight 6.7 kg (15 1b)

Without plastic enclosure

Revision FO8 and below

Revision FO9 and above

Height 13.3 cm (5.3 in)
Width 43.6 cm (17.2 in)
Depth 29.8 cm (11.7 in)
Weight 7.3 kg (16 1b)

Height 13.3 cm (5.3 in)
Width 43.6 cm (17.2 in)
Depth 29.8 cm (11.7 in)
Weight 4.5 kg (10 1b)

AC power-cord length

U.S. 1.83 m (6 ft)

Transceiver cable length

BNE4 12 m (39 ft) maximum

Others 2.5 m (8.2 ft)

BNE3 50 m (164 ft) maximum

Fiber-optic cable

Type

Corning 1508 (100/140)

Minimum bandwidth

300 MHz measured at a wavelength of 820 nm

Introduction

1-31

Fiber-optic connectors

Type

Stainless steel, Amphenol type 906, SMA style, or equival

Maximum attenuation

Less than 1.5 dB

ent

The plastic enclosure is easily removed, and brackets are provide
d to allow mounting of the unit on
a wall or in a cabinet. An optional kit (part number HO039)
is available for mounting the bridge on a
wall or partition without removing the plastic enclosu
re.

1.6.2

Environmental Specifications

The LAN Bridge 100 unit is designed to operate in a non-air

-conditioned environment or in an

exposed area of an industrial site. However, 50°C is the maximu

must not be exceeded at the air intake of the bridge.

m ambient temperature which

This applies even when the LAN Bridge 100

unit is mounted in a cabinet. The bridge is not intende

d to operate in an air plenum.

Operating Environment:

Temperature

5°Cto 50°C (41°F to 122°F)
R

Maximum rate of change

20°C/hr (36°F/hr)

Relative humidity

10% to 95% (noncondensing)

Wet-bulb temperature

32°C (90°F) maximum

Dew point

2°C (36°F) minimum

Altitude

Sea level to 2.4 km (8000 ft)

Air flow

37.5 CFM. About 10 to 15 cm (4 to 6 in) of space must be
provided
on both ends of the unit for adequate air flow.

Shipping Environment:

Temperature

-40°C t0 66°C (-40°F to 151°F)

Relative humidity

0% to 95% (noncondensing)

Altitude

Sea level to 9.1 km (30,000 ft)

—

1-32

LAN Bridge 100 Technical Manual

1.6.3

Electrical Specifications

The LAN Bridge 100 unit requires the following power for proper operation:
AC input power

Revision FO8 and below

Revision FO9 and above

Voltage

Range

120V

88 Vito 132V

88Vt 132V

240V

176 V to 264 V

176 Vto 264 V

Current

At 120V

1.6 amps

1.1 amps

At 240V

0.9 amps

0.7 amps

Frequency

47 to 63 Hz

Power

160 W

168 W

Consumption

The LAN Bridge 100 unit provides the following power features:
e

Self-contained power supply

Power-cord options for all major DIGITAL markets

Adequate power to drive two external transceivers

Introduction

1-33

OIS

——

2
Installation and Verification

2.1

Overview

This chapter provides all the information necessary for successful installation and subsequent
checkout of the LAN Bridge 100 unit. Included are instructions for:
e

Unpacking and inspection

Step-by-step installation

Verification of operation

2-1

2.2

Unpacking

Unpack the LAN Bridge 100 unit (Figure 2-1).

BUBBLE PACKING* --~...

LAN BRIDGE 100 HARDWARE

INSTALLATION /OWNER’S GUIDE
LAN TRAFFIC MONITOR

'IDENTIFICATION CARD

o @_—~ POWER CORD
GO

LAN BRIDGE 100
SHIPPING CARTON

"BUBBLE PACKING CONTAINS LOOPBACK CONNECTORS, MOUNTING BRACKETS, AND

SCREWS

LKG-0953

Figure 2-1:

Unpacking the LAN Bridge 100 Unit

2-2

LAN Bridge 100 Technical Manual

Check the contents of the carton for damage and missing parts (see Figure 2—2). In case of damage,
contact your delivery agent and your Digital sales representative. In case of missing parts, contact

your Digital sales representative.

Wfl

W
h

[[M]?J

/e0
me

LAN BRIDGE 100

MOUNTING BRACKETS

|1/ J

SCREWS

LOOPBACK CONNECTOR
PIN 12-22196-01

2 FOR DEBET-AA AND -AB
1 FOR DEBET-RC AND -RD

LAN BRIDGE 100 HARDWARE
INSTALLATION /USER’S GUIDE

LAN TRAFFIC MONITOR

POWER CORD

IDENTIFICATION CARD
LKG-0951

Figure 2-2:

Checking the LAN Bridge 100 Contents

Installation and Verification

2-3

2.3

Arranging for Software Installation

If you are installing the LAN Bridge 100 unit as a bridge, software installati

Section 2 .4.

on is not required. Go to

If you are installing the LAN Bridge 100 unit to operate as an LTM Listener, arrange
to have the LTM
distribution software installed on a load host while you install the hardware.

The identification card (ID card) that you received with the unit provides informatio

required by the person installing the software. This information includes the

n that is

LAN Bridge 100 serial

number and Ethernet address (see Figure 2—3), where the device is to be located, and your
Fill in this information at the top portion of the ID card, and then give the card

name.

and any software

cartons you received with the shipment to the system/network manager.
The system/network

manager can then fill in the required information as described on the ID card and

arrange for the

PO,

software installation. Ask to be notified when the LTM distribution software is
installed on a load

host and when the LAN Bridge 100 unit is configured in the load host database. You
can continue
with the hardware portion of the installation, but do NOT power up the unit until notified
that software installation is complete and that the LAN Bridge 100 unit is configured
in the load host database.

—

HAIIAGOI,

2-4

LAN Bridge 100 Technical Manual
[

U
SERIAL
NUMBER

ETHERNET
ADDRESS

AS72000199

S/N___

[ 08-00-2B—02—B3-AF ]

ASY

s A

56.3

”"*’W

38-22887-01-81

DEBET-AA

P/N__

INS

REVISION
LEVEL
LKG-0961

Figure 2:«3: Location of Serial Number and Ethernet Address

Installation and Verification

2-5

2.4

Installation

Installation consists of setting switches, placing the unit and connecting cables.

2.4.1

Setting the Voltage Switch

Determine the operating voltage for the LAN Bridge 100 unit by comparing your power
cord to
those shown in Figure 2—4. If necessary, set the voltage switch to match the power-cord
voltage.

Do not connect the power cord at this time.

CAUTION
An incorrect voltage setting can damage the LAN Bridge 100 unit.
I

100 V/115V/120V

SLIDE THE SWITCH SO
_

THAT 120IS VISIBLE

220 V/230 V/240 V SLIDE THE SWITCH SO
THAT 240 IS VISIBLE
\/

t—

LKG-0931

Figure 2-4:

2-6

Setting the Voltage Select Switch

LAN Bridge 100 Technical Manual

2.4.2

Verifying Configuration Switch Settings

Check to see that your configuration switch settings are properly set, as shown in Figure 2-5.
NOTE

If you are installing the LAN Bridge 100 unit as only an LTM Listener, set Switch 5,
Down-Line Load Enable, to ON (down = ON). You can use RBMS to switch between
LTM functions and bridge operation if Switch 5 is OFF (up = OFF).

SWITCHES
SWITCHES

1=LOOP SELF-TEST

2 = NVRAM RESET
3=PORT AACCESS
4 = PORT B ACCESS
5= DOWN-LINE LOAD ENABLE
6= NOT USED

BBQQUB
12

456

m—

B UP = OFF
U DOWN = ON

LKG-0959

Figure 2-5:

Verifying Configuration Switch Settings

Installation and Verification

2-7

The LAN Bridge 100 unit is housed in a plastic, table-top enclosure. This plastic enclosure can be

removed, and the sheet metal unit inside it can be either wall mounted or mounted in a standard

48-cm (19-inch) rack. A set of mounting brackets for both rack mounting and wall mounting is

included.

NOTE
Whichever installation you choose, leave enough clearance around the bridge’s air

inlets and outlets to ensure optimal air flow.

Rack mounting requires removal of the bridge’s plastic enclosure. For wall mounting, the bridge
can be installed with or without its plastic enclosure. To mount the LAN Bridge 100 unit with the

enclosure on a wall or partition, order the special wall-mounting kit (part no. H039). To remove
the plastic enclosure, take the following steps (see Figure 2-6).

Turn the bridge upside down.

Remove the 8 screws from the bottom of the unit.

Remove the plastic enclosure.

RSO

LKG-0952
S

Figure 2-6:

Removing the Plastic Enclosure from the LAN Bridge 100 Unit
o

CAUTION
A

Do not reinstall screws in the bridge’s metal casing. Doing so could damage the
bridge.

2-8

LAN Bridge 100 Technical Manual

p—

2.4.3

Rack-Mount Installation (Optional)

|
To rack mount the LAN Bridge 100 unit, take the following steps (see Figure 2-7).

1. Fasten the mounting brackets to the bridge with the screws provided.
2.

Fasten the bridge to the rack (screws not provided).

LKG-0932

Figure 2-7:

Installing the LAN Bridge 100 Unit in a Rack

NOTE
The air inside the rack may be hotter than the ambient room temperature. There-

fore, be sure that the air entering the bridge’s air inlet does not exceed the maximum

In addition, be sure that all cables connecting to a rack-mounted bridge are dressed
to the rack to relieve strain. If fiber-optic cables are used, the cables must not exceed

air-inlet temperature of 50°C (122°F).

their 15-cm (6-inch) bend radius.

Installation and Verification

2-9

2.4.4

Wall-Mount Installation (Optional)

To wall mount the LAN Bridge 100 unit without the plastic enclosure, take the following steps:
1.

Fasten the mounting brackets to the bridge with the screws provided (Figure 2-8).

NOTE
To meet federal safety codes, never install a bridge with the I/O panel facing
downward.

I/0 PANEL
-

GBS

AIR INLET

LKG~-0938

Figure 2-8:

Fastening the Wall-Mounting Brackets to the LAN Bridge 100 Unit

2-10

LAN Bridge 100 Technical Manual

Fasten the brackets to the wall (screws not provided) (Figure 2-9).

NOTE
o—

The air entering the bridge’s air inlet must not exceed the bridge’s maximum

temperature of 50°C (122°F). If two bridges are mounted side by side, one
bridge receives the heated exhaust air from the other. This usually does not
cause a problem unless the ambient air temperature is close to 50°C (122°F). Do
not mount several bridges side by side.

AIR INLET

SCREWS NOT
SUPPLIED

LKG-0954

—

Figure 2-9:

Positioning the LAN Bridge 100 Unit for Wall Mounting

Installation and Verification

2-11

2.4.5

Connecting the Transceiver Cables

If you are installing a DEBET-RC/RH or DEBET-RD/R]J bridge with a fiber-optic cable, there will be

only one transceiver cable (see Figure 2-10).
1.

Push the slide latch to the left to the unlocked position.

Plug the transceiver cable into the jack.

Push the slide latch to the right until it snaps into the locking position.

Gently pull on each connector to make sure that the latch is secure.

| PUSH SLIDE LATCH

| FIRMLY TO RIGHT
| TO LOCK CABLES

| IN PLACE

\2
P

AR,

LKG-0933

Figure 2-10:

Connecting the Transceiver Cable

——

SRS

2-12

LAN Bridge 100 Technical Manual
S

2.4.6

Connecting the Fiber-Optic Cable

If you are installing a DEBET-RC/RH or DEBET-RD/RJ bridge with a fiber-optic cable, use the
following procedure to connect it (see Figure 2-11):
WARNING

Never look into a fiber-optic connector or cable. High-intensity light can damage
your eyes.

Pull the protective caps off the fiber-optic connectors (Figure 2—-11).

FIBER-OPTIC

PROTECTIVE
CAPS

PROTECTIVE

CAPS

FIBER-OPTIC CABLE

LKG-0934

Figure 2-11:

Removing the Protective Caps from the Fiber-Optic Connectors

Installation and Verification

2-13

Note the polarity of the cable connections. The transmit cable (identified on the cable by an
arrow pointing away from the connector) must be connected to the fiber-optic output of the

bridge. The receive cable (identified on the cable by an arrow pointing toward the connector)
must be connected to the fiber-optic input of the bridge.

Connect the fiber-optic cable. Tighten the fiber-optic'connectors so they are only finger tight

(see Figure 2-12).

CAUTION
The fiber-optic cable will be damaged by sharp bends. Do not allow the bend radius

to be less than 15 cm (6 inches).
AR

FIBER-OPTIC FIBER-OPTIC

INPUT

OUTPUT

RECEIVE

BEND RADIUS
15 CM (6 IN)

FIBER-OPTIC CABLE

LKG-0955

Figure 2-12:

—

Connecting the Fiber-Optic Cable
TS

2-14

LAN Bridge 100 Technical Manual
O

2.4.7

Connecting the Power Cord

To connect the power cord, take the following steps (Figure 2—-13):

Read and remove the yellow CAUTION label.

Plugin both ends of the power cord.

Figure 2-13:

Connecting the Power Cord

2.5

Verifying the Installation

panel. The LED states vary depending whether the LAN Bridge 100 hardware unit is configured to

e s

operate as a bridge or as a LAN Traffic Monitor.

Proper installation of the LAN Bridge 100 unit is verified by the state of the status LEDs on the I/O

If you are installing the LAN Bridge 100 unit to operate as a bridge, go to Section 2.5.1 to verify corL

rect installation.
If you are installing the LAN Bridge 100 unit to operate as a LAN Traffic Monitor, go to Section 2.5.2

to verify correct installation.

Installation and Verification

2-15

2.5.1

Verifying the Bridge Installation

Whenever power is applied to the unit, the bridge performs its diagnostic self-test. The bridge’s
self-test normally takes about 15 seconds to complete. The bridge then spends about 30 seconds
communicating with other bridges in the network to determine if there are loops and build its ini-

tial database of forwarding addresses.

Allow up to 45 seconds for the bridge’s self-test and communications tasks to complete, then com-

pare the state of the five status LEDs on the bridge with those shown in Figure 2-14.

ON LINE
PORT A
ACTIVITY

SELF-TESTOK
DC OK

PORT B

ACTIVITY

“,

it
ol
ORI
K

i
——

o,
o
'

ON OR OFF OR BLINKING
(INDICATING NETWORK

TRAFFIC)
O

EITHER OF TWO STATES:

ON = DESIGNATED BRIDGE

OFF = BACK-UP BRIDGE

BRI,

If:

Then:

The status LEDs match

The LAN Bridge 100 hardware is

those shown here.

functional.

[

—

[——

The status LEDs DO NOT

Go to Chapter 5.

match those shown here.

LKG-0957

Figure 2-14:

Bridge Hardware Verification
NOTE

For definitions of the status LEDs on the bridge, refer to Chapter 1, Table 1-5.
PRAIORS

2-16

LAN Bridge 100 Technical Manual

—

Check the logical link by sending a message from a node on one side of the bridge to a node on the

other side. On a VAX/VMS host running DECnet, you can send a message by using the SET HOST

command.

Installation of the LAN Bridge 100 unit is now complete.

Installation and Verification

2-17

2.5.2

Verifying LAN Traffic Monitor (LTM) Installation

Whenever power is applied to the unit, the bridge performs its diagnostic self-test and, if success-

ful, initiates a request for a down-line load of the LTM Listener software image from a load host.
The bridge’s self-test normally takes about 15 seconds to complete, but the down-line loading of
the software image could take longer if the network is busy.
Allow up to 2 minutes for the bridge’s self-test and down-line loading of the LTM Listener software
image to complete; then compare the state of the five status LEDs on the unit with those shown in
Figure 2-15.

SELF-TEST OK

ON LINE

DC OK

PORTA
ACTIVITY

PORTB
ACTIVITY

ON OR OFF OR BLINKING
(INDICATING NETWORK

TRAFFIC)

FLASHING TWICE EVERY TWO SECONDS
INDICATES THAT THE LOAD HOST
SUCCESSFULLY DOWN-LINE LOADED THE
LTM LISTENER SOFTWARE IMAGE.
FLASHING ONCE EACH SECOND INDICATES
THAT THE LOAD HOST HAS STARTED

THE LTM LISTENER FUNCTION.
B

If:

Then:

The status LEDs match

The LTM Listener is operational.

those shown here.

Notify the system manager.

The status LEDs DO NOT

Go to Chapter 5.

match those shown here.
LKG-0956
R

Figure 2-15:

LTM Hardware Verification
B—

NOTE
For definitions of the status LEDs on the bridge, refer to Chapter 1, Table 1-5.
Installation of the LAN Bridge 100 unit as a LAN Traffic Monitor is now complete.

2-18

LAN Bridge 100 Technical Manual

3
Operation

3.1

Chapter Overview

This chapter describes the operational states and processes control operation of the LAN Bridge
100 unit. This chapter also describes the initialization block and descriptor ring data structures.

3.2

Operational States

The LAN Bridge 100 unit has six states that control its functions. An operating state transition diagram is shown in Figure 3—1. The LAN Bridge 100 unit has the following operational states:
e

SELF-TEST

BROKEN

INITIALIZATION

e PREFORWARDING
e

FORWARDING

BACKUP

3-1

C POWER UP )

—

A A

CSOFT RESET )

SELF-TEST
(LEDs FLASH

ONCE)

SELF-TEST
pAstD

—

NVRAM
NO

TES’I‘; FAIL

NVRAM

YES

LOOP-ON

YES

SWITCH ON
NO

ON?

SELF-TEST

—

RESET SWITCH >

.
3

BROKEN STATE
(NO LED SET)

.
WAIT ABOUT 15

SECONDS
-

SET SELF-TEST
LED

LKG-0737-87

Figure 3-1

Figure 3-1:

Operating State Transition (Sheet 1 of 3)

3-2

LAN Bridge 100 Technical Manual

(

INITIALIZE

)

OLL* SWITCH
IN NVRAM
SET?

pLL*

" INFORMATION
NVRAM

w:. T0
07

YES

DOWN—LINE
LOAD ENABLE

SWITCH (SW5)
SET?

owN—LUNE LoaD]
WITH FILE
NAMED BY ALL

DLL*INFORMATION|

Jpown—LINE LoaD
WITH DEFAULT
FILE SPECIFIED
BY LOAD HOST

I
(EXEC‘UTE LOADED)
SOFTWARE

* DLL STANDS FOR DOWN-—LINE LOAD
LKG~D737-87

Figure 3-1:

Operation

Operating State Transition (Sheet 2 of 3)

3-3

DETECTION
CONDITIONS

MET
?

YES

OGS

CONDITIONS
MET

LoOP
DETECTION
CONDITIONS
MET

AR,

YES
R

( SOFT RESET )
R

LKG-0737

Figure 3-1:

Operating State Transition (Sheet 3 of 3)

LAN Bridge 100 Technical Manual

3.2.1

SELF-TEST State

The SELF-TEST state is entered on power up, on a soft reset command received from bridge management software, or from the BROKEN state.

The SELF-TEST state sets the LAN Bridge 100 unit to a known condition and tests the hardware and
firmware by performing read/write operations to various hardware locations.
The SELF-TEST state loops (repeats) if the Loop Self-test switch (located on the LAN Bridge 100

connector panel) is set to the on position. A description of the self-test procedure is provided in
Section 3.9.

If the self-test fails (except for nonvolatile RAM [NVRAM] failures), the LAN Bridge 100 unit enters

the BROKEN state.

- If NVRAM fails and the NVRAM Reset switch (located on the I/0 panel) is enabled, the LAN Bridge
100 unit will enter the INITIALIZATION state using default parameters. If NVRAM fails and the
NVRAM Reset switch is disabled, the LAN Bridge 100 unit will enter the BROKEN state.
If the self-test passes, the LAN Bridge 100 unit enters the INITIALIZATION state.

3.2.2

BROKEN State

The BROKEN state is entered if errors occur in self-test. In the BROKEN state, the LAN Bridge 100

unit reenters the SELF-TEST state every 15 seconds until the LAN Bridge 100 unit is powered off.

3.2.3

INITIALIZATION State

The INITIALIZATION state is entered on completion of the self-test.

3.2.3.1

Bridge Initialization — The microprocessor uses the INITIALIZATION state to perform

the following operations:
®

Initialize various system counters and registers.

Allocate necessary buffer space in packet memory.

Write the initialization blocks and the descriptor rings into packet memory (see Section 3.3).

e Write the LAN Bridge 100 unit’s physical address into the Ethernet address table and into the
LANCEs.

Werite the pointers to the initialization block into the LANCE control and status registers.

Determine whether the bridge down-line loads as part of the INITIALIZATION state.

Write the initialization and start bits into the LANCE control and status registers.

If the bridge determines that a down-line load will not occur, the INITIALIZATION state exits to

the PREFORWARDING state after the initialization and start bits are written into the LANCE control and status registers.

Operation

3-5

3.2.3.2

Down-Line Loading — The LAN Bridge 100 unit may perform a down-line load as part

of the INITIALIZATION state.
The down-line load feature is controlled using a hardware ‘‘down-line load enable’’ switch and

two software parameters: DLL-switch and DLL-info (DLL stands for down-line load). The hardware switch must be manually set or cleared. Software parameter values for DLL-switch and
DLL-info are entered using RBMS.

An algorithm examines the hardware switch and software parameters to determine whether the
bridge down-line loads. If the bridge is configured to down-line load, the algorithm specifies which

file to load. A block diagram of the algorithm is contained in Figure 3-1.

MOP protocol messages control the down-line load operation. Once the down-line load operation
begins, the bridge attempts to complete the operation until the load is successful or the bridge
power is turned off.

When the down-line load is complete, the INITIALIZATION state exits to the state determined by

the loaded software.

3.2.4

PREFORWARDING State

The PREFORWARDING

state is entered after the initialize and start bits are written into the LANCE

control and status registers. This state may also be entered from the BACKUP state when the LAN
Bridge 100 unit becomes the root bridge in the extended LAN or a designated bridge on a LAN (see
Section 3.5).

The LAN Bridge 100 unit uses the PREFORWARDING state to:

Learn the location of active stations on the network.

Perform loop detection functions to determine whether it is the root bridge in the extended
LAN or a designated bridge on a LAN (see Section 3.5).

The operations performed in the PREFORWARDING and FORWARDING states (see Section 3.2.5)
are identical except that packets are not forwarded during the PREFORWARDING

state. This pre-

vents the formation of transient loops.
The duration (forwarding delay) of the PREFORWARDING state has a default value of 30 seconds.

The forwarding delay can be changed using RBMS (see the Remote Bridge Management Software
Guide).
The PREFORWARDING state exits to the BACKUP state if the LAN Bridge 100 unit determines that
the following loop detection conditions (see Section 3.5) are met:
®

The LAN Bridge 100 unit is not a root bridge.

The bridge is not a designated bridge on a LAN.

If the

loop

FORWARDING

detection

conditions

are

not

met,

the PREFORWARDING state exits to the

state.

ARSI

3-6

LAN Bridge 100 Technical Manual

3.2.5

FORWARDING State

The FORWARDING state is entered from the PREFORWARDING

state when loop detection condi-

tions are not met (refer to Sections 3.2.4 and 3.5).
The FORWARDING state is similar to the PREFORWARDING state except that packets may be forwarded. In this state, the LAN Bridge 100 unit performs address filtering functions. The LANCE
chips receive and store packets in memory. The packets are discarded or processed depending on a

determination made by the microprocessor and table lookup subsystem.
Other tasks carried out in the FORWARDING state include address table maintenance and loop
detection functions. Bridge management functions such as Remote Bridge Management Software
(RBMS) and maintenance functions which use Maintenance Operation Protocol (MOP) may also be

performed.
The FORWARDING state exits to the SELF-TEST state if a reset operation is performed.

The FORWARDING state exits to the BACKUP state if loop detection conditions are met (refer to
Section 3.5).

3.2.6

BACKUP State

The BACKUP state is entered when loop detection conditions are met (see Section 3.5). Remote

Bridge Management Software (RBMS) commands may be used to force the LAN Bridge 100 unit into

the BACKUP state by ‘‘disabling’’ a link in the bridge.
During the BACKUP state, the LAN Bridge 100 unit ‘‘listens’’ to network traffic. If the LAN Bridge
100 unit determines that loop detection conditions are not met, it enters the PREFORWARDING
state.

3.3

Packet Memory Data Structures

The LANCE initialization blocks and the descriptor rings are the data structures that manage packet
memory.

The LANCE chips use these structures to locate available buffer space for storing incoming packets
and for finding packets that are to be transmitted.

The microprocessor writes the initialization blocks and descriptor rings into packet memory during the initialization state. The microprocessor uses the descriptor rings to find packets in memory

and to identify packets that are ready for transmission by a LANCE.

3.3.1

Initialization Blocks

The microprocessor writes two initialization blocks into packet memory to initialize the two
LANCE chips. Each block uses 12 words of contiguous memory.

Each LANCE uses its own initialization block to configure its operating parameters. The fields in an
initialization block are identified in Figure 3-2.

Operation

3-7

e
—

MODE OF LANCE OPERATION

ETHERNET PHYSICAL ADDRESS
ETHERNET PHYSICAL ADDRESS

ETHERNET PHYSICAL ADDRESS
LOGICAL ADDRESS FILTER
LOGICAL ADDRESS FILTER
LOGICAL ADDRESS FILTER

LOGICAL ADDRESS FILTER

RECEIVE DESCRIPTOR RING POINTER

RECEIVE DESCRIPTOR RING POINTER
TRANSMIT DESCRIPTOR RING POINTER
Y

TRANSMIT DESCRIPTOR RING POINTER
LKG~1224—87

Figure 3-2:

LANCE Initialization Block Field Functions

The functions of each field in the initialization block are defined as follows:

Mode of LANCE operation — Promiscuous mode is selected, signifying that all addresses are

received.

Ethernet physical address — This field contains the 48-bit Ethernet address assigned to the LAN
Bridge 100 unit. The LANCE uses this address to detect packets (such as RBMS or MOP commands) addressed to the LAN Bridge 100 unit. This address is also transmitted by the LAN
Bridge 100 unit in loop detection (Hello) messages.

Logical address filter — The 64-bit filter helps to detect multicast addresses. This filter is not
used by the LAN Bridge 100 unit since the LANCE operates in promiscuous mode and receives
all packets.

Receive descriptor ring pointer — This field points to the base (lowest) address of the receive

YRGS

descriptor ring. Another part of this field defines the number of entries in the receive descriptor

ring (1, 2, 4, 16, 32, 64, or 128 entries).
R

IOR

3-8

LAN Bridge 100 Technical Manual
OSBRI

Transmit descriptor ring pointer — This field points to the base (lowest) address of the transmit
descriptor ring. Another part of this field defines the number of entries in the transmit descriptor ring (1, 2, 4, 16, 32, 64, or 128 entries).

3.3.2

Descriptor Rings

Four descriptor rings (one transmit and one receive ring for each LANCE) are written into packet
memory by the processor during the INITIALIZATION state.

A descriptor ring is a circular queue of tasks which the LANCE and processor use to locate buffer
space or packets in packet memory. Figure 3-3 illustrates the concept of descriptor rings.
Each descriptor ring is made up of entries that define the location and status of buffer space in
packet memory. Each entry is four words in length. Thus each entry starts on a quadword boundary in the descriptor ring address space. The number of entries in a descriptor ring is defined in the
ring pointer of the initialization block.

Receive descriptor rings are used by the LANCE chips to find available buffer space for writing
incoming packets. The processor uses the receive descriptor rings to find the destination and
source addresses in packets the LANCE chips receive (the destination and source addresses are contained in the first 12 bytes of a packet).

Transmit descriptor rings are written to by the processor to identify packets awaiting transmission.
The LANCE reads the transmit descriptor rings to find packets awaiting transmission.

Operation

3-9

RECEIVE DESCRIPTOR RING

LANCE CSRs

POINTER TO INITIALIZATION
BLOCK

ENTRYO
(>

| RECEIVE |

BUFFER ADDRESS

| BUFFER |

STATUS, BUFFER SIZE,

PACKET SIZE

ENTRY 1

| RECEIVE

BUFFER ADDRESS

BUFFER
|

STATUS, BUFFER SIZE,

PACKET SIZE

ENTRY n

INITIALIZATION

RECEIVE

BLOCK

BUFFER

BUFFER ADDRESS

MODE OF OPERATION

STATUS, BUFFER SIZE,
PACKET SIZE

PHYSICAL ADDRESS

n
T,

LOGICAL ADDRESS FILTER

POINTER TO RECEIVE RING
SO

NUMBER OF RECEIVE ENTRIES

POINTER TO TRANSMIT RING
NUMBER OF TRANSMIT ENTRIES

TRANSMIT DESCRIPTOR RING
ENTRYO

{ TRANSMIT}

BUFFER ADDRESS, STATUS
BUFFER SIZE, PACKET SIZE

BUFFER |

ENTRY 1

BUFFER ADDRESS, STATUS

BUFFER SIZE, PACKET SIZE

ENTRY n

BUFFER ADDRESS, STATUS
BUFFER SIZE, PACKET SIZE

TRANSMIT
| BUFFER

\ L

LKG-0739

P—

Figure 3-3:

Descriptor Ring Structures
AR

3-10

LAN Bridge 100 Technical Manual
WG

3.3.2.1 Receive Descriptor Ring Entry — Each receive descriptor ring entry is made up of four
words" that define the location and status of buffer space in packet memory. Figures 3—4 through
3-7 show the fields that make up each word in the receive descriptor ring entry. Tables 3-1
through 3—4 describe the bit functions of each word in the receive descriptor ring entry.

LKG-0740

Figure 3-4:

Bit Format of Word 0 in the Receive Descriptor Ring Entry

Table 3-1:

Bit Descriptions of Word 0 in the Receive Descriptor Ring Entry

Bit(s)

Name

<15:00>

LADR

Description

Low address. The low-order 16 bits of the address of the transmit data
buffer corresponding to this receive descriptor ring entry.

Operation

3-11

OWN

FRAM

OFLO

CRC

BUFF | STP | ENP

LKG-0742

Figure 3-5:

Bit Format of Word 1 in the Receive Descriptor Ring Entry

Table 3-2:

Bit Descriptions of Word 1 in the Receive Descriptor Ring Entry

Bit(s)

Name

Description

<15>

OWN

Ownership bit. When set, indicates that the LANCE owns or has use of the
receive descriptor ring entry and the corresponding receive data buffer.

The LANCE clears the OWN bit to give the entry to the microprocessor.
The microprocessor sets the bit to give the entry back to the LANCE.

<14>

ERR

Error summary. Indicates that one or more of the following error bits is

set: FRAM, OFLO, CRC, or BUFF.
<13>

FRAM

Framing error. Indicates that the received packet has both a CRC error and
that the number of bits in the packet is not a whole multiple of 8.

<12>

OFLO

Overflow. Indicates that the LANCE has lost part or all of the packet
because the internal data silo overflowed before the LANCE could write
the packet into packet memory.

<1l1>

CRC

CRC error. Indicates that the packet has a CRC error.

<10>

BUFF

Buffer error. Indicates that the LANCE filled the receive data buffer and
was unable to chain to another receive data buffer to store the remainder

of the packet. Either there was not another receive buffer available or the
LANCE was unable to read the receive descriptor ring entry quickly
enough to find the new buffer before the LANCE internal data silo over-

flowed. Data buffers must be at least 64 bytes long to allow the LANCE
time to find the next data buffer for buffer chaining.
<09>

STP

Start of packet. Indicates that this receive data buffer starts a new packet.

<08>

ENP

End of packet. Indicates that this receive data buffer ends a packet.

<07:00>

HADR

High address. The high-order 8 bits of the address of the receive data
buffer corresponding to this receive descriptor ring entry.
A

3-12

LAN Bridge 100 Technical Manual
M,

LKG-0743

Figure 3-6:

Bit Format of Word 2 in the Receive Descriptor Ring Entry

Table 3-3:

Bit Descriptions of Word 2 in the Receive Descriptor Ring Entry

Bit(s)

Name

Description

<15:12>

Must be ones.

<11:00>

BCNT

Buffer byte count. The length of the data buffer corresponding to this
entry in the receive descriptor ring expressed as a twos complement.

)

]

RESERVED

]

PCNT

)

LKG-0744

Figure 3-7:

Bit Format of Word 3 in the Receive Descriptor Ring Entry

Table 3-4:

Bit Descriptions of Word 3 in the Receive Descriptor Ring Entry

Bit(s)

Name

Description

<15:12>

RESERVED

Not used.

<11:00>

PCNT

Packet byte count. The number of bytes in the received packet.

Operation

3-13

3.3.2.2

Transmit Descriptor Ring Entry — Each transmit descriptor ring entry is made up of

four words that define the location and status of packets stored in packet memory. Figures 3-8
through

3-11

show the fields that make up each word in the transmit descriptor ring entry.

Tables 3-5 through 3-8 describe the bit functions of each word in the transmit descriptor ring
entry.

LKG-0745

Figure 3-8:

Bit Format of Word 0 in the Transmit Descriptor Ring Entry

Table 3-5:

Bit Descriptions of Word 0 in the Transmit Descriptor Ring Entry

Bit(s)

Name

Description

<15:00>

LADR

Low address. The low-order 16 bits of the address of the transmit data

RIS

buffer corresponding to this transmit descriptor ring entry.

3-14

LAN Bridge 100 Technical Manual

OWN

ERR

RES

ONE

DEF | STP | ENP

LKG-0746

Figure 3-9:

Bit Format of Word 1 in the Transmit Descriptor Ring Entry

Table 3-6:

Bit Descriptions of Word 1 in the Transmit Descriptor Ring Entry

Bit(s)

Name

<15>

OWN

Description
Ownership bit. When set, indicates that the LANCE owns or has use of the

transmit descriptor ring entry and the corresponding transmit data buffer.
The microprocessor sets the OWN bit to tell the LANCE to transmit the

contents of the data buffer as a packet. The LANCE clears the bit to give the
entry back to the microprocessor.

<14>

ERR

Error summary. Indicates that one or more of the following error bits is set
in the transmit descriptor ring entry word 3: UFLO, LCOL, LCAR, or
RTRY.

Not used.

<13>
<12>

More. Indicates that more than one retry was needed to transmit the
packet.

<1l1>

ONE

<10>

DEF

One. Indicates that exactly one retry was needed to transmit the packet.
Deferred. Indicates that the LANCE had to defer or wait for a pause in the
message traffic on the network before transmitting the packet.

<09>

STP

Start of packet. Indicates that this is the first buffer to be used by the
LANCE for this packet. It is used for data chaining buffers.

<08>

ENP

End of packet. Indicates that this is the last buffer to be used by the LANCE
for this packet. It is used for data chaining buffers.

<07:00>

HADR

High address. The high-order 8 bits of the address of the transmit data
buffer corresponding to this transmit descriptor ring entry.

Operation

3-15

P ——

LKG-0747

Figure 3-10:

Bit Format of Word 2 in the Transmit Descriptor Ring Entry

Table 3-7:

Bit Descriptions of Word 2 in the Transmit Descriptor Ring Entry

Bit(s)

Name

Description

<15:12>

Must be ones.

<11:00>

BCNT

Buffer byte count. The length of the data buffer corresponding to this

entry in the transmit descriptor ring expressed as a twos complement
number.

3-16

LAN Bridge 100 Technical Manual

—

TDR

LKG-0748

Figure 3-11:

Bit Format of Word 3 in the Transmit Descriptor Ring Entry

Table 3-8:

Bit Descriptions of Word 3 in the Transmit Descriptor Ring Entry

Bit(s)

Name

Description

<15>

BUFF

Buffer error. Indicates that the LANCE did not find the end of packet (ENP)

<14>

UFLO

Underflow error. Indicates that the LANCE has truncated the packet

_—

<13>

RES

Reserved. Not used.

<12>

LCOL

Late collision. Indicates that a collision occurred after the LANCE trans-

bit in the current buffer and did not own the next buffer.

because it was unable to read data words from memory quickly enough.

mitted the first 64 bytes of the packet. All nodes should have detected that

there was activity on the network by this time. This suggests that the colli-

sion detect circuitry of some other node on the network has failed. The

LANCE does not retry on late collisions.

<11>

LCAR

Loss of carrier. Indicates that the carrier-presence input to the LANCE

went false during the transmission of this packet. The LANCE does not
retry upon loss of carrier.

<10>

RTRY

Retry error. Indicates that the LANCE failed to transmit the packet in 16

<09:00>

TDR

Time domain reflectometry. An internal counter that counts system
clocks (10 MHz) from the start of a transmission to the occurrence of a col-

attempts because of repeated collisions on the network.

lision. This value is useful in determining the approximate distance to a

fault in the network cable. The TDR value is valid only if RTRY is set.

Operation

3-17

3.4

Learning and Forwarding

USROS

The main function of the LAN Bridge 100 unit is filtering network traffic. This minimizes network

traffic by keeping local traffic local.

The LAN Bridge 100 unit “‘listens’’ to network traffic and acquires a working knowledge of which
stations are associated with which LAN (LANSs are identified as being connected to port A or port B

of the bridge). This knowledge is acquired by reading the source addresses of incoming packets and
noting the port through which the packet entered. Thus each station address that is heard from

becomes associated with a port of the bridge.
The LAN Bridge 100 unit stores the information in a station address table that holds the station

address (48 bits) and the status of each station (16 bits). The status is written by the firmware and
defines the port associated with an address, the age of the address, and so on.

When subsequent packets are received that are intended for a known (previously stored) address,
the bridge can determine whether the packet should be forwarded or ignored. The bridge also
updates any existing status for the station that sent the packet. If the sending station was previously

unknown, the bridge adds the new address and its status to the table.
Figure 3—12 illustrates the table lookup (TLU) process. The Troll feature shown in the figure may be

set by RBMS. This feature controls access to certain stations on the network (see Section 3.4.5).

T—

3-18

LAN Bridge 100 Technical Manual
ST

ARRIVES

COMPARE DESTINATION

ADDRESS AGAINST TABLE

MATCH

{

YES

ANY
.

REASON NOT TO

YES

FORWARD PACKET

TROLL FEATURE

YES

SET
' NO

PORT

DESTINATION

= ORIGINATION

YES

NO fi

PORT

DON'T FORWARD PACKET l

FORWARD PACKET

—

LKG-0749

Figure 3-12:

Operation

Table Lookup Process (Sheet 1 of 2)

3-19

COMPARE SOURCE ADDRESS
AGAINST TABLE

MATCH

NO
e

YES
|

INSERT NEW ADDRESS

—

IN TABLE AND WRITE

I UPDATE STATUS WORD

]

STATUS WORD

B—

LKG-0749
A

Figure 3-12:

3.4.1

Table Lookup Process (Sheet 2 of 2)

Table Initialization

When the LAN Bridge 100 unit first becomes active, the network address table is empty. Because

the table is empty, the processor has no means of accessing status information to perform packet
filtering. The bridge must construct a list of active stations on the network. To do this, the bridge
processor places the bridge in the PREFORWARDING state to monitor all traffic on the two LANs

to which it is connected.

In the PREFORWARDING state, source addresses are extracted from the packets received from the
LANSs, but the packets are not forwarded. These source addresses and a status word for each station
address are inserted in ascending order into the network address table.

The bridge exits the PREFORWARDING

state after the preforwarding delay time of 30 seconds.

The preforwarding delay time can be changed by use of RBMS.

3-20

R—

LAN Bridge 100 Technical Manual
R

3.4.2

Address Table Entries

Address table entries are used to keep track of the status of known network stations. A station

becomes known when the LAN Bridge 100 unit receives a packet from the station.

An address table entry has the fields shown in Figure 3-13.

48 47
STATUS <63:48>

STATION ADDRESS <47:00>

LKG-0750

Figure 3-13:

Address Table Entry

The status is a 16-bit word stored in address table RAM. The status functions include:

Identifying the port associated with the station address.

Logging the age of the last packet received.

Noting any special status that may be set by RBMS such as a nonaging or limited access address
and so on.

The station address is the unique 48-bit station address of the originating station.

3.4.3

Binary Search

Under the heaviest network traffic conditions, the bridge must be able to filter or forward a contin-

uous stream of minimume-size (64 bytes) incoming packets. Thus it is essential that the TLU subsystem quickly find entries in the active station list or determine that an entry is not present. The

active station list is an ordered list (ascending numerical order) of 48-bit station addresses corresponding to active stations on the network.

A rapid search is accomplished using a binary decision tree search method. Figure 3-14 shows a
binary decision tree for 16 stations.
The binary search is started at the middle entry of the table. A comparison is made between the

address received and the middle entry and a ‘‘greater than,”’ “‘less than,’’ or “‘equal to’’ determination is made. This determination results in the following actions:

Ifthereceived address is greater than the accessed entry of the table, the search continues at the

center of the upper half of the tree. In Figure 3—-14, the search would continue at address 12.

If the received address is less than the accessed entry of the table, the search continues at the
center of the lower half of the tree. In Figure 3-14, the search would continue at address 4.

If the received address is equal to the accessed entry of the table, the search ends.

Operation

3-21

If the search continues, another ‘‘greater than,”” ‘‘less than,”’ or ‘‘equal to’’ determination is made
on the remaining part of the tree. This process continues until the address is found or is determined
to be missing from the table.

When the search ends, the results are made available to the search results register. The results
include a pointer to the last location probed, and bits that indicate whether the received address

was less than, greater than, or equal to the address in the last location probed.
The binary search algorithm requires 2 maximum of 13 probes in a table containing 8000 entries.
The example in Figure 3—14 requires a maximum of four probes. To optimize the LAN Bridge 100
processor performance, the results of a search are sought only after enough time has elapsed to
guarantee that the required number of probes has been performed.
Figure 3-15

illustrates the decisions that control the binary search process.

R —

<«—— LESS THAN mlm GREATER THAN —>
-

S\
/\

<«— SECOND PROBE

/\ /\ /\ /\ /\ /\ /\ /\
11

FIRST PROBE

THIRD PROBE

FOURTH PROBE

THIS ROW IDENTIFIES THE LOCATION WHERE
NEW ADDRESS SHOULD RESIDE IF IT IS NOT

FOUND IN THE TABLE.
S

LKG-0751

Figure 3-14:

Binary Decision Tree
B

3-22

LAN Bridge 100 Technical Manual

—

(

START

)

:I“:.".':

ACCESS MIDPOINT

OF (SUB-) TABLE

TABLE EXHAUSTED

—

COMPARE ADDRESS

FROMTABLE AGAINST

SEARCH ARGUMENT

C ENTRY NOT FOUND)

TABLE ADDRESS

GREATER THAN SEARCH

ARGU?ME.NT

TABLE ADDRESS

PROBE HIGHER HALF
OF TABLE

'EQUQ'JSMSEE,:‘?CH
2

YES

\
OF TABLE

ADDRESS FOUND

LKG-0752

Figure 3-15:

3.4.4

Binary Search Process

Forwarding

When the LAN Bridge 100 unit has found an incoming packet’s destination address to be stored in

whether the packet should be forwarded or ignored.

When the bridge has not found the destination address from an incoming packet to be stored in the

active station list of the address table, the bridge forwards the packet by default.

the active station list of the address table, it checks the corresponding status word to determine

Operation

323

NOTE
If certain conditions are set by use of RBMS, the LAN Bridge 100 unit places additional constraints on the decision of whether to filter or forward packets. Some of

these conditions are discussed in Section 3.4.5.

3.4.5

Writing the Ethernet Address Table

An optional software package called Remote Bridge Management Software (RBMS) can be used to
specify additional conditions for the bridge’s decision on whether to filter or forward packets.
Users of RBMS can write into the active station list to regionalize multicast packets or limit the

SRS

access of certain stations on opposite sides of the bridge.
More detailed information on implementing the following features is provided in the Remote
Bridge Management Software Guide (part no. AA-FY93A-TE).

Regionalization
A number of protocols use multicast packets for initializing, locating other stations, and so on. It is
easy to confine such protocols to a particular region of the network by instructing the bridge not to

pass such multicast packets. This is done by entering the multicast addresses into the Ethernet

address table of the bridge. In this way the bridge recognizes the addresses as being on the same side

of the bridge as the originating station and does not forward them.

Limiting Access
It may be desirable from a network management standpoint to limit the ability of some stations on

one side of the bridge to send packets to certain station on the opposite side of the bridge. This is
referred to as the Troll feature, which can be activated by RBMS.
)

The appropriate bridge management protocol packet instructs the bridge to flag none, some, or all
of the addresses in the active station list as controlled addresses. This causes the bridge firmware to
drop packets originating from one side of the bridge destined for these controlled addresses on the

opposite side. In other words, the bridge never forwards a packet to what is considered to be a con-

trolled address.
P

3.4.6

Special Cases

The LAN Bridge 100 unit handles the following special cases.

3.4.6.1

Stale Packets — A packet that is to be forwarded becomes stale if the bridge is not able to

forward the packet within two seconds because of outbound congestion. This stale packet time is
measured from the reception of the last byte of the CRC off the inbound local area network (LAN)

to the start of transmission of the destination address on the outbound LAN. The LAN Bridge 100
unit recovers from a stale packet condition by causing an unsuccessful transmission of the packet.

3.4.6.2

Swapped Sides — If a station is moved from one side of the bridge to the other, its pres-

ence on the new port supersedes its presence on the old port (the status word is updated).

3-24

LAN Bridge 100 Technical Manual

3.4.6.3 Address Aging — The relative age of station addresses in the Ethernet address table is

determined by the amount of time elapsed since the station was last heard from.
If two minutes have elapsed since the station was last heard from, the bridge considers the station

to have ceased to exist on that side of the bridge. Any data in the status word indicating that the station exists on that side of the bridge is ignored.

If that station later transmits from its previous side of the bridge, its address is relearned.
NOTE

Two minutes is the normal amount of time allowed for aging. However, for two
minutes following a topology change, only 30 seconds of aging are allowed. Topology changes can change the location (port) associated with an address and cause
partitioning of the extended LAN. Shortened aging time minimizes partitioning by
clearing the address table of entries whose locations were affected by the topology
change.

3.4.7

Ethernet Address Table Maintenance

When the LAN Bridge 100 unit finds the source address from an incoming packet to be in the active
station list of the Ethernet address table, it updates the corresponding status word. Updating the
status word includes resetting the aging timer, noting the port associated with the address and so
on.

When the source address from an incoming packet is not stored in the active station list, the bridge
must insert the new address (and status) into its place in the ordered table.

New addresses and their search status (last location probed and the equal to or greater than bits) are
first stored in the new address entry portion of the address table.

As a background operation (see Section 3.6), the new address entries are presorted in the new entry
portion of the table. Then each address is inserted into the active station list.
3.4.7.1 Inserting New Addresses — When a new source address is inserted into the active station list, it must be inserted in its proper location in the ordered table. Often, this requires
numerous existing entries to be shifted up so that the proper location is ‘‘vacated’’ to accommodate the new address.

As an example, Figure 3—16 shows addresses that must be moved to accommodate the new address
55. Address 55 must be inserted into location N to maintain the ordered nature of the table.
Since the new address 55 must be inserted into location N in the active station list, seven addresses
must be transferred up one location at a time until location N is vacated.

Operation

3-25

LOCATION

ACTIVE STATIONS
48 BITS

STATUS
16 BITS

A
B

c
D
E
F

H CC:

STATUS 09

}

STATUS 11

STATUS 12

STATUS 17

STATUS 20

STATUS 40

INSERT 55— N —]

STATUS 42

ACTIVE

STATUS 60

STATION

STATUS 62

LIST

STATUS 71

STATUS 82

STATUS 96

STATUS 56

STATUSS6

LAST

STATUS 61

STATUS 67

NEW ADDRESS

ENTRIES

LKG-0753
U

Figure 3-16:

Inserting a New Address into the Ordered Table

3.4.7.2

Single-Entry Compare — When several new address
es have accumulated in the new
address entry portion of the address table, they are presort
ed using a single-entry comparison
before they are inserted into the active station list. Presort
ing new entries quickens the process of
insertion into the active station list.

RIS

Under normal conditions, a binary search is initiated using
the middle location in the active station
list. A binary search continues until the address is found
or the table is exhausted.

3-26

LAN Bridge 100 Technical Manual

A special location (0000) in the address table is used to initiate single-entry comparisons. When

location 0000 is probed by the microprocessor, a single comparison is made and no further locations are probed.

Once the new entries are presorted, the microprocessor then uses 64-bit moves (see Section
3.4.7.3) to insert each address into its location in the active station list.

3.4.7.3 64-Bit Move — Shifting any address in the address table involves moving 48 bits of
address and the corresponding 16 bits of status, a total of 64 bits. Since the microprocessor can
manipulate only 16 bits of data at a time, each address shift would require four address-to-address
transfers.

To accelerate the task of table maintenance, a special address range is mapped such that when the

microprocessor addresses the special address range, all 64 bits (the address and status) are moved at
once. The low-order 48 bits that make up the Ethernet address are read from RAM, temporarily

stored in the comparator PALs and then written into the new location in RAM. The high-order 16
bits that make up the status word are written from the status RAM, temporarily stored in the microprocessor status register, and then written into the new location in the status RAM.

3.5

Loop Detection

LAN configurations cannot include any loops (redundant paths) because packets transmitted onto
the LAN will circulate around any loop indefinitely.

All LAN Bridge 100 units have a loop detection feature. This is an automatic process that logically
configures a loop-free extended network. This process, performed by a spanning tree computation

algorithm, provides the following features to enhance extended network performance:

Loop detection — If LAN Bridge 100 units are accidentally configured in a loop, the process
will compute a loop-free portion of the topology. This prevents packets from circulating
around the extended network indefinitely.

Backup — LAN Bridge 100 units can be deliberately configured in a loop in such a way that one
LAN Bridge 100 unit in the loop can serve as the backup for any one of the other LAN Bridge
100 units in the loop.

The process is self-maintaining — The process is a continuous one that responds dynamically to
changes in network topology. Changes may be caused by a malfunctioning LAN Bridge 100 unit

or by using Remote Bridge Management Software to change parameters in the spanning tree
computation algorithm. When changes are sensed, the algorithm recomputes the new network
topology and again becomes stable and self-maintaining.
e

The process is deterministic — A fixed set of rules controls the process so that when variables
are changed, the results can be predicted.

The process requires low network overhead — The messages that control the spanning tree are

transmitted by participating LAN Bridge 100 units at 1-second intervals, thus using a very small
percentage of the available network bandwidth. For example, if the interval between Hello
messages is 1 second (the default), the overhead caused by Hello messages in a stable configuration is 64 bytes per second (0.005% on the network).

Operation

3-27

NOTE
Hello messages

are 60 bytes in length. The Hello multicast address is
(HEX) 09-00-2B-01-00-01, and the type field is 80-38. Because of the propri-

etary nature of the Hello message, other fields are not discussed in this manual.

3.5.1

The Spanning Tree Principle

Although the LAN Bridge 100 units in an extended network can be physically placed in an arbitrary

fashion, the logical network that the LAN Bridge 100 units automatically create is always a

spanning tree. The spanning tree has the following basic properties:
®

There are no loops; that is, there is only one path between any two LAN Bridge 100 units.
NI

All LANSs are connected.

The process by which the LAN Bridge 100 units construct the logical spanning tree from an arbitrary physical configuration is called the spanning tree computation process.

3.5.2

The Spanning Tree Computation Process

HGONGRTE

The spanning tree computation process consists of the following steps:
S,

The extended network elects a single root bridge.

The extended network elects a designated bridge on each of its LANs. The designated bridge is
the LAN Bridge 100 unit with the lowest path cost to the root bridge. A path cost is the sum of

all the line-cost parameter values in any given path of lines. RBMS may be used to modify line
cost parameter values (see Section 3.7).

All LAN Bridge 100 units except designated bridges turn off all lines, except for the single line

that is the lowest path cost to the root bridge. Designated bridges turn off all lines except for the
single line that is the lowest path cost to the root bridge and any lines attached to LANs for

which they are the designated bridge.

When LAN Bridge 100 units are first activated (either by being powered on or by being reset using a
bridge management command), each LAN Bridge 100 unit assumes that it is the root bridge of the
spanning tree. Each LAN Bridge 100 unit sends out Hello messages on all its LANs declaring itself to

be the root bridge of the extended network. These Hello messages are multicast to all other bridges
that are directly connected to the same LANs. For example, in Figure 3—17, bridge B10’s Hello message is received by bridges B3, B4, B9, B11, and B14.

ARG

Because the LAN Bridge 100 unit is declaring itself the root bridge, it also declares itself the desig-

nated bridge on all its LANSs.
RN

As LAN Bridge 100 units receive Hello messages from other LAN Bridge 100 units, each bridge com-

pares the root bridge and designated bridge information in the received Hello messages to its own

information. When performing the comparison, the bridge compares the new root bridge, root
path cost, and designated bridge to the current root bridge, root path cost, and designated bridge.

The root bridge and designated bridge fields are simply the network address of the root bridge and
designated bridge with a root priority prefix. This root priority prefix can be set by bridge manage-

ment and defaults to a fixed value that is equal for all bridges. If a bridge finds that a Hello message

3-28

LAN Bridge 100 Technical Manual
S

contains better root information (lower values), it ceases to declare itself as the root, stores the new

root information just received, and begins to send out the new information.
In a stable environment (all bridges and lines remain active), this process eventually leads to a single
root bridge being elected as well as a single designated bridge per LAN. If no further events occur,

the network will remain configured in the original spanning tree created by the spanning tree computation algorithm. Note that given the same bridges, the same set of conditions, and the same set

of parameters, the spanning tree computation algorithm will always derive the same configuration.
Figure 3—-17 shows the end result of the algorithm for a 14-bridge extended network. In this net-

work, the bridges have the addresses B1 through B14. All root priority fields are assumed to be the
default (128). If address B1 is assumed to be arithmetically less than address B2, then B1 will

become the root bridge. Figure 3—17 also shows the lines that are put into the BACKUP state. In this
state the bridges continue to receive Hello messages over these lines; however, the lines are not
used for forwarding. This BACKUP state is what allows the bridges to form a loop-free tree from an

arbitrary physical connection and yet have the flexibility to reconfigure themselves if a line goes
down. Figure 3—-17 also shows the designated bridges that result from the algorithm: bridges B2,

B3, B5, B6, B8, B10, B12, B13, and B14.

| B12 |
it I
8

| B13 |
&

| B14
|
.|

e
B Lt |
10

¢ Broken lines indicate lines that are in the BACKUP state.
e Bridge B1is the root bridge.

e Bridges B2, B3, B5, B6, B8, B10, B12, B13, B14 are the designated bridges.
LKG-0754

Figure 3-17:

Operation

Result of the Spanning Tree Computation Algorithm

3-29

When a stable spanning tree configuration has been established and a bridge determines that it is
not the root bridge, it stops sending out Hello messages. Instead, it sends out a Hello message only

when it receives a Hello message from the designated bridge on the line to the root bridge. Other
than topology change notification messages, bridges do not send a Hello message on the line over

which they received the Hello message (the line to the designated bridge).
Therefore, in a stable network, each LAN has only one bridge (the designated bridge) that sends out

Hello messages. Figure 3—18 shows the Hello message propagation in the stable network shown in
Figure 3-17.

I)-——-?

""‘T"‘“

LANO

RO,

! and | indicate direction of Hello message propagation.
Broken lines indicate lines that are in the BACKUP state.
® Bridge B1 is the root bridge.
Bridges B2, B3, BS, B6, B8, B10, B12, B13, B14 are the designated bridges.
LKG-0755

Figure 3-18:

3.5.3

Hello Message Propagation

Spanning Tree Parameters

Certain parameters used in the spanning tree computation algorithm can be set using bridge management commands, and almost all parameters and variables involved in the algorithm can be read
using management commands. This capability allows network managers to monitor and influence

decisions made by individual bridges. For information on using this capability, refer to the Remote
Bridge Management Software Guide (part no. AA-FY93A-TE).

3-30

LAN Bridge 100 Technical Manual
O

3.5.4

Examples of the §Spanning Tree Algorithm

Figure 3—19 represents a possible physical network in which seven networks are tied together with
bridges to form a large extended LAN. This particular configuration may not be practical, but it
serves to explain the function of the loop detection process. When the bridges in the network of
Figure 3—19 are powered up, the following events occur as part of the spanning tree algorithm.
1. All bridges send out Hello messages which contain the bridge ID.

2. Since bridge 0 has the lowest ID, it is recognized as the root of the spanning tree.

3. The root bridge establishes itself as the designated bridge on each of the LANS it connects.

4. Each of the other bridges determines its distance from the root bridge in terms of the number of
bridges between it and the root bridge.

5. The algorithm elects a designated bridge on each LAN. The designated bridge is the one with
the lowest path cost to the root bridge.
Designated Bridge

LAN

6. In the case of LAN G, the distance to the root bridge is the same through bridge 5 or bridge 6.
Bridge 5 becomes the designated bridge because its ID is lower than that of bridge 6. Bridge 6
goes into the BACKUP state because it is neither a root bridge nor a designated bridge.

Operation

3-31

“

‘

(/ a

)

§<
N

!
. 0|

N
LKG-0756

Figure 3-19:

3-32

Hypothetical Extended LAN Using Bridges

LAN Bridge 100 Technical Manual

The logical network formed as a result of the learning algorithm is shown in FIgure 3-20.
G

BACKUP

\
O

a~_

ROOT BRIDGE

‘

(L9
N

e
LKG-0757

Figure 3-20:

Operation

Logical Extended LAN Formed by Learning Algorithm

3-33

If a bridge goes from FORWARDING state to a BROKEN state,
the spanning tree algorithm
reconfigures the network topology. For example, if bridge 4 malfuncti
ons, bridge 6 goes to the

FORWARDING state and becomes the designated bridge for LAN E. Bridge
0 is still the root bridge.
The new logical network topology is shown in Figure 3-21.

%B
1
D

1S
&

ROOT BRIDGE

N
Figure 3-21:

LKG-0758

Reconfigured Logical Extended Network

S
—

If the root bridge of Figure 3—-21 malfunctions:
®

Bridge 1 becomes both the root bridge and designated bridge for LANs A and

Bridge 6 goes into the FORWARDING state and becomes the designated bridge for

Bridge 4 becomes the designated bridge for LAN C.

Bridge 2 becomes the designated bridge for LAN B.

LANE.

3-34

LAN Bridge 100 Technical Manual
AT

The new logical network topology is shown in Figure 3-22.

NEW

- ROOT BRIDGE

i<
A

BROKEN

B
LKG-0759

Figure 3-22:

Operation

Reconfigured Logical LAN with a New Root Bridge

3-35

Another example of an extended network is shown

in Figure 3-23. Notice that bridges are connected in such a way that two physical loops exist. The
loop detection function of the bridge helps
to configure the network logically so that no loops exist.

S—

‘

'
O

LKG-0760

Figure 3-23:

Extended LAN Showing Physical Loops

After the network is configured by the spanning tree algorith

Bridge 0 is established as the root bridge of the spanning tree

LANs A and B.
®

Bridge 1 is established as the designated bridge for LAN C.

Bridge 3 is established as the designated bridge for LAN

Bridges 2 and 4 enter the BACKUP state.

and is the designated bridge for

The logical configuration is shown in Figure 3-24.

3-36

LAN Bridge 100 Technical Manual
AR

|
L

| ROOTBRIDGE

BACKUP

E:3::]

BACKUP

LKG-0761

Figure 3-24:

Logical Extended LAN with Backup Bridges

In the network of Figure 3—-24:

If bridge 1 malfunctions, bridge 4 will bcmme operational and the designated bridge for LAN C.
If bridge 3 malfunctions, bridge 2 will become the designated bridge for LAN D.

If bridge O (the root bridge) malfunctions, bridge 1 will become the new root bridge and designated bridge for LANs B and C, bridge 2 will become operational and the designated bridge for
LAN D, and bridge 3 will become the designated bridge for LAN A.

Operation

3-37

[

The new logical network topology formed as a result of the root

Figure 3-25.

bridge malfunction is shown in

BROKEN

)

o g
BACKUP

NEW ROOT |

*
o

C
BRI

(4
LKG-0762

Figure 3-25:

3.6

Logical Network Topology for New Root Bridge

Foreground and Background Operations

The main function of the LAN Bridge 100 unit is filtering and forwardi
ciently under the heaviest network traffic conditions, the LAN Bridge

or forward a packet every 32 ps.

ng packets. To operate effi100 unit must be able to filter
|

During normal operation, the LAN Bridge 100 unit performs a variety of operations that

includes:

Performing loop detection functions.

Receiving packets and writing them into packet memory.

Using the descriptor rings to allocate buffer space and to find packets in packet memory.

Transmitting packets from packet memory.

3-38

LAN Bridge 100 Technical Manual
Y

Performing a binary search of the addresses in the Ethernet address table.

Presorting new address entries.

Storing new address entries in the Ethernet address table.

e Responding to bridge management instfuctions,
Performing MOP operations.

Though all of the operations listed here support the main function of the LAN Bridge 100 unit,
some operations may be less critical; thus their execution may be delayed until an operation of
higher priority is completed.

To prioritize the various operations, those in the critical path of filtering or forwarding packets are
considered foreground operations. Less critical operations are considered background operations.
3.6.1

Foreground Operations

Foreground operations are those operations that must occur so that the LAN Bridge 100 unit can filter or forward packets. The LAN Bridge 100 unit must:
®

Receive packets from a LAN.

Scan and write the receive descriptor ring.

Write the packets into packet memory.

Compare the destination address to the Ethernet address table.

Decide whether to forward or filter the packets.

Scan and write the transmit descriptor ring.

Transmit (forward) the packets to the other LAN.

Compare the source address to the Ethernet address table.

Perform loop detection functions.

Most of the foreground operations are in the critical path of filtering or forwarding packets and are
called critical path operations.

Loop detection functions, though not in the critical path, have the highest priority and are performed as foreground operations. Loops in an extended network cause unnecessary packet traffic
that can degrade performance of critical path operations. This condition can ultimately prevent
background operations from executing, since background operations are deferred until foreground operations are completed.

Loop detection functions maintain LAN Bridge 100 unit performance levels by detecting and eliminating loops. This optimizes usage of the critical path by minimizing unnecessary traffic, thus

allowing time for execution of background operations.

Operation

3-39

3.6.2

Background Operations

Background operations support the main function of the LAN Bridge

100 unit but are not in the
critical path of filtering or forwarding packets. Table 3-9 identifies major
background operations
in order of their priority.

Table 3-9:

Background Operations

Operation

Description

Address table

Typical address table maintenance operations include:

maintenance

Bridge

Updating the address table status word

®*

Presorting new address entries

Storing new addresses in the address table

Responding to bridge management instructions.

management

operations

Maintenance

Operation

MOP functions include responding to requests for system identification, loop

data

messages, or counter messages.

Protocol (MOP)

functions
Timer functions

Counter
maintenance

functions

Timer functions include resetting various timers.
Counter functions include maintaining pointers for various queues such as the

mit and receive descriptor rings.

trans-

Control of state

changes

State changes are dictated by timers and counters that are controlled in the

ground.

back-

3-40

LAN Bridge 100 Technical Manual

3.7

Remote Software and Bridge Access

Remote software such as Remote Bridge Management Software and LAN Traffic Monitor software
use the network to communicate with the bridge. Bridge switches 3, 4, and 5 control communications between the bridge and the node running remote software. Table 3—10 describes the switch
positions that control bridge access.

Table 3-10:

Bridge Access Switches

Switch
Number

Name

ON (Down)

OFF (Up)

Port A

Stations on the LAN connected to
port A that have bridge management capabilities are allowed to
read and write (modify) bridge

Stations on the LAN connected to
port A that have bridge management
capabilities can read but cannot write

Access

bridge management parameters.

management parameters.

If 2 1oad host resides on port A, the
switch must be on for down-line
loading software.

Port B
Access

Stations

the

LAN connected to

Stations on the LAN connected to
port B that have bridge management capabilities are allowed to
read and write (modify) bridge

port B that have bridge management
capabilities are allowed to read but
cannot write bridge management

management parameters.

parameters.

If a load host resides on port B, the
switch must be on for down-line
loading software.
5

Down-Line
Load
Enable

Configures unit to operate as a

LAN Traffic Monitor. Enables the
unit to down-line load software
(such as LTM Listener software)
from a load host. When the switch
is on, the bridge does not forward

Configures unit to operate as a bridge.
Disables the down-line function.
RBMS can override the switch causing
the bridge to operate as a LAN Traffic
Monitor.

packets.

Applicable Port A Access or Port B
Access switch must be on so that
the load host can write to bridge
memory.

Operation

3-41

)

3.8

Maintenance Operation Protocol (MOP)

The LAN Bridge 100 unit implements network interconnect (NI) maintena

nce protocols as specified in the Digital Network Architecture Maintenance Operations Function
al Specification (part
no. AA-X436A-TK). The operational description and message formats of
these protocols are specified in that document and are not repeated here. The minimum functiona
l requirements for an
NI port are outlined in the Digital Network Architecture Ethernet Node
Product Architecture
Specification (part no. AA-X440A-TK).

The bridge supports the following minimum set of maintenance functions:
*

Request identification and system identification packets

®*

Loop data and loop data packets

Request counters and counters messages

These maintenance functions are further defined in Table 3-11.

Table 3-11:

S0,

MOP Request and Response Messages
e

Request Message

Response Message

Request Idenfication — A message from network
management

requesting

system

System Identification — A response

identification

message from the LAN Bridge 100

Loop Data — A message from network management instructing the LAN Bridge 100 to loop data.

from

the

LAN

bridge containing:

®*

MOP version number

Console state

Hardware address

Functions

Device

Looped
~

Data

—

message

from

Bridge 100 in response to a loop data message.

Loop messages are forwarded in the direction indi-

cated by the Ethernet address table.

3.9

Self-test

The self-test is resident in the LAN Bridge 100 firmware and is started on
power up or soft reset.
The self-test program is designed to:
®

Test 95% of all stuck-at (SA1 or SA0) faults.

Provide a maximum amount of fault isolation to the failing functional

block.

The self-test is run in one of two switch-selectable modes:
®

Normal
A

Loop-on-self-test

3-42

LAN Bridge 100 Technical Manual

In normal mode, the self-test starts from the power-up vector and begins testing at the lowest level
with the CPU and ends testing at the more functional network tests. If no errors are found, self-test
illuminates the Self-test OK status LED and jumps to the LAN Bridge 100 firmware. If an error is
found, a code corresponding to the failing test is written to NVRAM, and the Self-test OK status LED
remains off. The LAN Bridge 100 unit enters the BROKEN state, waits 15 seconds, and then

reenters the SELF-TEST state. If an error is again detected, the error is handled in the same manner
as before. The error code previously stored in NVRAM is overwritten with the most recent error
code.

When the loop-on-self-test switch is turned on, the self-test operates the same as it does in normal
mode except that it repeats the entire self-test until either an error is detected or the LAN Bridge

100 unit is powered off. The NVRAM write test is disabled after the first pass to avoid exhausting its
write capability. The Self-test OK status LED is turned on at the end of the first pass and remains on
unless an error is found. If an error is found, the Self-test OK status LED is turned off, and a code
corresponding to the failing test is written to the NVRAM and the manufacturing register. The selftest then loops on the failing test and generates a trigger pulse (once each pass) at one of the external pins. This trigger pulse is an aid to troubleshooting the LAN Bridge 100 module.
The self-test contains three main modules:
1.

Basic tests

LANCE tests

TLU tests

3.9.1

Basic Tests

Basic tests are run from ROM and include testing of the:
¢

Program ROM

Program RAM

NVRAM

Ethernet address ROM

Ethernet address RAM

Timer check

Packet memory

Packet memory refresh

Sections 3.9.1.1 through 3.9.1.9 provide a brief description of each of these basic tests.

Operation

3-43

3.9.1.1
®

—

Program ROM Test — This test does the following:

Performsa CRC32 calculation on the program ROM excluding the

ROM

last longword location in the

AN
RY

Compares the results of the calculated CRC with the CRC character

tion of the ROM

stored in the last word locaARG

3.9.1.2

Program RAM Test — Program RAM is tested for all stuck-at (SA1 or SAO)

pling faults using a modified version of the Nair, Thatte, and Abraham’

faults and cou-

s testing procedure (refer to

Functional Testing of Semiconductor Random Access Memories — Computin

g Surveys, Vol 15,

No. 3, Sept. 1983).

3.9.1.3

NVRAM Checksum Test — This test performs a checksum calculation on the

and compares the calculated checksum with the value stored in the NVRAM.
culated using the 16-bit ones complement binary arithmetic, shifting before

3.9.1.4

3.9.1.5

The checksum is cal-

adding.

NVRAM Write Test — This test verifies that the NVRAM can be written into. The

write test is disabled after the first pass of the self-test in manufacturing

NVRAM

mode.

Ethernet Address ROM Checksum Test — This test performs a checksu

on the Ethernet address ROM and compares the result of the calculation

m calculation

to the checksum value pre-

viously stored in the ROM. The checksum is calculated using the 16-bit
ones complement binary
arithmetic, shifting before adding.

3.9.1.6

Ethernet Address RAM Test — The Ethernet address RAM is tested for stuck-at

(SA1 or SA0) and coupling faults using 2 modified version of the N air, Thatte,

procedure.

faults

and Abraham’s testing
AR

3.9.1.7
®

Timer Test — The timer test does the following:

Starts the software clock and checks that the interrupt occurs at the correct

rect IPL.

Checks that the interrupt can be acknowledged or reset by reading address

Checks that the interrupt occurs at the correct interval (498 milliseconds).

3.9.1.8

location 4000.

Packet Memory Test — Packet memory is tested for stuck-at faults (SA1 or SA0) and cou-

pling faults using a modified version of the Nair, Thatte, and Abraham’s testing

3.9.1.9

vector and the cor-

procedure.

Packet Memory Refresh Test — This test verifies that packet memory refresh

ing. A data pattern is written into the RAM, the test waits 1 second, then

O EOTGS,

reads the location to verify

that the pattern read agrees with the pattern written. This is repeated until

SA1 and SAOQ faults.

is work-

all bits are checked for

3-44

LAN Bridge 100 Technical Manual

3.9.2

LANCE Tests

There are three categories of LANCE tests:
*

Reset

Internal loop

External loop

3.9.2.1 LANCE Reset Test — This test verifies that LANCE chips can be reset to a known state.
Bits in both LANCE chips are set to a predetermined state. LANCE A is reset and LANCE B is checked
to verify that the bits are still set, then LANCE A is checked to verify that it is cleared. Next, bits in
LANCE A are again set to the predetermined state. This time, LANCE B is reset and LANCE A is
checked to verify that the bits are still set. Finally, LANCE B is checked to verify that it is cleared.
3.9.2.2

Internal Loop Tests — The internal loop tests include:

Transmit CRC Logic Test

Receiver CRC Logic Test

Receive Bad CRC Test

Collision Test

Multicast Address Test

Reject Physical Address Test

Byte Swap and Broadcast Address Test

Transmit CRC Logic Test

This test clears the disable transmit CRC (DTCR) bit in the mode register to enable CRC generation
on the packet transmission. The test then transmits a packet and compares the received CRC character with a precalculated value.

Receive CRC Logic Test

This test sets the disable transmit CRC (DTCR) bit in the mode register to disable CRC generation on
packet transmission. The test then transmits a packet with a precalculated CRC character and
verifies that the packet was received correctly with no CRC errors.
Receive Bad CRC Test

This test sets the disable transmit CRC (DTCR) bit in the mode register to disable CRC generation on
packet transmission. The test then transmits a packet with a bad CRC character appended and
verifies that the receiver flags a CRC error.
Collision Test

In this test, a packet is transmitted in internal loop mode with the mode register collision bit set.
The test then verifies that the LANCE detects an error and retries the transmission 16 times. After
the sixteenth retry failure, a retry error is indicated.

Operation

3-45

Multicast Address Test
O

This test checks the ability of the LANCE to accept or reject a packet

destination address of the transmitted packet.

with the multicast bit set in the

The test uses the following procedure:
1.

Transmit a packet with an address that the logical address filter should

Verify that the packet is transmitted and received correctly.

Transmit a packet with an address that the logical address filter should

Verify that the packet is transmitted but not received.

accept.

reject.

Reject Physical Address
This test transmits a packet with the destination address not equal to
the LANCE address. The test
then verifies that the packet is transmitted correctly but is not accepted
by the LANCE receiver. The
test also checks the transmit status and looks for unexpected interrupt
s.

Byte Swap and Broadcast Address Check

The LANCE has the BYTE SWAP bit set in CSR3 and the destination address

of the packet is a broad-

cast address. When an internal loop packet is transmitted, the
receiver unit

packet. Transmit and receive status is checked along with the

3.9.2.3

should accept the

data.

External Loop Tests — These tests are used to determine the port-to-p

ration. A flag is set to enable the network port to network port

ort loop configu-

tion available. In port-to-port testing, packet sizes vary from 64 to 1500

nal loop tests:
®*

External loop test

Port-to-port loop test

testing if there is a valid configura-

bytes. There are two exter-

External Loop Test
This test takes the following steps to determine what network loopback

to the LAN Bridge 100 unit:
1.

Anexternal loop packet is sent on port A.

Ifthe packet is not received, the self-test considers it a hard error

An external loop packet is sent on port B.

Status and data are checked at both ports.

and halts.

If the packet is received back on port A, the self-test determines that

connected to both ports.

configuration is connected

there must be transceivers

If the external loop packet transmitted on port A is received on port B, the

that there must be a port-to-port loop configuration.

self-test determines

3-46

LAN Bridge 100 Technical Manual

7. If the external loop packet transmitted on port B is received on port A, the self-test sets a flag to
indicate that a port-to-port configuration exists.

8. If any of the status/data checks or transmit/receptions fail, the self-test indicates the hard error
and halts.

Port-to-Port Loop Test

The port-to-port test requires that both ports are connected together through a manufacturing

loopback cable. The test also requires that the port-to-port flag be set.
NOTE

This test will fail unless the loop-on-self-test switch located in the LAN Bridge 100

I/0 panel is set to the Enable position.

The port-to-port loop test creates a maximum amount of activity on the LAN Bridge 100 hardware
by starting operations with the binary search hardware and timer hardware at the same time as it
does port-to-port looping through the manufacturing loopback cable. The size of the packets transmitted varies from 64 bytes to 1518 bytes. Both transmit and receive data chaining is tested.
3.9.3

Table Lookup (TLU) Tests

The TLU tests test the following:
e

Status RAM

Basic binary search

Binary search engine

3.9.3.1

Status RAM Test — Status RAM is tested for stuck-at faults (SA1 or SA0) and coupling

faults using a modified version of the Nair, Thatte, and Abraham’s testing procedure.

3.9.3.2 Basic Binary Search Test — This test verifies that the search registers can be read from
and written into by the processor. The test also verifies that data from the search registers can be
written into the Ethernet address table RAM in 64-bit increments by writing to addresses
FCOO0OO0-FFFFF (this tests the 64-bit move function).

3.9.3.3 Binary Search Engine Test — This test verifies that the binary search engine can search
the network address RAM for entries scattered throughout its address range. A check is made to
ensure that a binary search is actually done during the search for stored network addresses.

Operation

3-47

PSS

AR5

[

4
-

Technical Description

4.1

Introduction

—

This chapter contains a technical description of the hardware and hardware subsystems that make
up the LAN Bridge 100 unit. The discussion takes a top-down approach, beginning with the system
level description (LAN Bridge 100 system) and ending with discussions at the circuit level.

4.2

Hardware Overview

The LAN Bridge 100 unit is made up of four subsystems, as shown in Figure 4—1. Each subsystem

provides the following functions:

The network interconnect (NI) subsystem receives packets from the LAN on one side of the
LAN Bridge 100 unit and stores them in memory. The NI may forward the packets to the LAN
on the other side depending on a determination made by the LAN Bridge 100 unit firmware.

The table lookup (TLU) subsystem performs searches of 48-bit Ethernet addresses stored in an

8K-by-48-bit address table. The results of the search are used by the LAN Bridge 100 firmware

to determine whether a packet should be forwarded to the LAN on the other side of the LAN

Bridge 100 unit.

e
e

The processor subsystem keeps track of packets stored in packet memory and determines
whether those packets will be filtered or forwarded. The processor maintains the Ethernet
address table in the TLU. The processor also responds to loop detection messages, Maintenance

Operation Protocol functions, and bridge management or Remote Bridge Management Software (RBMS).

J—

The power supply subsystem provides all of the voltages required to support the NI, TLU, and

processor subsystems as well as the transceivers connected to port A and port B.

4-1

LAN A

LAN BRIDGE 100

VT 7

| SUBSYSTEM
|
|

Lo——___4J

}

PROCESSOR |

! suBSYSTEM |
A

-—————— 9

TABLE

TM Lookup

| SUBSYSTEM|

mmmmmmmmmmmm

| POWER SUPPLY SUBSYSTEM |
AR

LAN B

LKG-0763

Figure 4-1:

LAN Bridge 100 Subsystems
S

4.2.1

Overview of LAN Bridge 100 Functional Blocks

The LAN Bridge 100 subsystems interact through a system of bus structures and control lines.

block diagram in Figure 4-2 shows relationships among the major circuits of the LAN Bridge

unit.

The design incorporates a system of bus switches (latches) that isolate memory structures

100
AT

Bridge

4-2

in each of

the subsystems. This important design element enhances the operating speed of the LAN
100 unit by allowing concurrent operation of the NI, TLU, and processor subsystems

The

LAN Bridge 100 Technical Manual

a—————s

1 ST

ADDRESS

| TaBLe

.._..._....l

88000
MICROPROCESSOR

A 01-19

LATCH

3
|

¢
|

A axxre

—1_-

‘

] TRANSC

COMPARATOR

RAM

LY,

praer

BUS

/| TRanscf

COMPARATOR
SEARCH

gé

RE
|
|.

SA 0143

EMmEIO-Apr

l
'

|
|
l

l
l
l
l

LADD-LAS

LMmMEIO=pr

ADDRE!
LATCH

\ IMKMr AN CE BOMBDO

Jal

0-1 D LATOH

120

4tl¢+

t vy by

]

i
l
l

|
I
l
i
!
l

|
|

L g

momaRS
e

|
|

I
I

BT PO

SPAEARE
T
W

LKG-0764

Figure 4-2:
4.2.2

LAN Bridge 100 Functional Block Diagram

The DataPath

This section provides a general discussion of one typical LAN Bridge 100 operation to illustrate
how the hardware is used to forward packets. Each of the circuits mentioned in the discussion here
is shown in Figure 4-2.

Technical Description

4-3

As an example of LAN Bridge 100 operation, assume that a minimum size packetis present at port

A. The LAN Bridge 100 unit performs the following steps:
1.

The serial interface adapter (SIA) receives the packet from port A transceiver and sends it to the

Local Area Network Controller for Ethernet (LANCE).

The LANCE stores the packet in packet memory and clears an OWN bit in the LANCE A receive

descriptor ring. This causes the OWN interrupt generator to send an OWN interrupt to the proCesSor.

The processor responds to the OWN interrupt by reading the LANCE A receive descriptor ring
that points to the buffer containing the packet. The processor then reads the 48-bit destination

address (locatedin the first 6 bytes of the packet) and copies it into the TLU.

The TLU performs a binary search, which compares the destination address against addresses

already stored (in ascending numerical order) in the address table. The processor uses the
results of the search to determine whether the packet must be forwarded.

OGO

One of the following will occur:
®

If no match is found (that is, if the destination address does not equal a known source
address), the processor, by default, forwards the packet to the LAN at port B. Go to step 4.

If a match is found (that is, if the destination address equals a known source address), the
processor checks the corresponding 16-bit status word in the processor RAM to determine

the location (port A or port B) of the station.
If the destination is on side A, the processor disregards the packet. Go to step 5.

If the destinationis on side B, the processor determines, by default, that the packetis to be

forwarded to the LAN at port B.

The LAN Bridge 100 unit transmits the packet to port B as follows:

The processor writes the data from the LANCE A receive descriptor ring into the LANCE B
transmit descriptor ring, resets the OWN bit in the descriptor ring, and sets the transmit

demand bit in LANCE B CSR.
®

When transmit conditions at port B are met, LANCE B reads the packet from packet memory and transmits it to the LAN at port B using the serial interface adapter (SIA).

The LAN Bridge 100 unit compares the source address of the incoming packet to addresses
stored in the Ethernet address table. This keeps the address table and status word up to date.

4-4

LAN Bridge 100 Technical Manual

4.3

Processor Subsystem

The processor subsystem is responsible for overall operation of the LAN Bridge 100 unit. The processor allocates buffers to the LANCE network controllers, initiates binary searches, and maintains
the address table of the TLU.

4.3.1

Processor Circuit Descriptions

The processor consists of the following circuitry (see Figure 4-3):
®

Motorola 68000 microprocessor
Interrupt controller
Processor memory

System timers
Reset circuitry

To minimize capacitive loading of the processor data bus (D00-D15), a secondary data bus (UDOO-

UD15) is incorporated into the architecture (see Figure 4-3). The secondary data bus is joined to
the primary data bus by a pair of octal bus transceivers. The secondary data bus is enabled by the
address decoding circuitry.

Technical Descriptioin

4-5

$9.0-9M1

SW3LSASENsS
.

|——

Sig-00a

WOHd

WVHAN

|Adonwaw

gHOSS30Hd

13INH3IHLI WSO3HAdv

WOHd3

Wvd

sJOnNgVYS1H34SO1SI934

rmmw

SSH31SI1934
NO3INVIT3OSIW

—

~ ANV IN

nNiLoL

ar-poqgr 5L

W3LSAS

SHIWIL

1dNY|ILNI

JSINTH.LD

1301310HMd

NV Hv31d

S1dNYILNI IAOYNH4

Blv

§1Q-000

00089

HOS30Hd

1dl
e

S——

Figure 4-3

4-6

Processor Subsystem

LAN Bridge 100 Technical Manual

4.3.2

The Microprocessor

The heart of the processor subsystem is the microprocessor. The microprocessor used in the LAN
Bridge 100 unit is the Motorola 68000, a single-chip device with address and data on separate
buses. The 68000 has built-in mechanisms for handling incoming interrupt information and for
controlling bus arbitration. The 68000 also provides status information that allows decoding of the
current bus operation.

The 68000 microprocessor has a 32-bit internal architecture and a 16-bit external data path. The
address bus is 23 bits wide with separate control lines for accessing the upper and lower bits either
as separate bytes or together as words. The processor is driven by a 10-MHz (100-ns) system clock
that results in a processor cycle time of 400 ns without WAIT states.
4.3.3

Interrupt Controller

The interrupt controller (see Figure 4—4) monitors the interrupt priority level (IPL) lines. The controller responds to the interrupt request if the priority of the request is higher than the priority of
the processor. If two or more interrupt requests are presented to the interrupt controller at the

same time, the highest level is passed on to the processor and the other levels remain pending.
The microprocessor acknowledges the interrupt by placing the interrupt level on address lines A1
through A3. When the interrupt vector generator is enabled by the function code lines (FCO-FC2),

it reads the address lines A1 through A3 and places a corresponding interrupt vector onto the data
bus. Table 4-1 identifies each interrupt, its priority, and its vector.

Technical Description

4-7

Table 4-1:

Hardware Interrupts

Interrupt Level

Vector Address

Interrupt

Watchdog timer.

LANCE B interrupt. Generated when a CSRO status flag is set for

LANCE B.

LANCE A interrupt. Generated when a CSRO status flag is set for
LANCE A.

OWN bit interrupt B. Generated when an OWN bit (set by
LANCE B) is detected.

OWN bit interrupt A. Generated when an OWN bit (set by
LANCE A) is detected.
|

Fiber-optic self-test. Generated during self-test if a fiber-optic
module is installed. This causes self-test to verify operation of

the fiber-optic circuitry.
S

FIBER-OPTIC l

SELF-TEST

LANCE A

OWN INTR
PAL

LANCEB
OWN INTR
PAL

WW O

‘

IPLO

PL 1

BA 1

ADDRESS
BUS

LATCH

—

BA2
BA 3

| INTERRUPT
VECTOR

GENERATOR

IPL 1

LANCE A
INTR

[ |

UDO00-UDO7

. ? SECONDARY

|— | DATABUS

IPL2

ENABLE
A

68000

"MICROPROCESSOR

FCO

FC 1

LANCEB

INTR

FC7

INT ACK

—

FC2

A""""’"

SDI

TIMING

WATCHDOG
TIMER

LKG-0766

Figure 4-4:

4-8

Hardware Interrupt Configuration

LAN Bridge 100 Technical Manual

4.3.4

Processor Memory

The processor memory mmains the following components:
e

PROM (stores the program code)

RAM (holds a copy of the program code)

RAM (holds the address table status)

PROM (holds the physical address of the LAN Bridge 100 unit)

NVRAM (records diagnostic and bridge management parameters)

Miscellaneous control registers

4.3.4.1 PROM — The PROM (programmable, read-only memory) section of processor memory
stores 16K words and contains all native LAN Bridge 100 firmware, including self-test.

4.3.4.2 Program RAM — Program RAM contains 8K words and is made up of two 8K-by-8-bit
static RAM devices. The program RAM contains a copy of the LAN Bridge 100 program code and
data structures. The program executes from program RAM after the LAN Bridge 100 unit has powered up and self-test has passed.

4.3.4.3 Address Table Status RAM — The address table status RAM contains 8K words and is
made up of two 8K-by-8-bit static RAM devices. The address table status RAM contains status words
that correspond to each entry in the Ethernet address table of the TLU. The status includes information such as the following:

The port (port A or port B) on which the address resides

The age of the address

Whether the address is used by the LAN Bridge 100 unit for special purposes

Whether the address is learned through normal LAN Bridge 100 operation or is fixed (written

into memory by bridge management so that it cannot age)

4.3.4.4 Physical Address PROM — The physical address PROM is a 32-bit-by-8-bit PROM

device. This PROM contains the unique 48-bit (6 bytes) station address of the LAN Bridge 100 unit

and a 2-byte checksum of the address.

The physical address is used for bridge management functions, Maintenance Operation Protocol

(MOP) functions and for loop detection purposes.

4.3.4.5 NVRAM — Nonvolatile RAM (NVRAM) contains 2K bytes and is a single 2K-by-8-bit
memory device. NVRAM stores diagnostic information and LAN Bridge 100 parameters during a
power-down period.

NVRAM is write protected by the reset circuitry during all power-up and power-down cycles.

Technical Description

4-9

4.3.4.6

Miscellaneous Control Registers —

The miscellaneous control registers monitor and

control operation of the LAN Bridge 100 unit. The registers and their functions are as follows:
®

Miscellaneous control registers enable interrupts and loopback functions. A memory map con-

trol bit causes RAM address space to be remapped after power up is completed (see Section
4.3.5.0).
®
®*

Aself-test error latch holds error status if errors occur during self-test.
A programming selection buffer holds LAN Bridge 100 configuration information such as

switch settings or whether a fiber-optic module is present. The processor uses this buffer to set

program functions.

4.3.5

Processor Memory Map

The LAN Bridge 100 unit has approximately 1 megabyte of address space. When the LAN Bridge

100 unit is powered on, the program executes from program ROM. Program RAM is hidden and is
not accessible. The memory map shown in Figure 4-5 illustrates the mapping used for powering up

P05

the LAN Bridge 100 unit.

ORI

4-10

LAN Bridge 100 Technical Manual

—

PROGRAM ROM

MISCELLANEOUS

(04000)

NONVOLATILE
MEMORY

(05000)
(O5FFF)

LANCE DESCRIPTOR
RINGS

(06000)
(O6FFF)

LANCE CSR

(07000)

REGISTERS

(O7FFF)

PROGRAM ROM

(10000)

PACKET
MEMORY

(20000)
(3FFFF)

ETHERNET
ADDRESS TABLE

(E0000)
(EBFFF)

ETHERNET ADDRESS
TABLE STATUS RAM

(EC000)
(EFFFF)

COMPARE-AND-MOVE
REGISTERS

(F4000)
(F4FFF)

ETHERNET ADDRESS
TABLE STATUS RAM

(FB0OO)
(FBFFF)

ETHERNET ADDRESS
TABLE 64-BIT MOVE

(FC000)
(FCFFF)

CONTROL

(00000)
(O3FFF)

(O4FFF)

(1FFFF)

LKG-0767

Figure 4-5:

Address Mapping Used During Power Up

After the bridge completes self-test, the address space for program RAM is remapped to the range of

00000-03FFF and the program firmware is copied from ROM into program RAM. Thereafter, the

program executes from program RAM. Executing the program from RAM provides more flexibility
by allowing the use of soft interrupt vectors.

The memory map shown in Figure 4-6 illustrates the mapping used for operation of the LAN
Bridge 100 unit after power up and self-test have completed.

Technical Description

4-11

PROGRAM RAM

(00000)

(O3FFF)
MISCELLANEOUS

CONTROL
NONVOLATILE
MEMORY
LANCE DESCRIPTOR

RINGS
LANCE CSR

REGISTERS

PROGRAM ROM

(04000)

(04FFF)
(05000)
(O5FFF)
(06000)

(O6FFF)
(07000)

(O7FFF)

(10000)
(1FFFF)

PACKET

(20000)

MEMORY

(3FFFF)

ETHERNET

(E0000)

ADDRESS TABLE

(EBFFF)

ETHERNET ADDRESS

(EC000)

TABLE STATUS RAM

(EFFFF)

COMPARE-AND-MOVE
REGISTERS

(FAFFF)

ETHERNET ADDRESS

(FB0OO)

TABLE STATUS RAM

(F4000)
E

(FBFFF)
R

ETHERNET ADDRESS
TABLE 64-BIT MOVE

(FC000)
(FCFFF)
A

LKG-0768

——

Figure 4-6:
4.3.5.1

Address Mapping Used for Normal LAN Bridge 100 Operation

Program ROM —

—

Program ROM contains the firmware that is used to control the oper-

ation of the LAN Bridge 100 unit.

During power up, program ROM is mapped into the ranges of 00000-03FFF and 10000-1FFF
F.
After power up and self-test are complete, RAM is mapped into the range of 00000—-03F
FF and the
firmware (except for self-test) is copied into RAM. Subsequent program execution is from
RAM.

4.3.5.2

Miscellaneous Control — The miscellaneous control region (04000-04FFF) contains

various registers used to monitor and control operation of the LAN Bridge 100 unit
(see Section

4.3.4.6).

4-12

LAN Bridge 100 Technical Manual
RGNS

4.3.5.3

Nonvolatile Memory — Nonvolatile memory (NVRAM) resides in memory address

is powered off. If power up self-test fails, NVRAM stores a diagnostic error code.

4.3.5.4 LANCE Descriptor Rings — The LANCE A and LANCE B transmit and receive descriptor rings reside in address space 06000—0GFFF. To clear the OWN interrupt, the OWN bit in the
descriptor rings are cleared using the range of 06000—-06FFF.
The LANCE descriptor rings are used to allocate and find buffer space in packet memory.

Note that the descriptor rings also reside within the packet memory address space and range from

26000 through 26FFF. This is the range used by the LANCE processors to access the descriptor

rings.

4.3.5.5 LANCE CSR Registers — The LANCE A and LANCE B CSR registers (a total of eight
registers) are used to monitor and control the operation of the LANCE processors. The registers are
mapped into the range of 07000-07FFF.

4.3.5.6

Program RAM — Program RAM is used for program execution after the LAN Bridge

100 unit powers up and self-test is complete.

During power up of the LAN Bridge 100 unit, program RAM is hidden and is not accessible. After
power up and self-test are complete, RAM is remapped into the range of 00000-03FFF and the firmware (except for self-test) is copied into RAM. The interrupt vector address space is included in
RAM after RAM is remapped.

4.3.5.7 Packet Memory — Packet memory (20000-3FFFF) is used to store packets while the
LAN Bridge 100 unit is determining whether to forward or filter the packets. The LANCE A and
LANCE B descriptor rings and the initialization blocks also reside in packet memory.

4.3.5.8 Ethernet Address Table — The Ethernet address table (ECO0O0-EBFFF)is 8K by 48 bits
and contains 48-bit Ethernet station addresses. A station address (and a corresponding status word)
is stored when a packet is received from that station.

4.3.5.9 Ethernet Address Table Status RAM — The Ethernet address table status RAM is 8K
by 16 bits and contains 16-bit status words that correspond to each station address stored in the
Ethernet address table (see Section 4.3.4.3).
The Ethernet address table status RAM is mapped into two regions of address space: ECOO0-EFFFF,
and FS000-FBFFF. Both regions contain the same data.

4.3.5.10 Compare-and-Move Registers — Three 16-bit compare-and-move registers (F4000F4FFF) are used to store search arguments (48-bit Ethernet addresses), while magnitude comparisons are performed by the TLU. These registers are also used in the read/write operation associated
with a 64-bit move.
Each register and address is identified as follows:

F4xx0
F4xx2

Search register with address bits <47-32>
Search register with address bits <31-16>

F4xx4
F4xx06

Search register with address bits <15-00>
Results register

Technical Description

4-13

—

Writing to address F4xx4 triggers the binary search of the Ethernet address table. The binary search
starts at a midpoint in the address table defined by address lines BA13-BA03. If the address lines
BA13-BAO3 are all zeros, then a single entry comparison is performed using location 0 in the

Ethernet address table.

Writing to address F4xx6 also triggers a binary search of the Ethernet address table. In this case, the
TLU searches the address table using the same search argument that was used for the previous

search.

4.3.5.11

Ethernet Address Table 64-Bit Move —

Address space FCOO0-FCFFF accesses the

Ethernet address table and the Ethernet address table status RAM. The memory configuration
addressed in this range is 8K by 64 bits (64 bits include a 48-bit Ethernet address and a 16-bit status

word). This capabilityis necessary for performing a 64-bit move (see Chapter 3 for more informa-

tion on 64-bit moves).
4.3.6

System Timers

The system timer is driven by the 10-MHz system clock from the NI subsystem and provides the following functions:
O

Clock interrupt to the processor

Hardware watchdog timer

(Clock to the refresh counter

The clock interrupt to the processor occurs about every 0.5 seconds and drives a real-time clock.
This clock is used to keep track of how long each network address has remained unused in the
address table. If a station has been inactive for 2 minutes, the LAN Bridge 100 unit considers the

corresponding address to be invalid.

The watchdog timer is used as a hardware check on the firmware. It is the responsibility of the firmware to reset the watchdog timer once every 15 seconds. If the timer is not reset, it will expire and
cause a system reset. If the firmware is operating properly, this situation will not occur.

The clock to the refresh counter occurs every 12.5 [s and is used to increment the refresh counter.

4.3.7

Reset Circuitry

The reset circuitry is responsible for the following functions:
¢

Resetting the hardware on power up

Resetting the hardware upon a watchdog timer reset

Write protecting the NVRAM during power up, power down, and reset sequences

The reset circuitry monitors VCC (5 volts) to sense when power is stable. During power up, the
reset circuitry holds the LAN Bridge 100 unit in reset mode and keeps the NVRAM write protected

until all state-saving devices are stable.
B

[

4-14

LAN Bridge 100 Technical Manual

4.4

Network Interconnect Subsystem

The network interconnect (NI) subsystem provides the interface between the other LAN Bridge
100 subsystems and the physical channel of the extended LAN by performing the following functions:

Receiving and decapsulating packets from the LAN connected to port A or port B of the LAN
Bridge 100 unit.

Writing packets into packet memory in locations determined by the processor.

Writing packet status information into the receive descriptor rings.

Reading the transmit descriptor rings to determine the location(s) of any packet that is ready for
transmission.

Reading packets from packet memory and encapsulating packets for transmission to the LAN

Transmitting encapsulated packets to the specified LAN according to the Ethernet channel

s
-

=
-

connected to side A or side B.
access method.

4.4.1

NI Subsystem Circuits

The NI subsystem is made up of two physical channel interface circuits (NI ports), a LANCE CSR

control circuit, and a packet memory. Figure 4-7 is a block diagram of the NI subsystem.
The NI subsystem bus structures and packet memory are isolated from processor and TLU bus
structures and memory. This allows concurrent operation of the NI, processor, and TLU and helps

minimize the time required to perform LAN Bridge 100 functions. In addition, the LANCE processor and microprocessor buses have independent access to packet memory. This allows maximum
utilization of the memory cycle time and guarantees minimal delay for access.

Technical Description

4-15

PROCESSOR ADDRESS BUS BAD1-BATS
BN

LANCE A

INTERRUPT

LANCEB
OWN INTERRUPT

OWN INTERRUPT

LANCE A

LANCE B

OWN INT.

GENERATOR

PORT

LANCER
INTERRUPT

RIS

AR SS

LANCE l
LANCE
A
|
C_—_._.> ADDRESS
A

SIA
— A

LATCH

PORT
B
Ay

|
LAQOLATS

N/ N\ /

5SIA

LBOO-LB1S

LANCE

8:1 ADDRESS

MULTIPLEXER

LANCE A

LANCE B

ADDRESS AND DATA BUS

Eaind

L
[

o
«

<
8

2
8

o |

84K x 16 BITS

£
&

PACKET

MEMORY DATA BUS

1 >
LANCEB

‘

<‘

CSR BUS
REGISTER

PROCESSOR DATA BUS D0O-D15

)
Y

LRG-D4BS

Figure 4-7:
4.4.2

NI Subsystem

NiPorts
B,

The NI subsystem has two ports. Each interface circuit connects to one of the two LANSs joined by

the LAN Bridge 100 unit. Each NI port consists of one SIA chip and one LANCE chip. This discussion describes one NI port; the other NI port is identical.

NOTE
RS

In a remote LAN Bridge 100 unit, a fiber-optic module is included in port A. The

fiber-optic interface is described in Section 4.4.2.5.
S

4-16

LAN Bridge 100 Technical Manual

4.4.2.1

SIAChip — The SIA chip is a 24-pin device that interfaces the LANCE chip to the net-

work transceiver. The SIA chip performs the following functions:
¢

Detection of data presence on the network

Conversion of collision-presence signals to TTL levels

Manchester decoding of data

Manchester encoding of data

Interfacing of differential signal line pairs and TTL levels

The SIA chip requires a 20-MHz crystal for its internal oscillator. The oscillator generates the clock
for the Manchester-encoded data stream. The 20-MHz is halved to provide the 10-MHz transmit
clock and receive clock. The LANCE uses the transmit clock to set the data rate. The LANCE uses
the receive clock to gate sampling of the receive data line.

4.4.2.2 LANCE Chip — The LANCE is a 48-pin, very-large-scale-integration (VLSI) device
designed to provide data link services over a CSMA/CD local area network. The LANCE chip performs the following functions:

Direct memory access to packet memory

Buffer management by using descriptor rings in packet memory

Runt packet filtering

CRC generation and checking

Transmission backoff and retry

The LANCE operates in promiscuous mode. This mode causes the LANCE to accept all packets
regardless of their destination address. The LANCE is monitored and controlled by control and
status registers (CSRs), which are physically separate from the LANCE. CSR control is described in
more detail in the following section.

The LAN Bridge 100 unit does not use the LANCE’s ability to generate CRC. However, the bridge

does preserve incoming CRC sequences for use on outgoing packets. Note that the probability of
externally undetectable errors originating within the LAN Bridge is no greater than 1 in 109.
A table called the initialization block that is stored in system memory is loaded into packet memory
upon initialization. The initialization block is used by the LANCE to do the following:
e

Determine the size and location of the descriptor rings in packet memory

Set operating parameters for the LANCE (such as promiscuous mode)

LANCE Control and Status Registers — There are four control and status registers (CSRs) within
each LANCE that are programmed by the processor. The CSRs are accessed through a register
address port (RAP) and a register data port (RDP) as shown in Figure 4-8.

Technical Description

4-17

Data is read from or written into a CSR in a two-step operation.
AR

The CSR is addressed by writing the register address into the regis'ter address port (RAP) of the

LANCE.
2.

Data that is read from or written into the register data port (RDP) is also read from or written

into the CSR that is selected by the RAP.

During the final stage of LAN Bridge 100 initialization, the microprocessor sets the RAP to CSRO.
Once the RAP is set, it remains set until rewritten, and it is not rewritten once the LANCE starts

operation. With the RAP set to CSRO, each time the RDP is accessed, CSRO is accessed.

memmmmmm

mmmwmmmm

NI SUBSYSTEM

PROCESSOR SUBSYSTEM

|
——

CSR

} ADDRESS AND

CONTROL
|

CONTROL LINES

ARBITRATION |

processor
DATA BUS

CSR
BUS
REGISTER
16-BIT

" LATCH

DATA
PORT

|
|

ADDRESS

- PORT

<‘

LANCE ADDRESS

AND DATA BUS

|
|

HEORBG

LKG-0464

Figure 4-8:

CSR Control

4-18

LAN Bridge 100 Technical Manual

4.4.2.3

LANCE CSR Bus Registers — The CSR bus registers are bidirectional latches that

interface the processor data bus to the LANCE address/data bus (see Figure 4-8). The latches isolate
the two buses to allow concurrent operation of the NI and processor subsystems. They also mini-

mize processor WAIT states since the processor does not have to wait for the LANCE address/data
bus to read the LANCE CSR registers.

4.4.2.4

LANCE CSR Control Circuit — The LANCE CSR control circuit (see Figure 4-8) man-

ages the transfer of data between the LANCE internal CSRs and the CSR bus registers.

After the LANCE

is initialized, the register address port (RAP) is set to CSRO (see Section 4.4.2.3).

CSRO remains selected as long as the LANCE continues to operate.

When the processor wants to read from CSRO, it actually reads from the CSR bus register, which is
part of the processor’s memory map. Reading from the bus register triggers the CSR control to initiate a read from the RDP into the bus register. Therefore, each time the bus register is read, the data
in the bus register is updated. This guarantees that the processor does not have to wait for the
LANCE bus, but it also means that the data obtained from the bus register is old and must be treated
as such. The processor must read the bus register twice to get fresh data. However, the processor
will incur WAIT states if it attempts back-to-back accesses of the CSR while the LANCE is performing a DMA operation.

When the processor wants to write to CSRO, it writes to the CSR bus register. The CSR control arbitrates for the LANCE bus and writes the data into CSRO as soon as the LANCE bus is available. The
maximum delay incurred is 4.8 [s.
4.4.2.5 Fiber-Optic Module — The fiber-optic module is included only in remote LAN Bridge
100 units and performs functions similar to those of a transceiver.
The fiber-optic module converts emitter-coupled logic (ECL) levels from the SIA to optical signals
for transmission through a fiber-optic cable. Conversely, optical signals received from the fiberoptic cable are converted to ECL levels for use by the SIA circuitry.
4.4.3

LANCE Address Latch

There are two LANCE address latches, one for each LANCE (see Figure 4-7). This section describes
the address latch for one LANCE. An identical circuit exists for the other LANCE.

The LANCE address latch multiplexes the LANCE address/data bus to provide a separate address
bus and data bus for packet memory access.

The LANCE latches a packet memory address into the latch by asserting the address latch enable
(ALE) signal to the latch and writing the address onto the LANCE address/data bus. The latched
address is then placed onto the LANCE address bus where it is used by the address multiplexer to
access packet memory.

When the ALE signal is deasserted, the LANCE address/data bus becomes available for data functions.

Technical Description

4-19

4.4.4

OWN Interrupt Generator

There are two OWN interrupt generators, one for each LANCE (see Figure 4-7). This section
describes the OWN interrupt generator for one LANCE. An identical circuit exists for the other
LANCE.
The OWN interrupt generator is a programmed array logic (PAL) device that generates an interrupt
when it detects that an OWN bit has been set in the LANCE receive descriptor ring. The interrupt is
sent to the interrupt controller in the processor subsystem.

In packet memory, every fourth address in the range of receive descriptor ring addresses contains

an OWN bit. When the OWN interrupt generator detects these specific addresses on the LANCE

address bus, it looks at bit < 15> of the LANCE address/data bus (bit < 15> carries the OWN bit). If
the PAL detects that an OWN bit is cleared, it generates an interrupt signal. This signals the proces-

sor that a packet memory buffer is ready for processing by the processor and TLU subsystems. See

Chapter 3 for more information on the OWN bit function.

4.4.5

Packet Memory

Packet memory (see Figure 4-7) stores packets during the forward/filtration process. In addition,

this memory contains the descriptor rings (transmit and receive descriptor rings) and the initializa-

tion blocks for each LANCE.

The packet memory architecture allows access by either of the LANCES, by the processor, or by the

refresh address counter.

The cycle times of the devices that access packet memory are inherently slower than the memory

cycle access time, which is about 300 ns. To minimize memory circuit WAIT states, data bus latches
hold the data retrieved from packet memory until it can be read by the requesting device. This
improves memory performance by allowing the memory to begin its next cycle immediately without waiting for slower devices.

Packet memory is made up of the following circuits:
®

Address multiplexer

Memory control circuit

Dynamic RAM (DRAM)
NN

Data latches

4.4.5.1
Address Multiplexer — The address multiplexer (see Figure 4-7) determines which
device (LANCE A, LANCE B, the processor, or the DRAM refresh counter) addresses packet mem-

ory. Steering inputs to the multiplexer come from the memory control circuits (see Section

4.4.5.2).

When a 16-bit address is present at the input to the multiplexer, the steering inputs apply the high-

order (row) address and then the low-order (column) address onto the DRAM address bus.

4-20

LAN Bridge 100 Technical Manual

4.4.5.2

Memory Control Circuit — The memory control circuit (see Figure 4-7) coordinates

memory activity and arbitrates the priority of the LANCE, processor, and refresh counter.

Table 4-2 identifies the priorities assigned to the four devices. Figure 4-9 is a block diagram of the
_—

memory control circuit. The memory control circuits perform the following functions:
¢

The synchronization circuit synchmmzes the control signals of memory accessing devices with
the memory control circuit.

The arbitration circuit selects which device (processor, LANCE A, LANCE B, or refresh counter)
addresses packet memory.

®*

The timing circuit ensures that row and column address strobes and the write signal are properly gated to the DRAM.

Table 4-2:

||
s

Priority

Device Priorities

Device

Refresh

Processor

LANCEA *

LANCEB *

* Note: LANCE A and LANCE B alternate priority.

Technical Description

4-21

PROC CYC

(

_JFLIP-FLOP I:lj RFSHCYC

PROC CYC

LANACYC

LANCE READ,

TIMING

ARBITRATION

SYNCHRONIZATION

TIMING

\ 4

*
LANBCYC
PN

-—
AND
ALE,
DAS SIGNALS
7

SIG
A

coLset.

& F L

—

>| ip
LNB 100 ns

BUS SELECTED | SIGA SIGB
LANCE A

PROCESSOR

LANCE B

REFRESH

PROCESSOR BUS

N
DN

LANCE ABUS

MUX

LANCE B BUS

PACKET
MEMORY
A

REFRESH ADDRESS
SO

LKG-0770

Figure 4-9:
4.4.5.3

Memory Control Circuit

Dynamic Packet Memory —

The packet memory stores packets while the processor

and TLU perform the forward filtration process. The memory also contains the LANCE descriptor

rings and the LANCE initialization blocks. Refer to Chapter 3 for a description of the initialization

blocks and descriptor rings.
A

DRAM memory is made up of sixteen 64K-by-1-bit devices (each device is an array of 128 rows by

512 columns).
S

The DRAM address bus is 16 bits wide but is multiplexed into 8-bit row addresses and 8-bit column

addresses. Row address strobe (RAS) and column address strobe (CAS) signals select which row and
column are read from memory. RAS, CAS, and WRT (write-enable) signals are driven by the memory control circuit (see Section 4.4.5.2).

The DRAM data bus is 16 bits wide. When CAS strobes the DRAM, the data is either read from or
written to the data bus depending on the state of WRT.

Each DRAM row must be refreshed at least once every 2 ms. Since each DRAM contains 128 rows,
one row must be refreshed every 15 s (128 x 15 s = 1.920 ms). To allow for any delays, the

refresh counter addresses a row in memory every 12.5 [s (see Section 4.4.5.4).
RO

4-22

LAN Bridge 100 Technical Manual
[

4.4.5.4

Refresh Counter —

The refresh counter increments and accesses a row address in

packet memory every 12.5 ps.

The refresh counter is made up of a dual 4-bit binary counter that is configured as an 8-bit binary
counter. The incoming 12.5-Ws clock that increments the counter comes from the system timer (see
Section 4.3.6).

For additional information on the DRAM refresh cycle, see Section 4.4.5.3.
4.4.5.5 DataBusLatches — Packet memory data bus latches (see Figure 4-7) are controlled
by the memory control circuit (see Section 4.4.5.2). The latches interface the packet memory data
bus (PMD 00-15) to the following bus structures:
e

ADAL 00-15 (LANCE A address/data bus)

BDAL 00-15 (LANCE B address/data bus)

DO00-15 (processor data bus)

The data latches and buffers perform two functions:

They isolate the memory data bus from external data buses to allow concurrent operation of
the NI and other subsystems.

They hold data retrieved from memory until the data can be read by the requesting device. This
improves memory performance by allowing the memory to begin its next cycle without waiting for slower devices.

4.5

Table Lookup Subsystem

Each time a LANCE writes a new packet into packet memory, the processor reads the destination
address of the packet and sends the address to the table lookup (TLU) subsystem.
The TLU compares the destination addresses from incoming packets to addresses stored in mem-

ory. The results of each comparison are made available to the processor, which determines
whether the packet should be forwarded or ignored.

The TLU, with assistance from the processor subsystem, also maintains the Ethernet address table.
Table maintenance functions include comparing source addresses from incoming packets to

addresses stored in memory. The results of each comparison are made available to the processor,

which determines whether the address should be stored in the address table or ignored.
4.5.1

Overview of the TLU

The functional blocks of the TLU are shown in Figure 4—10. This illustration can guide you through
the overview of TLU operation.

Technical Description

4-23

HOSS3004Hd
snaviva
$1Q-00Q

LLO-ET
DG

Sne

.ST1O7HLNISN3OyDHa31nS19v3Y

13NOa

A
svd

Eivd-iovd

13INH3HLI

NOSIHV-JINOD

SNLVLS

$HOLVHVdWO

s171NsS3y

n__},v

-eLvs-iovsKgrxy8 NOSIHYJINODLig-spS3YAQvSNgHOYV3S

IR,

SS3yaav

TXNNS

1°_

XNW
RO
Ds:

Figure 4-10

Table Lookup Subsystem
R

4-24

LAN Bridge 100 Technical Manual
GG

1. The processor initiates the TLU binary search process in the following way:

The processor writes the search argument (the new 48-bit Ethernet address) from the data
bus into the search compare registers of the comparator PALs. The processor uses three
write cycles to write the three address words into the search compare registers. The words
are written in sequence of the low, middle, and high word of the address.

The search control and status PAL detects when the final (highest) byte is written into the
search compare registers by the processor. After a delay of about 50 ns, the search control
and status PAL deasserts the DONE signal. The deasserted DONE signal does two things:
After 50 ns, it steers the multiplexer to select subsequent addresses from the search control
and status PALs.

After 100 ns, it prevents the processor from addressing any registers in the TLU.

Before the DONE signal is deasserted (during the 50-ns delay), the processor addresses the
location in the Ethernet address table, where the binary search begins. This address is used
to start the binary search. Addresses for subsequent probes of the address table are provided

by the search control and status PALs.

2. Once the processor has written the first address to be probed in the Ethernet address table, the
search control and status PALSs in the TLU assume control of the binary search process.
NOTE

Ethernet addresses are stored in order of magnitude in the address table, higher
locations having the higher addresses. This ordering is necessary for the binary
search process to function. See Chapter 3 for a detailed description of the binary
search process.

The comparator PALs perform bit-by-bit comparisons of the latched search argument
against addresses probed in the address table.

The search control and status PALs use the results of each comparison (> [greater than],
< [less than], or = [equal to]) to determine the next step in the search process.

“Greater than’’ causes a search to be performed at a higher location in the address table.
‘““Less than’’ causes a search to be performed at a lower location in the address table.
‘““Equal to”’ indicates a match. A match terminates the search.

* The binary search ends either when a match is found or when the bottom address of the
binary tree (an odd address) is probed and no match has been found.

The search control and status PALs assert DONE. This enables the decoder to allow the processor to read the search results register. This register indicates whether a match was found
and contains the location of the last address searched.

3. After the search, the search control and status PALs report the results of the search by setting
““equal to” or ‘‘greater than’’ bits in the search results register. The address of the last location
probed is included in the register data. These results dictate what the processor does with the
location address.

Technical Description

4-25

If the ““‘equal to”’ bit is set, it means that a match was found.
If a destination address was compared, the processor checks a corresponding 16-bit status
word in RAM to determine the location (bridge port A or port B) of the Ethernet address and
whether the packet should be forwarded or ignored. The status word location in RAM corresponds to the Ethernet address location in the address table.

If a source address was compared, the processor updates the status word in RAM.

If the “‘greater than’’ bit is set, it means the search argument was greater than the address in

the last location probed. This also indicates that no match was found.

If a destination address was compared, the processor forwards the packet since it cannot
determine the location (port A or B) of the station with that address.
A

If a source address was compared, the processor adds the new address and status word to

the Ethernet address table as a background task.
R

If neither the “‘greater than’’ nor the ‘“‘equal to’’ bit is set, it means the search argument was

less than the address in the last location probed. This also indicates that no match was

found.

If a destination address was compared, the processor forwards the packet, since it cannot
determine the location (port A or B) of the station with that address.
If a source address was compared, the processor adds the new address and status word to
the Ethernet address table as a background task.

4.5.2

Circuit Descriptions

The TLU contains the following circuits:

Search address multiplexer
R

48-bit address comparator circuitry, which includes:
Ethernet address table memory
L

Bus transceivers

Ethernet search address comparator PALs
O

Binary search control and status PALs

Search results register

4.5.3

Search Address Multiplexer

The search address multiplexer (see Figure 4-10) selects the processor bus (BAO1-BA13) or the

search control bus (PAO1-PA13) to address the Ethernet address table memory.
Four quad 2:1 multiplexers make up the multiplexer circuit. The single steering input to the multiplexer is a signal called SMUXL. This signal is derived from the DONE signal when DONE is

deasserted by the search control and status PALs. SMUXL is gated to the multiplexer by the 100-ns

clock signal.

4-26

LAN Bridge 100 Technical Manual

4.5.4

48-Bit Address Comparator

The 48-bit address comparator performs bit-by-bit comparisons of the 48 bits that make up an
Ethernet address. The PALs contain the compare and move registers (see Section 4.3.5.10) that are
used for 64-bit moves (see Chapter 3).

4.5.4.1 Functional Description — The 48-bit comparator is made up of three 16-bit comparators. Figure 4-11 shows the structure of a 48-bit comparator. This illustration shows the circuits
that compare the low 16 bits, the middle 16 bits, and the high 16 bits of the 48-bit Ethernet address.

o
‘o

The address table memory is 8K by 48 bits. Thus one location in memory accesses the entire 48-bit

Ethernet address.

Technical Description

4-27

ETHERNET
ADDRESS

TABLE

gggpmmm
|

ouR
COMPARATOR

BUS

tgfii%PRsOCESSOR

TRANSCEIVERS

{s arrsu
(1
)

TWO

PALs

]

| FOUR

§ COMPARATOR

}

8-BIT

XCVRs

| PALSs
-

FOUR

COMPARATOR
PALs

lRESU LTS
SEARCH |

SA01-SA13

CONTROL

AND

STATUS

TO RESULTS

PAO1-PA13

LKG-0772

Figure 4-11:

48-Bit Comparator

RO,

Figure 4-12 shows the structure of a 16-bit comparator. This illustration shows the circuits that
perform a bit-by-bit comparison of the four nibbles that make up a 16-bit portion of the 48-bit

address.

The results of each byte comparison are sent to the search control PALs, which determine the next
step.

4-28

LAN Bridge 100 Technical Manual
f
e

DATA

BUS

SA01-SA13
!

_J\

RAM

HIGH-

LOW-

COMPARE |

COMPARE

RESULTS

LOW-BYTE COMPARE BUS

HIGH-

—— NIBBLE

COMPARE

8K x 8

LOW-

——]NIBBLE

COMPARE

XCVR

Do08-D15

HIGH-BYTE COMPARE BUS

RESULTS
TO SEARCH
CONTROL AND
STATUS PALs
LKG-0773

Figure 4-12:

16-Bit Comparator

Technical Description

4-29

4.5.4.2

Address Comparator Circuit Descriptions —

the following circuits:
®*

Ethernet address table memory

Bus transceivers

Ethernet search address comparator PALSs

Binary search control and status PALs

Search status register

4.5.4.3

The address comparator is made up of

Ethernet Address Table Memory —

The Ethernet address table memory stores (in

ascending numerical order) Ethernet addresses. The addresses can be either or both of the follow-

ing:

Addresses of stations from which the LAN Bridge 100 unit has received packets

Addresses that have been down-line loaded by Remote Bridge Management Software (RBMS)

The memory is mapped as 8K by 48 bits and is made up of six 8K-by-8-bit RAM devices. The devices

are addressable by the processor or by the search control PALs.

Data (48-bit addresses) can be read out of address table memory for the following purposes:
®

The comparator PALs use addresses for performing address comparisons.

The comparator PALs temporarily store addresses during table maintenance operations.

Memory data may be used by RBMS or for Maintenance Operation Protocol (MOP) purposes.

4.5.4.4

Bus Transceivers —

The bus transceivers are bidirectional and allow data to be trans-

ferred between the data bus (D00-D15) and the internal compare bus.
In normal LAN Bridge 100 operation, Ethernet addresses are transferred from the data bus to the

compare bus. The addresses are latched into the comparator PALs for comparison against

addresses stored in the Ethernet address table.
Bidirectional transfer capabilities allow bridge management to read addresses from the Ethernet

address table.
The transfer direction is controlled by the processor BWRT control line and the SXCO0, SXC1, and
SXC2 control lines from the decoder.

4.5.4.5

Ethernet Search Address Comparators —

The comparator PALs perform bit-by-bit

comparisons between data received through the processor’s data bus and data accessed from the

Ethernet address table.

When the bus transceivers place the 48-bit Ethernet address onto the compare bus, each PAL
latches one nibble (4 bits) of the address. As the search control and status PALs probe the address

table, each comparator PAL compares its latched nibble against a corresponding nibble accessed

from the address table. A “‘greater than,” “‘less than,”” or ‘‘equal to’’ result is sent to the search con-

trol and status PALs.

The comparator PALs are also used during a 64-bit move for temporarily storing bits that make up

the Ethernet address (see Chapter 3 for more information on 64-bit moves).

4-30

LAN Bridge 100 Technical Manual

4.5.4.6

Search Control and Status PALs —

The search control and status PALs have two

functions. They control the binary search process once it is initiated by the processor and they
write the search status to the search results register.

When the search control PALs detect that the final (highest) byte of the 48-bit Ethernet address is
latched into the comparator PALs, they read the address that is on the search address bus (SA01-

SA13). This address, which is used for the first probe of the Ethernet address table, is also used by
the search control PALSs to calculate successive addresses to be probed during the binary search of
the address table.

The search control and status PALs deassert the DONE

signal. This has two effects.

It inhibits the decoder from allowing the processor to address any
of the TLU hardware.

It steers the multiplexer to select search addresses from the search control and status PALs.

The PALs continue the binary search process by monitoring the results (‘‘greater than’’ or ‘‘equal

to’’) from the address comparator PALs. Depending on the results, the PAL determines the next
step in the search process as follows:
e

When the results are ‘‘greater than,’’ the PAL equation calculates the next address to be probed.
The calculated address is placed onto the address bus.

When the results are ‘‘less than’’ (actually, not ‘‘greater than’’ and not ‘“‘equal to’’), the PAL
equation calculates the next address to be probed. The calculated address is placed onto the
address bus.

When the results are ‘‘equal to,”’ the PALs terminate the search and reassert the DONE signal.
The reasserted DONE signal enables the decoder to allow the processor to read the search

results register. It also steers the multiplexer to select the search address from the processor.

4.5.4.7

Search Results Register —

The search results register contains the address and status

of the last location probed by the search control and status PALs. The register data is latched in the
search control and status PALs and is addressed through a buffer on the processor data bus.

The 16 bits of register data are written onto the PAL address lines (PA0O0O-PA13) and PAL control
lines. The register is addressable by the processor only when the binary search is not in progress

(the DONE signal is deasserted). The 16 bits have the following definitions:

Table 4-3:

Search Results Register Bit Descriptions

Bit

Definition

<15>

Equal to (EQ) bit. Asserted when a search results in a found entry.

<14>

Not used.

<13:01>

Search address bits corresponding to the last address probed by the search PALs.
When accompanied by the EQ bit being set, the address indicates where the entry is located.

When EQ is not set, the address indicates where the entry should reside.
<00>

Greater than (GT) bit. Asserted when a search does not result in a found entry and the last

comparison found the search argument to be greater in magnitude than the table entry.

Technical Description

4-31

4.6

Power Supply

The power supply subsystem is a switching type, regulated ac-to-dc converter that uses a transformer in a half-wave transformer-coupled mode.
Power regulation is achieved by modulating the pulse width of the inverter’s primary current con-

duction time. An increase in the input power decreases the pulse width of the inverter’s primary
current conduction time to lower the output voltage. Conversely, a decrease in the input power
increases the pulse width of the inverter’s primary current conduction time to raise the output voltage.

The voltage input range is switch selectable and can be either of the following:
®

88Vacto 132 Vac (115 Vac nominal)

176 Vacto 264 Vac (230 Vac nominal)

The power supply provides the following regulated output voltages that are used by the LAN

I —

Bridge 100 circuitry:
Revision FO8 and below

Revision FO9 and above

+5 volts at 16 amps

+5 volts at 20 amps

+12 volts at 3 amps

+12 volts at 2.5 amps

—12 volts at 1 amp

—12 volts at 2.0 amps

Ground

AR,

[

ORSAGS

4-32

LAN Bridge 100 Technical Manual
5

5
Maintenance

5.1

Scope

The purpose of this chapter is to provide information on maintaining the LAN Bridge 100 unit. This
chapter includes sections on:

Maintenance philosophy

Preventive maintenance

(orrective maintenance

LAN Bridge 100 disassembly

5.2

Maintenance Philosophy

The maintenance philosophy for the LAN Bridge 100 unit is option swap. In other words, maintenance is concerned with determining whether or not a LAN Bridge 100 unit is faulty, and replacing
the LAN Bridge 100 unit when necessary.

Maintenance of the LAN Bridge 100 unit consists of corrective procedures only. Instructions for
replacing faulty LAN Bridge 100 units are provided as part of these procedures.

5.2.1

Required Equipment

The controlled distribution (CD) spares kit contains a LAN Bridge 100 unit and loopback connec-

tors. The loopback connectors allow off-line testing of the LAN Bridge 100 unit. Table 5-1
describes the CD kits that are available for local and remote LAN Bridge 100 units.

5-1

Table 5-1:

Controlled Distribution Kits for the LAN Bridge 100 Unit

Part Number

Description

A2-W0948-10

Local LAN Bridge 100 Spares Kit (120 V Version)
This kit contains:

A2-W0948-11

1LAN Bridge 100 (part no. DEBET-AA)

2loopback connectors (part no. 12-22196-01)

Local LAN Bridge 100 Spares Kit (240 V Version)
This kit contains:

A2-W1043-10

1LAN Bridge 100 (part no. DEBET-AB)

2]oopback connectors (part no. 12-22196-01)

Remote LAN Bridge 100 Spares Kit (120 V Version)
This kit contains:
O

A2-W1043-11

1 LAN Bridge 100 (part no. DEBET-RA)

1 loopback connector (part no.

1 fiber-optic loopback connector (part no. 29-25037-01)

12-22196-01)

Remote LAN Bridge 100 Spares Kit (240 V Version)
This kit contains:
P

1 LAN Bridge 100 (part no. DEBET-RB)

1 loopback connector (part no. 12-22196-01)

1 fiber-optic loopback connector (part no. 29-25037-01)
F

5.2.2

Optional Equipment

The H4080 test connector may be helpful in performing some corrective maintenance procedures.
Note that this test device is not supplied with the CD Kkit.

The H4080 test connector is a transceiver that is connected to a short length of coaxial cable. This
test connector can replace an on-line transceiver for off-line self-testing of the LAN Bridge 100 unit.

NOTE

The H4000-TA (or H4000-TB) Ethernet transceiver testers cannot be used to test
the data path through a LAN Bridge 100 unit because all H4000-Tx testers have the

same Ethernet address. This prevents the LAN Bridge 100 unit from forwarding the
test packets. However, testers may be used to test the links that normally connect to

the ports of the bridge.

5-2

LAN Bridge 100 Technical Manual
R

5.3

—

Preventive Maintenance

There are no preventive maintenance (PM) procedures for the LAN Bridge 100 unit. However, the
LAN Bridge 100 self-test should be run when network PM is performed. To run self-test, perform
the following steps:

1. Start the self-test either by interrupting power or by remote bridge management intervention.

2. Observe the status LEDs shown in Figure 5—1. The Self-test OK status LED should light within
10 seconds after self-test is started. If loop-detection conditions are not met, the On-line status

LED should light within 20 seconds after the Self-test OK status LED lights. If a failure is noted,

refer to Figure 5-2.

PORTAACTIVITY

PORT
B ACTIVITY

ON LINE l SELF-TEST OK

|
| 00000

"
&

DCOK

Figure 5-1:

Maintenance

MKV85-2776

LAN Bridge 100 Status LEDs

5-3

5.4

Corrective Maintenance

Corrective maintenance should be performed when there are indications that a LAN Bridge 100

unit is faulty. Corrective maintenance comprises verification and replacement of the faulty LAN
Bridge 100 unit. This is done by analyzing the results of self-test and isolating the LAN Bridge 100

SRR

unit from the network by using loopback connectors.

5.4.1

Troubleshooting Tips

The following tips on troubleshooting a LAN Bridge 100 unit are also referenced in the trouble-

shooting flowchart (Figure 5-2).
®

Understand the symptoms of the problem before starting to troubleshoot. This can help prevent misinterpretation of the symptoms.

Always look for the obvious problems first, such as the following:
Loose cable connections: Make sure that transceiver cables are locked in place with the slide

latches. Make sure that fiber-optic cables are not misthreaded, that they are screwed firmly
together, and that transmit and receive cables are not interchanged. (Cable connections are

illustrated in Chapter 2.)

Bridge switches set incorrectly (see Chapter 1).
GO,

Power cord not plugged in.
Incorrect power or no power at the electrical outlet.

Circuit breaker (located on the bridge’s I/O panel) tripped.
Note whether the fan is running. If it is not, power problems may exist.

Operator error.

If the self-test fails, it automatically executes again after 15 seconds. It is not necessary to cycle

bridge power.

Test communication through the bridge by attempting to communicate to a node on the other

side of the bridge. VMS systems use the SET HOST command to establish a logical link.

Consider possible environmental problems, such as power fluctuations, high ambient tempera-

ture, interference from other equipment, and so on.

When down-line load is enabled using the hardware switch or RBMS, ensure that a load host is

established for down-line loading the apropriate LAN Traffic Monitor software. For software

installation details, refer to the LAN Traffic Monitor Installation Guide.

ORI

5-4

LAN Bridge 100 Technical Manual
Ao

5.4.2

Fault Diagnosis

This section includes a troubleshooting flowchart (Figure 5-2) that is designed to isolate problems
and identify a faulty LAN Bridge 100 unit, associated transceivers, transceiver cables, or fiber-optic
cables.

It is possible that a fault exists in the transceiver or other equipment electrically close to the LAN

Bridge 100 unit. In some cases, this type of fault initially appears to be in the LAN Bridge 100 unit.
However, careful execution of the troubleshooting procedures in this section will either isolate the
fault to the LAN Bridge 100 unit or point to other possible sources of the malfunction.

When the troubleshooting flowchart leads to a malfunction in a transceiver or remote repeater, use

the appropriate documents and tools to troubleshoot those devices. These include the H4000
Transceiver Technical Manual (part no. EK-H4000-TM) and the Etbernet Repeater Technical
Manual (part no. EK-DEREP-TM).
Notes

1. After troubleshooting the LAN Bridge 100 unit, be sure to reconnect all cables
and to reset any switches to their correct positions.

When NVRAM failure occurs, the fault can be bypassed and the bridge can continue to operate. However, the bridge should be replaced as soon as possible.
e

Even when NVRAM has failed, RBMS parameters may be set over the network if S2 is placed in the down position. However, the bridge reply to
RBMS commands indicates partial success in setting parameters, since any

set parameters will be lost in the event of subsequent hard or soft resets
(including power-down periods). The parameters must be set each time a
reset occurs. As long as this condition exists, the Self-test OK status LED
blinks to indicate that NVRAM has failed.

Maintenance

Default parameters are used if S2 is placed in the down position. The Self-test
OK status LED is steadily lit to indicate a successful self-test.

5-5

(

START

—

)

r
"‘

CYCLE BR!DGE
AC POWEH

ALL

STATUS LEDs
ON THEN OFF
EXCEPT DC OK
REMAINS ON

DISCONNECT CABLE
O

FROM PORT B

TROUBLESHOOT
TRANSCEIVER AND |
| TRANSCEIVER CABLE
|

DC OK

STATUS LED

ON?

FOR PORT B

LOCAL

REMOTE

OR REMOTE
(FIBER-OPTIC)

REPLACE BRIDGE

—+®

RIS

BRIDGE?

DISCONNECT CABLE

FROM PORT A

DC OK

STATUS LED

| REPLACE BRIDGE

ON?

TROUBLESHOOT

TRANSCEIVER AND
TRANSCEIVER CABLE

FOR PORT A

LKG-0774-87
RO

Figure 5-2:

Troubleshooting Flowchart (Sheet 1 of 8)
S

LAN Bridge 100 Technical Manual

SELF-TEST OK

STATUS LEDON
STEADY? *

YES

NVRAM
RESET SWITCH (S2)

IN DOWN POSITION?

YES

| 3

PUSH S2 DOWN

'
SELF-TESTOK |

STATUS LED ON
STEADY? *

SELF-TEST
STATUS LED
BLINKING

YES

TRANSIENT NVRAM
FAILURE; BRIDGE OK |

NVRAM FAILURE-

vES | REPLACE BRIDGET |

* in 15 seconds

1 see section 5.4.2; Item 2 under “Notes.”

Figure 5-2:
Maintenance

LKG-0774-87

Troubleshooting Flowchart (Sheet 2 of 8)
5-7

BRIDGE

PROPERLY

PROPERLY CONNECT

BRIDGE TO LANs t

LOCAL
OR REMOTE
BRIDGE?

LOCAL

CONNECT LOOPBACK
CONNECTOR
TOPORTSA&B
A

REMOTE

| CONNECT FIBER-OPTIC
LOOPBACK

CONNECTOR
TOPORTB

l’

CYCLE BRIDGE AC
POWER

S
* After 15 seconds

1 Referto Troubleshooting Tips (Section 5.4.1).
LKG-0774-87

ORI

Figure 5-2:

5-8

Troubleshooting Flowchart (Sheet 3 of 8)

LAN Bridge 100 Technical Manual

ON—LINE

STATUS LED

NJES |

2 MINUTES

ON WTHN

SET HosT'

1
|}

| NO

OWN

LINE-LOAD
—L

ENABLED? 2

BRIDGE

IN LOOP
CONFIGURATION?

YES

REPLACE BRIDGE |
ON—LINE
STATUS LED

YES
STATUS LED
FLASHING?

BLINKING? °

REPLACE BRIDGE

+ NO

DOWN—LINE LOAD
IN PROGRESS *#

[

LTM SOFTWARE
LOADED

1 FROM HOST ON PORT A TO HOST ON PORT B

2 DOWN—-LINE SWMITCH MAY BE ENABLED BY
SWITCH 5 BEING SET OR VIA A
REMOTE COMMAND BY RBMS. SEE

LAN BRIDGE 100 HARDWARE INSTALLATION/
OWNER'S GUIDE.

INSTALLATION
VERIFIED

3 FLASHING TWICE EVERY 2 SECONDS INDICATES

THAT THE LOAD HOST SUCCESSFULLY DOWN-—LINE

LOADED THE LTM LISTENER SOFTWARE IMAGE.

FLASHING ONCE EACH SECOND INDICATES THAT
THE LOAD HOST HAS STARTED THE LTM

LISTENER SOFTWARE.

4 CHECK THAT THE DOWN-LINE LOAD

HOST HAS BEEN SET UP. SEE
LAN TRAFFIC MONITOR INSTALLATION GUIDE
FOR DETAILS ON SETTING UP A LOAD HOST.
LKG~0774~87

Figure 5-2:

Maintenance

Troubleshooting Flowchart (Sheet 4 of 8)

5-9

SELF-TEST OK

REPLACE BRIDGE

STATUS LED
ON? *

—

LOCAL
Oggg‘gg‘f
,

REMOTE |

TROUBLESHOOT

TRANSCEIVER AND
TRANSCEIVER CABLE
FOR PORT B

REMOVE LOOPBACK

CONNECTOR FROM
PORT A AND

RECONNECT CABLE

SELF-TEST OK

STATUS LED
ON?

TROUBLESHOOT

TRANSCEIVER AND
TRANSCEIVER CABLE

FOR PORT A

TROUBLESHOOT

TRANSCEIVER AND
TRANSCEIVER CABLE

FORPORTB

*AFTER 15 SECONDS

LKG-0774-87

Figure 5-2:

Troubleshooting Flowchart (Sheet 5 of 8)
T

5-10

LAN Bridge 100 Technical Manual

HAS

BRIDGE BEEN
REPLACED?

‘ NO

REPLACE BRIDGE

LocaL |

TROUBLESHOOT

| vES

LOCAL

DATALINKS TO
RESPECTIVE
NETWORK

OR REMOTE
BRIDGE?

| REMOTE

SECOND
TIME THROUGH
THIS FLOWCHART?

BRIDGETO-BRIDGE

BRIDGE-TOBRIDGE

OR BRIDGE-TOREPEATER

CONFIGURATION?

ESCALATE

| BRIDGE-TO-REPEATER

REPEATER

SELF-TEST

STATUS LED
ON?

TROUBLESHQOT
REPEATER

YES

LKG-0774-87

Figure 5-2:
Maintenance

Troubleshooting Flowchart (Sheet 6 of 8)
5-11

Ll
.

1. CONNECT LOOPBACK

CABLE
TO BRIDGE PORT A
2. SET LOOPING

SELF-TEST ON

—

3. CYCLE BRIDGE AC

POWER

SELF-TEST OK
STATUS LED
ON*

IS THIS
1ST BRIDGE
OF PAIR?

REPLACE BRIDGE

REPEAT FOR

H 1

2ND BRIDGE OF PAIR |

)

o
ST

TROUBLESHOOT
FIBER-OPTIC

CABLE

*AFTER 15 SECONDS

LKG-0774-87
RO

Figure 5-2:

Troubleshooting Flowchart (Sheet 7 of 8)
A

5-12

LAN Bridge 100 Technical Manual

PRESS REPEATER

| TST SWITCH (LOCATED |
|

ONREPEATER)

REPEATER

TROUBLESHOOT

TST STATUS LED
ON?

DATA LINES TO
|RESPECTIVE NETWORKS

SECOND
TIME THROUGH
THIS FLOWCHART

CONNECT LOOPBACK
CABLE TO REPEATER
FIBER-OPTIC

INTERFACE
\

PRESS REPEATER TST

ESCALATE

1.CONNECT LOOPBACK

SWITCH

CABLE
TO BRIDGE PORT A
2.SET LOOPING
SELF-TEST ON
3.CYCLE BRIDGE AC

REPEATER
TST
STATUS LED

ON?

YES

POWER

‘

BRIDGE

SELF-TEST OK
STATUS LED
ON*

REPLACE BRIDGE

}-—>D

TROUBLESHOOT
FIBER-OPTIC CABLE

*AFTER 15 SECONDS

LKG-0774

Figure 5-2:

Maintenance

Troubleshooting Flowchart (Sheet 8 of 8)

5-13

5.5

LAN Bridge 100 Replacement Procedures

Use the following procedures to replace a LAN Bridge 100 unit.
1.

Unplug the power cord from the wall outlet and from the defective LAN Bridge 100 unit.

Disconnect the transceiver cables and/or the fiber-optic cable.

Test the replacement LAN Bridge 100 unit by installing loopback connectors and connecting

ORORE I

power to the LAN Bridge 100 unit.
R

If the LAN Bridge 100 unit is rack mounted or wall mounted, go to step 5. If the LAN Bridge 100
unit is a desk-top unit, replace it with another LAN Bridge 100 unit and go to step 8.

Support the LAN Bridge 100 unit and remove the screws holding the mounting brackets to the
rack or wall.

Remove the brackets from the LAN Bridge 100 unit.

Install a replacement LAN Bridge 100 unit in the rack or on the wall according to the instructions in Chapter 2.

Connect the transceiver cables and/or fiber-optic cable to the replacement LAN Bridge 100 unit
according to the instructions in Chapter 2.

5-14

LAN Bridge 100 Technical Manual

5.6

Bridge Disassembly

This section describes the disassembly procedure for the LAN Bridge 100 unit. Removal procedures for the following parts are provided:
e

Plastic enclosure

(Chassis cover

Fiber-optic interface

Power supply

Logic board

WARNING

To prevent electrical shock and damage to components, disconnect the power cord

a
e

from the LAN Bridge 100 unit before opening the chassis.

The instructions in this section assume that:

All external cables to the bridge have been removed. (Cables should be marked for proper

The LAN Bridge 100 unit has been removed from its rack or wall mounting.

replacement.)

Maintenance

5-15

5.6.1

Plastic Enclosure Removal

To remove the plastic enclosure (see Figures 5-3 and 5-4), follow these steps:
1.

Remove the 4 screws (see Figure 5-3) that secure the rubber feet to the bottom
of the plastic
enclosure. These screws release the top plastic of the enclosure and the two side
pieces.

Remove the 4 screws (see Figure 5-4) that hold the bottom of the enclosure to

the chassis.

TR~y
T
—

o«

LKG~-0775
S

Figure 5-3:

5-16

Removing the Top Plastic Enclosure and Side Pieces

LAN Bridge 100 Technical Manual

BRIDGE

BOTTOM
PLASTIC
COVER

e i————

LKG-0776

Figure 5-4:

Maintenance

Removing the Bottom Plastic Enclosure

5-17

5.6.2

Removing the Chassis Cover

To remove the chassis cover for Revision FO8 and below, see Figure 5-5 and follow these steps.

NOTE
To identify the revision level on the unit, the label that contains the Part Number
(PN) and Serial Number (SN) also lists the revision level as Rev. Older revision

units are labeled as Rev FO8 and below. The newer units are designated as Rev FO9

and above.

RIS

Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).

Remove 24 chassis cover screws as follows (see Figure 5-5):

Remove 10 screws from the top of the chassis.

Remove 14 screws (7 on each end of the chassis).

Lift off the chassis cover (see Figure 5-5).

ORI

To remove the chassis cover for Revision FO9 and above, see Figure 5-6 and follow these steps:
1.

Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).

Remove 14 chassis cover screws as follows (see Figure 5-6):
R

Remove 2 screws from the top of the chassis.

Remove 12 screws (6 on each end of the chassis).

Lift off the chassis cover (see Figure 5-6).
IO

5-18

LAN Bridge 100 Technical Manual

LRG-D7TT

Figure 5-5:

Maintenance

Removing the Chassis Cover for Revision FO8 and Below

5-19

ehterta
el P

Btote

iy.fl..n.fl,.«mw.fl.% .
i

xXAeA«nr, om‘HflOMuWw\“mL.wtWu&U‘t.uet%3405

,0.
A\ Set i
e

ALK

v.Mw"TRty STo t»%«.Wum.ua

iy et Ll

GROUND WIRE

NLR

~16-PIN POWER SUPPLY

LAtty

ietl Lt

M. Nw.

4 PIN POWER SUPPLY"

STANDOFF

CLIP
L Q i o @ o— Q

INTERFACE MODULE

MKV88-1686

Figure 5-6

Various Removal Procedures for Revision F09 and Above
T

—

5-20

LAN Bridge 100 Technical Manual
AR

5.6.3

Removing the Fiber-Optic Interface

To remove the fiber-optic interface, follow these steps:

1. Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).
2.

Remove the chassis cover (refer to Section 5.6.2).

3. Remove the 2 screws (see Figure 5-7) that hold the interface bracket to the LAN Bridge 100
chassis.

=
-

4. While releasing the standoff clips that hold the interface board in place (see Figure 5-7),

gently remove the interface board and bracket from its connector on the logic board, as shown
in Figure 5-8.

NOTE

A fiber-optic interface is present only in a remote LAN Bridge 100 unit.

FIBER-OPTIC INTERFACE

REMOVE 2

SCREWS

INTERFACE

BRACKET

STANDOFF
CLUIPS

RELEASE

s
Lo

—

LAN BRIDGE 100 WITH
CHASSIS COVER REMOVED

Figure 5-7:

Fiber-Optic Interface Retaining Hardware for Revision FO8 and Below
NOTE

See Figure 5-6 for fiber-optic interface retaining hardware for Revision FO9 and
above.

Maintenance

5-21

NOTE
Figure 5-8 shows the angle at which the fiber-optic interface should be removed

from the LAN Bridge 100 unit.

INTERFACE
CONNECTOR

FIBER-OPTIC
INTERFACE

LAN BRIDGE 100 WITH

CHASSIS COVER REMOVED
R

LKG-1228-87

Figure 5-8:

Removing the Fiber-Optic Interface for Revision FO8 and Below

NOTE

RS0

See Figure 5-6 for removing the fiber-optic interface for Revision FO9 and above.

5-22

LAN Bridge 100 Technical Manual

5.6.4

Removing the Power Supply

To remove the power supply, follow these steps.

Power supply removal procedure for Revision FO8 and below:
1.

Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).

Remove the chassis cover (refer to Section 5.6.2).

3. Unplug the 3 connectors (A) shown in Figure 5-9.

Remove the 2 retaining screws (B) that secure the power supply bracket to the LAN Bridge 100

I/O panel.
W

Remove the fan power harness from the plastic retaining clips (A) shown in Figure 5-10.
Loosen the 2 captive screws (B) on the power supply bracket until they pop up.

Carefully lift the power supply assembly out of the chassis. Figure 5-12 shows the power
supply assembly after it is removed from the chassis.

Power supply removal procedure for Revision FO9 and above:
1.

Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).

Remove the chassis cover (refer to Section 5.6.2).

Remove the 6 screws that secure the power supply module to the top chassis cover. See Figure
5-11.

Remove the fan power harness from the power supply module.

Remove the Ground wire from the top chassis cover by removing the retainer nut.

Carefully lift the power supply assembly out of the chassis. Figure 5—-13 shows the power
supply assembly after it is removed from the chassis.

Maintenance

5-23

UNPLUG
CONNECTOR (A)
REMOVE RETAINING

UNPLUG

SCREWS (B)

CONNECTOR (A)

UNPLUG
,

——

(A)

LAN BRIDGE 100 WITH
CHASSIS COVER REMOVED

LKG~1227-87

Figure 5-9:

Removing the Power Supply Connectors and Retaining Screws for Revision FO8
and Below

NOTE
See Figure 5-6 for removing the power supply connectors and retaining screws for
Revision FO9 and above.

T —

SR8

5-24

LAN Bridge 100 Technical Manual

REMOVE FAN POWER
HARNESS FROM

RETAINING CLIPS (A)

FAN

AN POWE

POWER

HARNESS

LOOSEN CAPTIVE

SCREWS (B)

1
i

i
4

LAN BRIDGE 100 WITH

p—

CHASSIS COVER REMOVED
LKG-0780

Figure 5-10:

Maintenance

Removing the Power Supply for Revision F08 and Below

5-25

FrTm—

REMOVE SCREWS

(6 PLACES)

POWER SUPPLY

FAN POWER
HARNESS
AR

TOP CHASSIS COVER
D

MKV88-1688

Figure 5-11:

Removing the Power Supply for Revision F09 and Above

SRR

5-26

LAN Bridge 100 Technical Manual

SUPPLY

LKG-0781

pre—
Figure 5-12:

Power Supply Assembly Removed from the Chassis (Revision FO8 and Below)

Maintenance

5-27

POWER SUPPLY

H7859-A

MKV88-1687

Figure 5-13:

5-28

Power Supply Assembly Removed from the Chassis (Revision FO9 and Above)

LAN Bridge 100 Technical Manual

5.6.5

Removing the Logic Module

To remove the logic module (see Figure 5-14), follow these steps:
1.

Remove the top and bottom plastic enclosures and side pieces (refer to Section 5.6.1).

Remove the chassis cover (refer to Section 5.6.2). Remove the power supply (refer to Section
5.6.4).

If necessary, remove the fiber-optic interface (refer to Section 5.6.3).
. Unplug the remaining interface connector (A) shown in Figure 5-14. To remove the interface
connectors completely, loosen the captive screw (see Figure 5-14, inset) on each connector
bracket until it pops up. Lift each interface connector out of the chassis.
NOTE

For Revision FO9 and above, the interface connector next to the ac power area is fixmounted and, therefore, cannot be removed.

. Remove the retaining screw (B) that holds the logic module to the LAN Bridge 100 chassis.

While releasing the standoff clips (C) that hold the logic module in place, carefully lift the
logic module out of the chassis (see Figure 5-14).

Maintenance

5-29

REMOVING THE

INTERFACE

CONNECTOR

CAPTIVE
SCREW

—

CONNECTOR
CONNECTOR

CABLE

UNPLUG INTERFACE

REMOVE RETAINING

CONNECTOR (A)

SCREW (B)

LOGIC
MODULE

S
*I

‘<.

LAN BRIDGE 100 WITH

POWER SUPPLY REMOVED

RELEASE STANDOFF

CLIPS (C)
LKG-0782

Figure 5-14:

Locations of Logic Module Retaining Hardware

5-30

LAN Bridge 100 Technical Manual

A
Glossary

Baseband Ethernet
An unmodulated, single-channel, CSMA/CD network. One station at a time can be transmitting
information on the medium without disruption. (See also CSMA/CD.)

Bridge physical address
A unique, 48-bit address that distinguishes a bridge from all other addressable network devices
(including other bridges).

Broadband Ethernet
A single CSMA/CD channel, modulated to occupy a frequency bandwidth channel on a broadband coaxial medium.

CSMA/CD
Carrier-sense, multiple-access with collision detection. This is the generic term for the class of
link management procedure used by Ethernet.

Cost
See Path Cost.

Data link
A logical connection between two stations on the same circuit.

Data Link layer
The higher of the two lowest layers in the International Standards Organization model for open
systems interconnection. The Data Link layer implements a medium-independent link-level

communication facility on top of the physical channel provided by the Physical layer.

Descriptor ring
A circular queue of tasks that the LANCE and microprocessor use to point either to buffer space

for storing received packets or to packets in packet memory that are to be transmitted.
SRS,

Designated bridge
The bridge on a local area network that has the lowest path cost to the root or the lowest ID if

more than one bridge has the lowest path cost. (See also Path cost.) The ID consists of the physi-

cal address (which is fixed) and a priority prefix (which can be set using RBMS).

Destination address
A field in an Ethernet packet that consists of the first 6 bytes (48 bits) of the packet once the preamble and start bit have been removed. This field identifies the destination of the packet.

Down-line load
The process of sending a software image from a load host to the bridge.

Extended LAN
See Extended local area network.

Extended local area network

A LAN whose capabilities, such as length and number of stations, are extended beyond normal
Ethernet

IEEE 802.3 LAN limitations through the use of certain devices such as the

ARSI

LAN Bridge 100 unit.

Fiber-optic link
A link between devices whose transmission medium is made up of optical fibers.

Filtering

The LAN Bridge 100 decision not to forward a packet from one side of a bridge to the other
side.

Forwarding
The LAN Bridge 100 decision to pass a packet from one side of a bridge to the other side.

Hello message
A special message transmitted among bridges in an extended LAN and used for loop detection

purposes. The message is transmitted as part of the spanning tree algorithm.
SR

IEEE 802.3
A standard of the Institute of Electrical and Electronic Engineers that defines the CSMA/CD
Media Access Method and Physical Layer Specifications of a network.

LAN
See Local area network.

A-2

LAN Bridge 100 Technical Manual
R

LAN backbone

The central segment in a LAN configuration to which multiple segments can be connected using
bridges. A typical use for a configuration with a LAN backbone might be in a multistory office
building. A vertical backbone could connect to segments that serve individual floors in the
building.

LAN Traffic Monitor (LTM)

An optional mode of LAN Bridge 100 operation that provides data about network traffic. A load
host down-line loads the LTM software image to the LAN Bridge 100 unit.

LANCE

Local Area Network Controller for Ethernet. A VLSI (very-large-scale-integration) device that

provides data link functions for Ethernet controllers.
Local area network

A network in which communications are limited to a moderately sized geographic area, such as
a single office building, a warehouse, or a campus. The maximum length of an Ethernet LAN is
2800 meters (9184 feet). A LAN depends upon a communications medium of moderate to high
data rate (10 megabits/second for Ethernet) and normally operates with a consistently low error
rate.

Local bridge

A bridge that can connect LANs separated by not more than 100 meters. The distance from the

local bridge to either LAN cannot exceed the maximum allowable transceiver-cable length of
50 meters (164 feet).

Loop detection

An implementation of a spanning tree algorithm that ensures a single data path between any
two stations on an extended LAN.

LTM

See LAN Traffic Monitor

MOP

Maintenance Operation Protocol. A maintenance protocol that is specified in the Digital
Network Architecture (DNA) Low-Level Maintenance Operations Architectural Specification.
Multicast message

A message that is addressed to a group of logically related stations.
NI subsystem

Network interconnect subsystem. The subsystem in a LAN Bridge 100 unit that receives and
stores packets from and transmits packets to LANs that are connected to the bridge.

Glossary

A-3

Node
See Station.

Packet
The basic Ethernet network message unit that is made up of a preamble and data stream. The
data stream is preceded by a preamble and includes the destination address and source address
fields. The minimum packet size is 512 bits (64 octets). The maximum packet-size length is

12,144 bits (1518 octets). Packets must be preceded by the interpacket gap of at least 9.6 micro-

seconds.

Path cost
The sum of the circuit costs along a path between two stations. A circuit cost is a positive inte-

ger value associated with use of a given circuit.

Physical address

The unique, 48-bit address value associated with a given station on the network. An Ethernet
physical address is defined to be distinct from all other physical addresses on all Ethernet networks.

Physical layer
The lower of the two lowest layers in the International Standards Organization model for open
systems interconnection. The Physical layer is implemented by the physical channel using a

specified medium. The Physical layer insulates the Data Link layer from medium-dependent

physical characteristics such as baseband, broadband, or fiber-optic technologies.

Processor subsystem
The subsystem in a LAN Bridge 100 unit that controls overall bridge operation.

RBMS
Remote Bridge Management Software. An optional, VMS-layered software package residing on

a station that is used to monitor and control LAN Bridge 100 units in an extended LAN.

Remote bridge

A LAN Bridge 100 unit that also converts electrical signals to optical signals that are suitable for
transmission over a fiber-optic cable. Two remote bridges and a fiber-optic cable can be used to

connect LANs separated by more than 100 meters (328 feet). The maximum length of the fiber-

optic cable is limited by its attenuation characteristics. The distance from either remote bridge
to its associated LAN cannot exceed the maximum allowable transceiver-cable length of 50

meters (164 feet).

Repeater
OIS

A device used in a LAN to extend the length, topology, or interconnectivity of the transmission

medium beyond that imposed by a single transmission segment. A segment can be up to 500
meters (1640 feet) in length.

A-4

LAN Bridge 100 Technical Manual
A

Root

The bridge in an extended LAN that has the lowest ID. The ID consists of the physical address
(which is fixed) and the root priority prefix (which can be set using RBMS).

SIA
Serial interface adapter. A VLSI (very-large-scale-integration) device that provides the electrical
interface between a LANCE and a transceiver cable. (See also LANCE.)

Source address

A field in an Ethernet packet that encompasses the second 6 bytes (48 bits) of the packet once
the preamble and start bit have been removed. This field identifies the source of the packet.
Spanning tree algorithm
An algorithm used to ensure a single data path between any two stations on an extended LAN.

Station
A single addressable site on an Ethernet network, generally implemented as a computer and

appropriate peripherals and connected to the network through a controller and transceiver.
Also called a node.

Station address
See Physical address.

TLU
See Table lookup subsystem.

Table lookup subsystem

The subsystem in a LAN Bridge 100 unit that performs address comparisons. The results of
comparisons are used by the processor to help determine whether packets should be forwarded
or filtered.

Transceiver
The portion of the Physical layer implementation that connects directly to the coaxial cable

and provides both the electronics that send and receive the encoded signals on the cable and
the required electrical isolation.

Transceiver cable
A four-pair, shielded cable used to connect a transceiver to the LAN Bridge 100 unit.

Troll feature
A feature in a LAN Bridge 100 unit that is controlled by RBMS. This feature can prevent stations
on one side of the bridge from sending packets to some stations on the opposite side of the
bridge. (See also RBMS.)

Glossary

A-5

PO,

BRI

SIS

B
Fiber-Optic Link Analysis

B.1

Introduction

This appendix discusses the general tradeoffs in a fiber-optic system and identifies and defines the

characteristics of the components in such a system. Since a system is composed of many independent entities, it is possible to make a system work while violating one or more of the recommendations. Digital Equipment Corporation assumes no responsibility for providing engineering
assistance if failure occurs in such cases. If the guidelines are followed properly, the system will
perform to the engineering specifications. If the installer or consultant wishes to ignore the

published limits and use actual measurements in forming a fiber-optic system, he or she does so
with the understanding that the system may work marginally, may fail prematurely, or may fail
entirely.

NOTE

Guidelines should be used to aid in determining the feasibility of a system. Following the guidelines minimizes risk.

The material provided in this appendix assumes that the reader is familiar with fiber-optic

technology and has been trained in proper fiber-optic installation and maintenance techniques.
A clear understanding of the guidelines included in this appendix is essential for successful link
installation, especially where many variables are involved.

WARNING
Be especially careful when working with fiber-optic cables and devices.
When exposed, optical fibers (contained within the fiber-optic cable) are
extremely brittle and fragments from the fiber can easily penetrate the skin or eyes.
Wear protective goggles and clothing when working with the optical fibers.
Some fiber-optic equipment may emit laser light that can injure your eyes. Never
look into an optical fiber.

B-1

B.2

System Conventions

This section describes the conventions used to develop the data included in this appendix.

B.2.1

Absolute and Relative Measurements

Absolute values of physical quantities in a fiber-optic system require a laboratory environment and
sophisticated laboratory techniques. It is assumed that the system under consideration has been
thoroughly characterized, that parameters have been identified and specified, and that the sensitivities are understood.

Relative measurements such as end-to-end cable loss or connector loss can be done in the field.
Since losses are relative in nature, the decibel (dB) is the common unit of measure rather than an

absolute value such as the microwatt. The system designer must use the allowable system loss
(relative measurements) as a guideline for system implementation.
DGR

NOTE
The actual power levels of emitted light and detected light in Ethernet devices are

not essential in the planning and prediction of a fiber-optic network interconnect.

The key measurement is the allowable system loss.
A5

You should assume that the product is a certified product that meets the minimum specifications.
The equation shown in Figure B-1 is the basis for the design engineer’s system limits. The key

parameter is the allowable loss, not the absolute values of light intensity.

ENGINEERING DETERMINATION

i«——- NETWORK OWNER’S——P-‘
CONCERN

MANUFACTURING ASSURANCE

MINIMUM

LAUNCH
POWER

—

MINIMUM

RECEIVER

SENSITIVITY

MAXIMUM

ALLOWABLE
LOSSES

INSTALLATION GUIDE

CABLE VENDOR GUARANTEES
INSTALLER GUARANTEES
ON-SITE MEASUREMENTS

LKG-0783

Figure B-1:

Fiber-Optic System Responsibilities
e

B-2

LAN Bridge 100 Technical Manual

—

In fiber-optic systems characterized by Digital Equipment Corporation, the range of absolute
values of light has been determined by engineering and is the amount of light in the fiber just
after the launch connector specified as the amount of light in the fiber just after the launch
connector (see Section B.3.1).

Light loss through the last connector of the link is negligible since fiber-to-receiver coupling
losses are minimal; unless the connector is damaged or the fiber end is marred, smudged, or dirty

(see Section B.3.3). Therefore, the connectors on both ends of the link should not be considered
in any loss calculations. They have been accounted for in the issuance of the allowable system loss

number. Figure B-2 identifies the portion of the fiber system where allowable loss can occur.

AMOUNT OF LIGHT IN
THE FIBER AFTER THE

AMOUNT OF LIGHT
HITTING THE PIN

FIRST CONNECTOR

RECEIVER DIODE

¢—— ALLOWABLE SYSTEM LOSS

l TRANSMIT ij

(11

(113

5:1 RECEIVE l

NOTE: [I_I] REPRESENTS INTERMEDIATE BARREL CONNECTIONS
10 REPRESENTS CONNECTOR TO DIGITAL DEVICE
LKG-1226-87

Figure B-2:

B.2.2

Area of Network Consultant’'s Concern — Allowable System Loss

Worst-Case Definition

Parameters specified in this appendix represent the worst case unless otherwise specified. That is,

all parameters have been characterized and the point specified as worst case is 2-sigma (two
standard deviations from the normal data distribution). 2-sigma implies that 98% of the values

obtained were better than the given worst-case number.

B.3

Coupling Considerations

There are three types of optical coupling in fiber-optic systems. Each type has a different coupling
mechanism; therefore, coupling losses that are specified for each type cannot be interchanged.
The three types of optical coupling are as shown in Table B-1.

Fiber-Optic Link Analysis

B-3

[

Table B-1:

Types of Optical Coupling

Type of Coupling

Measurement Expertise

LED (emitter)-to-fiber

Laboratory

Fiber-to-fiber

Field

Fiber-to-pin diode (receiver)

Laboratory

IMPORTANT
It cannot be overemphasized that the quality of the polished ends of the fiber is the
major cause of connector loss. Connector loss is significantly increased by dirt,

scratches, finger smudges, and so on. Keep the connector ends clean.

B.3.1

LED (Emitter)-to-Fiber

Emitter-to-fiber measurements are confined to the laboratory environment. There is no field

method established for determining quantitative data regarding the LED-to-fiber coupling. When
using a characterized fiber type, the engineering and manufacturing groups have ensured that

minimum coupling specifications are met.
The LED emits a spherical shaped pattern (see Figure B-3) that may vary from the form r(cos@)
to rr[(cos(@)* ]. The value of m is typically from 25 to 50 in the case of a microlensed LED. In the
case of a laser, m can be as high as 200. The coupling depends on mechanical alignment of the

actual LED die and the core of the fiber, the emission and acceptance angle of the LED and fiber,

mmm S

and the differences in refractive indexes of the materials between the die and the fiber.

LOST POWER

/
|

& /7,//////////

CLADDING

IB&§&w.—

»w

/--m-........

/4,,;////////////

FIBER

ACCEPTANCE

CORE

ANGLE

CLADDING
%MM

SOURCE RADIATION
PATTERN
LKG-0785

Figure B-3:

B-4

LED Coupling into a Fiber

LAN Bridge 100 Technical Manual

B.3.1.1 LED-to-Fiber Coupling Derating Factors — Tables B-2 and B-3 identify the Allowable
System Loss that can be assumed for those systems that use DIGITAL fiber-optic daughterboard
54-16053-01 and stainless steel Amphenol 906 type SMA connectors. The 54-16053-01 daughterboard has two revisions that give differing power budgets. Be sure to check the variation of the
product and follow the appropriate table based on the variation you are configuring.
The amount of Allowable System Loss must be compared to the sum of the fiber link losses in order
to determine whether or not the link will function. The losses that must be considered are listed
below. Methods for calculating system budgets are outlined in Section B.8.
Cable Attenuation
Splice Loss

dB/km X cable length
dB/each X # of splices

Margin

1.5 dB for system margin

Connector Loss
Bandwidth Derating

dB/connector X # of additional connectors
0.5 dB/km X cable length

NOTE

The 1 dB for LED aging and .7 dB for temperature-caused variations are already
included as part of the fixed losses and do not have to be included when making
budget calculations.

Table B-2: LED-to-Fiber Coupling Derating Factors for -RC/-RD Variations
(Using 54-16053 Revision A2 or lower)
Allowable

System Loss

Fiber Type

Dimensions

Derating Factor

Corning 1508

100/140 pm

0 dB

12.5 dB

Corning 1519

85/125 um

—2.4 dB

10.1 dB

AT&T Multimode

62.5/125 um

—4.2 dB

8.3 dB

Corning 1509

62.5/125 um

—4.2 dB

8.3 dB

Corning 1517

50/125 um

—9.4 dB’

3.1 dB

AT&T Multimode

50/125 pm

—-9.4 dB’

3.1 dB

*The greater loss incurred by these cable types results from a smaller numerical aperture and a smaller core
diameter.

Fiber-Optic Link Analysis

B-5

Table B-3:

LED-to-Fiber Coupling Derating Factors for -RH/-RJ Variations

(Using 54-16053 Revision B2 or higher)
Allowable

Fiber Type

Dimensions

Derating Factor

System Loss

Corning 1508

100/140 um

0 dB

16.0 dB**

Corning 1519

85/125 um

—2.5 dB

13.5 dB

AT&T Multimode

62.5/125 um

—4.0 dB

12.0 dB

Corning 1509

62.5/125 um

—4.0 dB

12.0 dB

Corning 1517

50/125 um

—8.0 dB’

8.0 dB

AT&T Multimode

50/125 um

—8.0 dB*

8.0 dB

" The greater loss incurred by these cable types results from a smaller numerical aperture and a smaller core

diameter.

"'DEBET and DEREP (-RH/-R]J versions ONLY) links using 100/140 micron fiber, must incur a minimum of

3 dB loss. Refer to section B.9 for use of the fiber-optic attenuator if your system does not have at least 3 dB

worth of loss.

B.3.2

Fiber-to-Fiber Coupling

Fiber-to-fiber coupling occurs in two instances: a splice or a bulkhead/barrel connection. Measurement of fiber attenuation, splice loss, and fiber-to-fiber barrel connections are the only mea-

surements that can be done in the field.
IR

NOTE
If possible, do not mix fiber types or sizes in a system. In a system with mixed fiber
sizes, all fibers must be considered to act as the smallest fiber size in the system.
Losses occur in splices and barrel connections because of mechanical misalignment, mismatch of

core diameters, mismatch of refractive index profile, and mismatch of numerical aperture. Statisti-

cal data is available for most of these parameters, but worst-case numbers should be used in
predicting performance of a system that is going to be installed.

SRR

NOTE
It is extremely difficult to measure attenuation in a system with mixed fiber
diameters. Light traveling from a small fiber to a larger one can yield erroneous test
readings, such as apparent light increases instead of attenuations (decreases).

B-6

LAN Bridge 100 Technical Manual

Splice and Barrel Connector Losses — In typical fiber-optic interconnect systems,

B.3.2.1

there may be several kinds of cables (using the same fiber type), several sections of cables, and
usually some type of a junction box strategy. This implies that there are barrel connections and

splices in the fiber-optic cable.

Table B—4 shows the loss in decibels for the connectors and splices that have either been characterized by engineering or are available and can be characterized by the installer.
IMPORTANT

The values in the following table are results that can be achieved only by a person
who is trained and experienced in using proper tools and correct procedures. The
values were obtained under laboratory conditions.

Table B-4:

Splice and Barrel Connector Losses
Typical Loss

Worst Case
Loss in dB

100/140 Amphenol type 906 (stainless steel)

0.8

1.5

50/125 Amphenol type 906 (stainless steel)

1.5

Note 2

62.5/125, 85/125 Amphenol type 906

1.5

Note 2

50/125 AT&T ST type

1.0

1.5

62.5/125 AT&T ST type

0.8

1.2

85/125 AT&T ST type

Note 2

100/140 AT&T ST type

Note 2

Elastomeric

0.1

0.2 (Note 3)

Capillary tube

0.1

0.2 (Note 3)

Fusion

.05

0.1 (Note 3)

Connector or Splice Type

in dB

(Note 1)

Barrel Connectlors

Splices

NOTES:

1. Assumes that the barrel connector is close to the LED source. This is the worst-loss condition.
2. Consult vendor information.

3. Loss assumes the same lot (reel) of fiber is used. Splicing different lots of fiber yields results based on the
fiber differences rather than on the mechanics of the splice.

Fiber-Optic Link Analysis

B-7

B.3.2.2

Guidelines for Connectors and Splices on Ethernet Products — This section provides

guidelines that should be followed when connecting or splicing fiber-optic cable for use with

Ethernet products.

All fibers must be terminated in an Amphenol type 906 stainless steel connector for attachment of the cable to the Ethernet device.

For 100/140 fiber systems, splices or barrel connections can be used. Use Amphenol type 906

stainless steel barrel connectors and follow the appropriate connection procedures for the
connectors and alignment sleeves.

For 62.5/125 or 85/125 fiber systems, splicing is preferred for intermediate connections, but
Amphenol type 906 or AT&T type ST connectors may be used for these connections.
For 50/125 fiber systems, all intermediate connections should be spliced.

Be aware of the types of misalignment or separation that are illustrated in Figure B—4.

NOTE
ARSI

No guarantee of a specified attenuation can be made for splicing or barrel connecting fibers with different core diameters.

The ideal treatment of loss from one fiber size to another is represented by the following equations. These generally do not apply to real-world situations since the loss depends on the spot size

of the light in the core, not on the core diameter.
Ideal cases:
S

Loss due to diameter mismatch:
20 log

diameter 1

diameter 2

where diameter is the diameter of the fiber core.
Loss due to mismatch of numerical aperture:

20 log

NA 1

RS,

NA 2

Where NA is the numerical aperture of the fiber.

B-8

LAN Bridge 100 Technical Manual
D

LATERAL MISALIGNMENT

END SEPARATION

ANGULAR MISALIGNMENT
LKG~0786

Figure B-4:

Types of Fiber Misalignment

Fiber-Optic Link Analysis

B-9

B.3.2.3

Typical Cable Attenuation Values — Table B-5 contains typical cable attenuations.

Each type of cable may have a different value for attenuation based on the application, cable type,
and vendor guarantee for the installation environment.

This table should be used as a reference for typical values only.

Table B-5:

Typical Cable Attenuation

Fiber Type

Typical Cable Type

Typical Value

Corning 1508

100/140; DEC BN25B-xx

4.5 dB/km*

Corning 1519

85/125; Siecor loose tube**

3.5 dB/km

Corning 1519

85/125; Siecor tight buffer

4.0 dB/km

Corning 1509 or

62.5/125; various loose tube

3.0 dB/km

AT&T Multimode

Corning 1517 or

50/125; various loose tube

2.5 dB/km

50/125; tight buffer tube

4.0 dB/km

AT&T Multimode
R

Corning 1517 or
AT&T Multimode

* This cable is guaranteed for 6 dB/km under specified applications for all temperature, humidity, and rated

RSO

tension conditions.

**Loose tube cable is usually very difficult to terminate with a connector. Splicing a connector cable end

onto the cable is recommended.

NOTE

[

To achieve the maximum link distances, premium grade fiber is recommended.
B

B.3.2.4

Single-Window and Dual-Window Fiber — When the goal of the installation is to

provide a transmission medium for an 835-nanometer (nm) system, single-window fiber can be

specified but dual-window fiber is recommended. If the cable contains extra fibers, or if future

[

changes or upgrades are anticipated, the person measuring the cable should also specify or list the

loss and bandwidth of the fiber at 1300 nm (the second window). Dual-window applications
imply a high bandwidth and distance (200-500 MHz'km for LED applications). Similarly, low

losses are implied (<2 dB/km) at this operating wavelength. Current DIGITAL general-purpose
cable, BNE25B-xx, uses a dual-window fiber.

B.3.2.5

Single-Mode Fiber — Currently, single-mode fibers are for laser/telephony applica-

tions. Telephony and computer data communications are two entirely different communications
systems. There is no practical method currently established that uses single-mode fiber in data
communications.

LEDs are the transmission device of choice for optimal cost and system

performance.

B-10

LAN Bridge 100 Technical Manual

B.3.3

Fiber-to-Pin Diode (Receiver)

The lowest-loss coupling type is fiber-to-receiver. Generally the receiver area is larger than the
spot size of the light exiting the fiber (see Figure B-5). Losses are due to large-scale mechanical

_—

misalignment, surface quality of the polished fiber end, fresnel losses due to different materials
with different refractive indexes, and efficiency of semiconductor devices. Fiber-to-receiver loss is

accounted for in the engineering design phase and should not be a concern in the field other than

for determining that the connector surface is not scratched, marred, smudged, and so on; conditions that would create greater losses than the design allows for.

=
FIBER

el
h

RECEIVER
-

h
LKG-0787

Figure B-5:

B.4

Fiber-to-Receiver Coupling

Numerical Aperture and Coupling Diameters

Coupling losses are not easily quantified in absolute terms. The following discussion describes the

apparent variability in

measurements.

When the connection is close to the source end, the fiber core is usually completely filled with

loss

measurements and the range

(optimistic to pessimistic)

of the

light, resulting in more loss from misalignment. When the connection is at the far end, the center
of the core is only partially filled with light, resulting in less loss from mechanical misalignment

(see Figure B-6).

NOTE

—

Losses near the source are usually higher and give pessimistic results. Measurement
of losses at distances greater than 100 meters (328 feet) from the source usually

gives optimistic results.

Fiber-Optic Link Analysis

B-11

FAR END

NEAR END

LIGHT LOST TO CLADDING

CLADDING

LATERAL

MISALIGNMENT

T R

CLADD
3
E
R
//
/
TBER i
CLADDING

CLADDING

LIGHT LOST

| |
l

LIGHT LOST

7
ANGULAR
MISALIGNMENT

Figure B-6:

B-12

/////////

Near-End and Far-End Coupling Losses

LAN Bridge 100 Technical Manual

B.4.1

Bending Losses

Typical cables will incur an additional —.5 dB loss because of changes in fiber bend radius. This
loss occurs because the light in the outer edge of the core is more easily lost to the cladding,
especially if the fiber is bent. In most systems, it is wise to assume the cable will be installed with
several bends; therefore, bend losses are included as part of the the fixed cable losses and do not
have to be separately included as part of the loss calculations (see Figure B-7).
NOTE

The manufacturer’s recommendation for minimum bend radius must be adhered to
for this to be valid.

DEBET l
MINIMUM BEND}

RADIUS

15 cm (6 in.)

FIBER OPTIC CABLE /
MKVv88-1839

Figure B-7:

Minimum Bend Radius

Fiber-Optic Link Analysis

B-13

B.5

System Losses and Variability

At the engineering level, a variety of factors must be quantified and specified. Many of these
factors are not measurable in the field. Listed below are a number of factors which may incur

losses in a system budget. These are calculated and taken into consideration for determining the
available budget, and do not have to recalculated when installing a fiber link. They are included
here for reference purposes only.
¢

LED Specification

LED Aging — Temperature Variations

LED Drive — Circuit Variations

Mechanical Coupling

(Connector Variations

Fiber Variations/Polishing

Temperature

(Cable Aging

Connector/Polishing

Coupling to Detector

Receiver Specification

SRR

Threshold Variations

B.6

Attenuation Measurement Procedures

This section provides three test methods that may be used to measure attenuation of a fiber-optic

link. The tests performed in the field can help determine whether the attenuation (loss between

the light launched into the fiber through the first connector and the light emitted at the other end

of the fiber) is within specified loss limits.
A

Though many methods of testing fiber-optic links are available, Digital Equipment Corporation
recommends using the first of the three methods described, which involves using an optical test
set (FOTEC model T302D). This will allow for the accuracy necessary to achieve the greater

distance capabilities of the higher power -RH/-RJ variations. The single-cable test (second method

outlined), is similar to the first method, but is not the preferred method for testing all fiber types.
This method is left here for informational purposes only. The third method is provided mainly for
measuring splice losses through the fiber-optic cable.

As described in Section B.4, the value of connector-to-connector loss may change depending on the
position of the connector relative to the light source and the direction in which the light is

traveling. Though it is desirable to determine that a given connection is within acceptable limits,
determining the value of “system loss” is the essential factor.

B-14

LAN Bridge 100 Technical Manual

B.6.1

Field Measurement Procedure for Relative Power Loss

The following describes the procedure for determining the acceptance of an installed fiber-optic
cable.

The test performed measures the relative loss of a system cable under test. It consists of launching
a pre-set quantity of optical power into an installed cable under test, and then measuring the
received power at the opposite end using an optical power meter. The value of the received power
represents the loss associated with the installed cable link.

B.6.1.1

po—

—
Lo
e

Equipment Required —

FOTEC optical power meter equipped with ST and SMA adapters

FOTEC optical LED source equipped with ST and SMA adapters

200/240 pum launch cable terminated with an ST and an SMA connector

Connector cleaning paraphernalia

ST and SMA barrel connector adapters

B.6.1.2 Preparation — Fiber-optic cables and connectors should always be handled with care.
Cables should not be bent or twisted tightly, especially near the connectors, as damage or breakage
may occur. Care should be taken to ensure that connectors are not dropped or physically impacted
in any way. If this happens, the connector may become damaged, impairing optical transmission.

—
i

All connector faces should be cleaned before testing or inserting into any optical port. This is done
by using Tex-wipes or similar cleaning pads dipped in alcohol to remove dust, debris, or fingerprints from the connector/fiber end. When removing and reinserting connectors, cleanliness
should be checked. If it appears that there is dirt or debris on the connector endface, it should be

re-cleaned.

—

Before any testing is performed, the equipment must be removed from the carrying case and
brought to room temperature. If the temperature in the testing environment is approximately 20
degrees above or below the previous environment the test equipment came from, allow up to
thirty minutes for this stabilization to occur.

B.6.1.3

Install the adapter that matches the cable connector into the FOTEC optical power meter.

Install the opposite type of adapter on the FOTEC optical LED source (this allows for the use

R
3.

Procedure — Perform the following set-up procedures

of the 200 um launch cable).

Clean and connect the launch cable to the optical source.

4. The optical source contains a sticker that indicates the wavelength of the transmitter in that
particular optical source. This wavelength is in the range of 790 to 850 nanometers (nm).
Make a note of the wavelength value. This value is needed during the Measurement Correction Calculation Procedure.

Fiber-Optic Link Analysis

B-15

To ensure that the launch cable-to-optical source connection is not disturbed, immobilize
both the launch cable and optical source as follows:
a.

Tape the optical source down to a stable surface using masking tape.

Tape the launch cable to the same surface as the optical source about 7.5 to 12.7 cm (3 to

5 in) away from the optical port of the source.

CAUTION
Take care not to bend the launch cable beyond its minimum bend radius of 12.7 cm

(5 in).

B.6.1.4

Test Source Adjustment Procedure — Perform the test source adjustment procedure

as follows:

Clean and connect the launch cable to the power meter.

Refer to Table B-6 and set the reading on the optical power meter to the correction setting that

corresponds to fiber size of the cable being tested. Set the reading by adjusting the trim pot

located on the source with a screwdriver.

NOTE
It is important that the trim pot not be readjusted or disturbed for the duration of

the measurement.

Table B-6:

Fiber Size Correction Settings for the Power Meter Reading

Fiber Size

Correction Setting

50 um

15.5 dBu

62.5 um

11.0 dBu

85 um

9.3 dBu

100 um

7.0 dBu

oGO,

ORI

Disconnect the launch cable from the power meter and connect it to the cable being tested

ARSI

using the corresponding barrel adapter (for example, if ST connectors are being used, the ST

barrel adapter must be used). The barrel adapter must be free of dirt and debris to ensure

optimum coupling of power.

When using SMA connectors, attach the plastic full sleeve that goes inside the barrel adapter.
The plastic sleeve must be free of dirt and debris and placed on the end of the first connector

that is to be inserted into the adapter before the connector is inserted.

Take the power meter to the opposite (receive end) of the cable being tested.

B-16

LAN Bridge 100 Technical Manual
A

At the receive end of the cable, clean the connector on the cable being tested and attach it to
the power meter.

NOTE
If a measurement is being made from one patch panel to another patch panel

without the drop cables to the installed equipment, it will be necessary to use a
jumper from the patch panel to the power meter. In this case, use a jumper with a
fiber core that is larger than the fiber in the cable under test.

B.6.1.5

Measurement Correction Calculation Procedure — During this procedure, complete

the End-to-End Cable Loss Calculation Worksheet. This worksheet is used to arrive at the corrected
cable loss by recording the measured end-to-end cable loss and adding in correction values.
Figure B—-8 is a blank End-to-End Cable Loss Calculation Worksheet. Before proceeding with the
measurement correction calculation procedure, make at least one photocopy of this worksheet for

each fiber optic cable to be certified.

END-TO-END CABLE LOSS
- WORKSHEET

MEASURED RELATIVE LOSS

(1)

LENGTH OF TESTED CABLE (TO NEXT .1 km)

(2)

WAVELENGTH CORRECTION VALUE

(3)

TOTAL WAVELENGTH CORRECTION
(MULTIPLY LINE 2 BY LINE 3)

(4)

CORRECTED CABLE LOSS: SUBTRACT
(SUBTRACT LINE 4 FROM LINE 1)

(5)
MKVv88-1838

Figure B-8:

End-to-End Cable Loss Calculation Worksheet

Take one copy of the blank worksheet (Figure B-8) and perform the following:
1.

On the Measured Relative Loss line of the worksheet, write the value that is displayed on the
power meter.

On the Length of Tested Cable line of the worksheet, write the rounded-off length of the
cable under test by taking the cable length and rounding it upwards to the next tenth of a
kilometer (for example; if the cable length is 1375 meters, round it off to 1.4 km)

Fiber-Optic Link Analysis

B-17

Refer to Table B-7 and choose the wavelength correction value that applies to the optical

source wavelength and to the Ethernet option being installed (the optical source wavelength is
the wavelength value that was noted during Step 4 of the Set-Up Procedure in Section

B.6.1.3). Write the value from Table B-7 on the Wavelength Correction Value line of the
worksheet.

Table B-7:

Wavelength Correction Values

Wavelength

DEBET-RC/RD/RH/RJ
|

—.4

795

—-.3

800

—.2

805

-.1

810

830

BN
N = O

790

835

815

820
825

840
845

.6
SRR

850

Multiply the Wavelength Correction Value line by the Length of Tested Cable line and
write the result on the Total Wavelength Correction line of the worksheet.

Calculate the corrected output power by subtracting the Total Wavelength Correction line

value from the Measured Relative Loss line value and write the result in the Corrected

Cable Loss line of the worksheet.
R

NOTE
AN

To subtract two negative values, drop the minus sign (—) on both numbers, subtract
the smaller number from the larger value, and add a minus (—) to the result. For
example —.5 subtracted from —7.4 becomes 7.4 minus .5 which results in —6.9
with the minus sign added.

To subtract a positive number from a negative number, drop the minus sign (—)

from the negative number, add the two numbers together, and add a minus sign (-)
to the result. For example, .5 subtracted from —7.4 becomes 7.4 plus .5 which
results in —7.9 with the minus sign added.

B-18

LAN Bridge 100 Technical Manual
AR

B.6.1.6 System Certification Prodedure — Table B-8 defines the minimum and recommended
maximum loss values for fiber-optic cable by fiber size and Ethernet option.
-

Table B-8:

Cable Certification Values

—

Recommended Maximum
Fiber

Size

Minimum

Loss*

Loss**

DEREP/DEBET-RC/RD

50 um

None

—1.1 dB

62.5 um

None

—6.3 dB

85 um

None

—8.1 dB

100 um

None

—-10.5 dB

DEREP/DEBET-RH/R]

50 um

None

—6.0 dB

62.5 um

None

—-10.0 dB

85 um

None

—11.5 dB

100 um

3 dB

—~14.0 dB

* If the relative loss is less than this value, an optical attenuator must be installed on the transmit end of the
system cable (see Section B.9).

**This recommended maximum loss value allows for 1 dB of additional loss for future repairs.

—
f

Compare the value on the Corrected Cable Loss line of the worksheet to the values defined in
Table B-8. If the Worksheet value is equal or less than the values from Table B-8, the installation
is acceptable.

—

If the planning and design of the cable plant was executed correctly, allowances should have been
made for splice loss, connector coupling, and future maintenance of the installation. The recommended maximum loss in Table B-8 represents the loss that allows for future maintenance of the

cable plant. The future maintenance can include splices or swapping of system boards.

Fiber-Optic Link Analysis

B-19

B.6.2

Optical Test Set (Single-Cable) Method

This method uses an optical test set to measure optical attenuation through a fiber-optic link. A
reference cable (1 meter in length) is used to calibrate the light meter relative to the light source.

Once calibration is complete, the reference cable should not be disconnected from the light
source.

Use the following procedure to measure attenuation of the fiber-optic link with an optical test set.

NOTE
Any reference or adapter cables used in the test procedures must have the same

type of optical fiber (such as 100/140, 62.5/125, and so on) as the cable to be
tested. In addition, the connectors used at the ends of the reference cable must

match both the test equipment connector and the connector of the cable under test.

Use of a reference cable that has a different fiber type than the cable under test

yields inaccurate measurements.

Calibrate the test configuration as follows (see Figure B-9):
a.

Attach the necessary connector adapters to the light source and the light meter.

Connect the reference cable to the light source and turn on the light source. Allow the

light source several minutes to stabilize.

Connect the light meter to the other end of the reference cable.
Use the calibration knob on the light meter to set the reading to 0 dB (do not alter the

calibration knob once it has been set).

NOTE
When measuring small fiber sizes such as 50/125, it may not be possible to

calibrate the light meter to 0 dB. In this case, adjust the meter as close as possible
to 0 dB and note the reading.
-

NOTE
When connecting cables to test equipment, use a half-length alignment sleeve.

When connecting cables together using a barrel connector, use a full-length alignment sleeve. A procedure for using alignment sleeves is provided in the Fiber-Optic
Pocket Guide EK-FIBOP-IS.

Be sure that the light meter is designed to measure the desired wavelength of light.
A

That is, if the fiber system operates at 820 nm, the light meter must be designed to

measure light loss at 820 nm.

B-20

LAN Bridge 100 Technical Manual

Disconnect the light meter from the reference cable.
NOTE

Once calibration is completed, do not disconnect the light source from the reference cable. Also do not move (tighten or loosen) the connector that is attached to
the light source.

REFERENCE

LIGHT

SOURCE

CABLE

A
|

LGHT

METER

HALF-LENGTH
ALIGNMENT SLEEVES
LKG-0790

Figure B-9:

Calibrating the Optical Test Set

2. Measure each channel of the fiber-optic link in the direction that light will be transmitted
under normal operating conditions (see Figure B-10).

Connect the light source and reference cable to the end of the cable under test that is
labeled “Transmit”.

b. Connect the light meter to the other end of the cable under test that is labeled “Receive”.
c.

Record the reading obtained from the light meter. If a calibration other than 0 dB was
obtained in Step 1, subtract that number from the present reading on the light meter.

For example, if the reading obtained in Step 1 was —1.5 dB and the reading obtained in
Step 2 is —11.9 dB, the corrected reading is —10.4 dB. The equation for this example is:
~11.9 — (-1.5) = —10.4

Repeat the measurement in Step 2 for the remaining channel(s).

Fiber-Optic Link Analysis

B-21

BARREL

CONNECTOR

e REFERENCE-—)I<————-—- CABLE UNDER TEST
|

CABLE

LIGHT

SOURCE

CE-\:D

NFIGURATION
| Gyt

METER

TESTING

HALF-LENGTH

FULL-LENGTH

HALF-LENGTH

ALIGNMENT SLEEVE

TRANSMITTER

Recever | | CONFIGURATION

DIRECTION LIGHT

(DETECTOR) | | UNDER NORMAL
|

| oPeraTION

LED

IS TRANSMITTED
B

LKG-0791

Figure B-10:

B.6.3

Measuring the Cable Under Test

Optical Time Domain Reflectometry

This method uses an optical time domain reflectometer (OTDR) to measure the length and

attenuation of a fiber-optic cable. This method may also help to identify the location and relative

ARG

value of losses.
P

WARNING
Do not look into a fiber-optic connector or cable end while the cable is connected

——

to an OTDR. Follow safety procedures recommended by the OTDR manufacturer.

NOTE

It is important to perform an OTDR measurement from each end (transmit and
receive) of both channels. This ensures that a broken fiber at the far end of a

channel is not missed (mistaken for the end of the fiber).
This procedure requires that all test equipment has been calibrated to the manufac-

AOTAE:

turer’s specifications within the recommended calibration period.
A

B-22

LAN Bridge 100 Technical Manual

NOTE

Any reference or adapter cables used in the test procedures must have the same
type of optical fiber (such as 100/140, 62.5/125, and so on) as the cable to be
tested. In addition, the connectors used at the ends of the reference cable must
match both the test equipment connector and the connector of the cable under test.
Use of a reference cable that has a different fiber type than the cable under test
yields inaccurate measurements.

B.6.3.1

Equipment Setup — Adjust any settings on the OTDR according to the manufacturer’s

instructions.

NOTE

The OTDR must measure light at the same wavelength used by the optical fiber
system. Both the LAN Bridge 100 unit and the Remote Ethernet Repeater use a
wavelength of 820 nm.

B.6.3.2 Link Measurement — Link measurement involves recording the readings obtained from
the OTDR. A permanent record of fiber-optic link measurement may be possible using a strip chart
or camera. Such records should be obtained whenever possible and stored for future reference.

Measure the fiber-optic link using the following procedure:
1. Connect the cable under test to the OTDR in accordance with the procedures outlined in the
OTDR manual.

NOTE

The pigtail, if needed (see Figure B-11) must be of the same fiber type and have
the same mating connectors as the cable under test and the OTDR. Some OTDRs
may have a different connector than the cable under test.

PIGTAIL

—T
D)

CONNECTORS

CABLE UNDER TEST

SPLICE

mmmn

CONNECTORS

LKG-0784

Figure B-11: OTDR Test Configuration

Fiber-Optic Link Analysis

B-23

Measure the length and attenuation of each fiber (transmit and receive). Note the splice and
connector losses (see Figure B—12). Look for abnormalities in the slope of the display.

—

e r——

Move the OTDR to the other end of the fiber system and measure the length and attenuation of

each fiber (transmit and receive). Note that the splice and connector losses are similar to those
obtained in Step 2.

CONNECTOR

BEGINNING
OF FIBER

END OF
SPLICE

FIBER

;

TM
PO

0.0 dB

-1.0dB

—— e ——

|_04dB

—20dB

0.8 dB {
Loss

L _f__ _|[—-_A_

-3.0 dB

-4.0 dB

-5.0dB
S

600 METERS

200

METERS
A

LKG-0792

Figure B-12:

B-24

OTDR Trace

LAN Bridge 100 Technical Manual
BT

Bandwidth Analysis and Derating Factors

B.7

B.7.1

Bandwidth in a fiber cable is often discussed in terms of its dispersion characteristics because
narrow light pulses launched into a fiber will be broadened (dispersed in time) at the fiber’s end.
This pulse broadening forces an increase in the amount of spacing between bits or pulses so as to

avoid Inter-Symbol Interference (ISI) (overlapping of the dispersed bits). Consequently, the

Fiber Bandwidth Rating System

greater the amount of fiber dispersion, the lower the bandwidth of the fiber.
There are several types of fiber dispersion. For multimode fiber with LED light sources there are
two dominant types: modal and chromatic.

Modal dispersion occurs because light modes within a fiber have different group velocities.
Thus in multimode fiber, where hundreds of modes are excited, a narrow pulse at the input

will have each mode or group of modes arriving at the output at slightly different times. This
broadens or disperses the output pulse. Modal dispersion is one of the dominant forms of

dispersion in multimode fibers with 1300-nm LED sources, thus limiting the bandwidth of
such systems to 100-450 MHzkm.

A light-ray analysis supports the concept of modal dispersion. Light rays emitted from a LED

—

each enter the fiber at a different angle and continue to travel through the fiber along separate

-l

paths. Thus, the pulse broadening is the difference between the time of the longest ray path
and the time of the shortest ray path.

Chromatic dispersion occurs because each wavelength of light within a fiber travels at a
slightly different speed. As a result, light from wide-spectrum sources (LEDs) is very suscepti-

ble to chromatic dispersion. The dominant form of dispersion in the 800- to 900-nm region is

chromatic dispersion, thus limiting the fiber bandwidth to 25-70 MHz*km for LEDs. Chromatic
dispersion lessens around the 1300-nm region, making modal dispersion the other major bandlimiting factor.

The various forms of dispersion cause the fiber bandwidth to depend greatly upon the spectrum

characteristics of the light source. Thus the bandwidth measurements reported have limited

meaning until they are correlated with the LED spectrum data in the derived relating equations.

—

Fiber bandwidth is specified in units of megahertz times kilometers because both dispersion
types continue to broaden (disperse) the light pulse as fiber length increases. The fiber

bandwidth is traditionally specified at 1 km since it decreases with fiber length.
B.7.2

Distance Derating Factors for Ethernet Products

—

It is assumed that the cable is rated at 300 MHz°-km (modal bandwidth) or greater. For LEDs, this

translates to a fiber bandwidth of about 50 MHz at a distance of 1 km. Using cable with a higher
rating does not decrease the bandwidth penalty (derating factor) enough to change the derating

factor. Bandwidth derating will be .5 dB/km.

—

bandwidth rolloff. Bandwidth rolloff is similar to the effect of a low-pass RC filter. The derating

It is apparent that the fiber bandwidth limits cable length by attenuating the signal with
factor compensates for bandwidth attenuation by ensuring a larger signal at the input.

Fiber-Optic Link Analysis

B-25

B.8

Suggested System Design Methodology

The worksheet shown in Figure B—13 can be used to calculate the LAN Bridge 100 budget loss for
a system installation. All values needed for the worksheet are available in this document or are

specified to be provided by the vendor or installer.
The fiber-optic system has a specification called allowable system loss. This is the maximum loss

that can occur between the two cable connectors that attach to the receptacles of the fiber-optic

interconnect devices (given in Tables B-2 and B-3). Note that the system is specified for the
amount of loss after the light is in the cable. That is, the connectors that attach to the LED and to

the fiber-optic receiver are already accounted for in the calculations of the maximum allowable
system losses. DO NOT ADD IN THEIR LOSSES AGAIN.

Compare the sum of the losses to the allowable system losses listed in Table B-2 or B-3. The sum of
the losses must be less than or equal to the allowable system losses. MAKE SURE THE PROPER
VALUE IS USED FOR THE LAN Bridge 100 OR DEREP VARIATION BEING USED.

'PRODUCT NAME =

BRI

Fiber Type

Allowable System Loss

(see Tables B-2 or B-3 based on product variation)

FACTOR

VALUE

— ——dB/conn.

QUANTITY

LOSS

BARREL

CONNECTIONS

——

of connectors

(B.3.2.1)

SPLICES

— —— dB/splice

——

of splices

(B.3.2.1)
CABLE

ATTENUATION

——— dB/km

X ———

km of fiber

0.5 dB/km

km offiber

(B.3.2.3)
BANDWIDTH
DERATING

———

(B.7.2)
R

MARGIN

1.5 dB

Sum of losses

=
MKV88-1799

Figure B-13:

B-26

Sample Worksheet for Budget Loss Calculation

LAN Bridge 100 Technical Manual

—

B.8.1

Example Worksheets

This section contains two sample worksheets. The first example (Figure B-14) shows sample

values and loss calculations for a DEBET-RH using 62.5/125 micron fiber in a 2 km link. The
second example (Figure B—15) shows system loss calculations for a DEBET-RC using 100/140

micron fiber in a 1 km link.

These loss calculations are for products that use the 54-16053-01 fiber-optic interface module.
Digital Equipment Corporation products that use this interface include the following:

LAN Bridge 100 (DEBET-RC)

LAN Bridge 100 (DEBET-RD)

LAN Bridge 100 (DEBET-RH)

LAN Bridge 100 (DEBET-R))

Remote Ethernet Repeater (DEREP-RA)

Remote Ethernet Repeater (DEREP-RB)

Remote Ethernet Repeater (DEREP-RC)

Remote Ethernet Repeater (DEREP-RD)

Remote Ethernet Repeater (DEREP-RH)

Remote Ethernet Repeater (DEREP-RJ)

Fiber-Optic Link Analysis

B-27

B.8.1.1

Sample Loss Calculation — Figure B—14 is a sample worksheet of a loss calculation for

a 2 km link using 62.5/125 micron optical fiber cable and a DEBET-RH.
Using the allowable system losses listed in Table B—-3, and the cable loss values listed in the
manufacturer’s specifications, Figure B—14 shows the allowable total loss.

PRODUCT NAME = DEBET
- RH

Fiber Type
Allowable System Loss

= 62.5/125 micron
=12 dB

(see Tables B-2 or B-3 based on product variation)

FACTOR

VALUE

QUANTITY

TOTAL LOSS

BARREL

CONNECTIONS

1.5 dB/conn. X O connectors

0.0dB

SPLICES

0.2 dB/splice X 4 splices

0.8 dB
DO

CABLE

ATTENUATION

3.0 dB/km X 2.0 km of fiber

6.0dB

DERATING

0.5 dB/km X 2.0 km of fiber

1.0dB

MARGIN

1.5 dB

1.5dB

9.3 dB

BANDWIDTH

Sum of losses

MKV88-1797

Figure B-14:

Sample Worksheet for DEBET-RH Loss Calculation

The total system losses are less than the allowable system losses, therefore, the system design is

valid.

B-28

LAN Bridge 100 Technical Manual
R

Figure B—15 is a sample worksheet of a loss calculation for a 1 km link using 100/ 140 micron

optical fiber cable and a DEBET-RC.

Using the allowable system losses listed in Table B-2, and the cable loss values listed in the
manufacturer’s specifications, Figure B—15 shows the allowable total loss.

- RC
PRODUCT NAME = DEBET

= 100/140 micron

Fiber Type

—
-

=12.5dB
Allowable System Loss
(see Tables B-2 or B-3 based on product variation)

FACTOR

—

BARREL

VALUE

QUANTITY

TOTAL LOSS

CONNECTIONS

0.8 dB/conn. X 1 connectors

0.8dB

—

SPLICES

0.2 dB/splice X 2 splices

0.4 dB

ATTENUATION

4.5 dB/km X 1.0 km of fiber

45 dB

BANDWIDTH

DERATING

0.5 dB/km X 1.0 km of fiber

0.5 dB

MARGIN

1.5 dB

1.5dB

7.7 dB

CABLE

Sum of losses

MKV88-1798

Figure B-15:

Sample Worksheet for DEBET-RC Loss Calculation

The total system losses are less than the allowable system losses, therefore, the system design is
valid.

e
|

It should be noted that a number of choices are available to link planners to minimize the link
losses. Various types of splices produce different losses. Cables have a range of attenuation and
lower values of attenuation may be used. Connectors may be eliminated in the link.

Fiber-Optic Link Analysis

B-29

B.8.2

Worst-Case and Statistical Results

The data referenced in this section is based on a statistical analysis of fiber-optic systems. In a
system with many variables, actual measured system loss will not be likely to match the worst-case
calculations. In most cases, the calculated loss will exceed actual measured loss since it would be

statistically improbable that every system component would have the worst-case characteristics.

However, if worst-case estimates are ignored so that the installed system depends entirely on a

statistical spread, then the network installer or consultant is responsible for ensuring adequate
performance of the system. The installer or consultant is responsible for diagnosing and modifying

the system in the event of failure.

B.9

PSRN

Attenuator Installation

The fiber-optic attenuator (P/N: 12-30068-01) is a device that induces a loss of 3 dB in a fiberoptic system. The attenuator (Figure B-16) is designed to be used with short cables and
high
POWEr sources.

PTG

ATTENUATOR

FIBER OPTIC

CONNECTOR

e
FIBER OPTIC CABLE
(SMA 906 CONNECTOR)

HALF

(f‘;%JRCE

DELRIN

MKV88-1412
RN

Figure B-16:

B-30

Attenuator

LAN Bridge 100 Technical Manual

B.9.1

When to Install the Attenuator

The attenuator can be used on the LAN Bridge 100 (DEBET-RH/RJ) and Remote Ethernet Repeater
_—

(DEREP-RH/R]).

NOTE

The attenuator can be used only on LAN Bridge 100 or DEREP devices that have the
latest revision fiber-optic module (54-16053-01 Revision B1)

Use the attenuator when a:

DEBET-RH/RJ or DEREP-RH/R]J is connected to a DEBET-RH/RJ or DEREP-RH/RJ (see Figure
B-17).

—

— Install the attenuator for 100/140 fiber-optic link that is 1000 m (3281 ft) or less.

— Do NOT install the attenuator for 100/140 fiber-optic link beyond 1000 m (3281 ft).

— Do NOT install the attenuator for 50/125, 62.5/125, and 85/125 optical fiber links.
e

DEBET-RH/RJ is connected to a DEBET-RC/RD or DEREP-RC/RD.

— Install the attenuator on one side (DEBET-RH/R]J) by attaching the attenuator to the output
(transmit) connector only (see Figure B—18).

— If an older DEBET-RC/RD is being used with a newer DEBET-RH/R], the guidelines and
budgets of the DEBET-RC/RD must be followed. Refer to LAN Bridge 100 Installation/User’s Guide, Section 3.4 (Fiber-Optic Cables).
NOTE

If a special condition exists and more budget is required, the attenuator can be
removed. This will increase the budget by 3 dB.

Fiber-Optic Link Analysis

B-31

ATTENUATOR IS INSTALLED

ON BOTH SIDES
ORI

LR,

LB,

\-————f_': RX

~=——1000 M OR LESS g

100/140 FIBER
DEREP OR DEBET

- DEREP OR DEBET

— RH/RJ VARIATION

Figure B-17:

MKV88-1413

Installing the Attenuator on Both Sides of the Link

ATTENUATOR IS INSTALLED

ON ONE SIDE ONLY
AR

:]-_—-\

\———C RX
~a—— 1000 M OR LESS
100/140 FIBER

DEREP OR DEBET

DEBET

— RC/RD VARIATION

— RH/RJ VARIATION

MKV88-2046
ARSI

Figure B-18:

Installing the Attenuator on -RH/-RJ Side Only
IR

B-32

LAN Bridge 100 Technical Manual

B.9.2

Where to Install the Attenuator

The attenuator is:

Installed ONLY on the SMA 906 connector that is on the fiber-optic cable end attached to the
transmitter (output) of the bridge or repeater.

Only one attenuator per fiber-optic cable.

Do not install the attenuator onto the RX end of the fibér-—optic cable.
Do not install attenuators on both ends of the same optical fiber.

Do not install the attenuator on a DEBET-RC/RD or DEREP-RC/RD unit.
If an attenuator is required for the second fiber, it will be installed on the SMA 906 connector
attached to the transmitter of the bridge or repeater at the other end of the link.

When the attenuator is installed, the cable should be marked with a label (36-18460-01)
indicating “3 dB attenuator installed” and the loss of the cable (if known).

Fiber-Optic Link Analysis

B-33

B.9.3
1.

How to Install the Attenuator

SOOI

Pull the protective caps from both the fiber-optic cable (SMA 906 connector) and the LAN
Bridge 100 or DEREP fiber-optic connectors (see Figure B—19).

FIBER OPTIC

CONNECTORS
L

PROTECTIVE
CAPS

FIBER OPTIC

CABLE
MKV88-1660

Figure B-19:

B-34

Removing the Protective Caps from the LAN Bridge 100

LAN Bridge 100 Technical Manual
PO

2. Remove the attenuator (P/N: 12-30065-01) from the plastic bag.
CAUTION

The attenuator must be kept free of dust and dirt to ensure proper installation.

3. Install the attenuator between the source (TX) fiber-optic output connector and the fiberoptic cable connector (SMA 906). See Figure B-20.
NOTE

It is possible for the attenuator to fall out of the connector during installation.
Before screwing the cable into the connector, ensure that the attenuator is still in
place.

Connect the fiber-optic cable. Finger tighten the fiber-optic connectors.

FIBER OPTIC

OUTPUT

INPUT

RECEIVE

BEND RADIUS
15 CM (6 IN)

FIBER OPTIC CABLE
MKV88-1673

Figure B-20:

Connecting the Fiber-Optic Cables to the LAN Bridge 100

Fiber-Optic Link Analysis

B-35

Place the attenuator label on the transmit cable of the fiber-optic cable (see Figure B-21).

FIBER OPTIC

OUTPUT

ATTENUATOR LABEL
FIBER OPTIC

ON TX SIDE

INPUT

FIBER OPTIC CABLE
MKV88-1417

Figure B-21:
6.

Fiber-Optic Cable Installed to the LAN Bridge 100

Refer to the

Fiber-Optic Attenuator Installation and Configuration

Card or the

DEBET/DEREP Fiber-Optic Upgrade document for complete installation.
B —

B.10

Reference Material

The following documents may be of general help in consulting on fiber-optic systems.
®

DEREP-RA Remote Ethernet Repeater Installation/Owner’s Manual (EK-DEREP-IN)

Ethernet Repeater Technical Manual (EK-DEREP-TM)

Ethernet Installation Guide (EK-ETHER-IN)

Digital Fiber-Optics Service Plan, May 1984

Fiber-Optic Pocket Guide (EK-FIBOP-IS)

FOTEC Practical Testing of Fiber-Optic Systems

Siecor Fiber-Optic Handbook for Use With DEC Ethernet Repeaters

Ampbenol Fiber-Optic Designer’s Handbook

ATET Lightguide Cable Specification, Issue 5

ANSI Standard Z136.2: Fiber-Optic Safety

DEBET/DEREP Fiber-Optic Upgrade (EK-DEFOU-IN)

Fiber-Optic Attenuator Installation and Configuration Card (EK-DEFOE-RC)

B-36

I0;

LAN Bridge 100 Technical Manual
R

P
|

Circuit Descriptions (Cont.)
search results register, 4-31

Adfi;’ess
iltering, 37
Address Table

TLU, 4-26 to 4-31
Configuring Extended LANs, 1-16

Connectors, 1-9, 1-10, 1-14