Remote Automation Solutions Guide: Bristol Network 3000 Communications Configuration Guide Manuals & Guides

Network 3000

Communications

Configuration Guide

(A user's guide to software configuration
and troubleshooting for the Network 3000
Communications system.)

Bristol Babcock

D5080 Issue: April, 2001

Notice

Copyright Notice

The information in this document is subject to change without notice. Every effort has been
made to supply complete and accurate information. However, Bristol Babcock assumes no
responsibility for any errors that may appear in this document.

Bristol Babcock does not guarantee the accuracy, sufficiency or suitability of the software
delivered herewith. The Customer shall inspect and test such software and other materials to
his/her satisfaction before using them with important data.

There are no warranties, expressed or implied, including those of merchantability and fitness
for a particular purpose, concerning the software and other materials delivered herewith.

Request for Additional Instructions

Additional copies of instruction manuals may be ordered from the address below per attention
of the Sales Order Processing Department. List the instruction book numbers or give the
complete model, serial or software version number. Furnish a return address that includes
the name of the person who will receive the material. Billing for extra copies will be according
to current pricing schedules.

ACCOL is a trademark and Bristol is a registered trademark of Bristol Babcock. Other
trademarks or copyrighted products mentioned in this document are for information only, and
belong to their respective companies, or trademark holders.

Copyright (c) 2001, Bristol Babcock, 1100 Buckingham St., Watertown, CT 06795. No part of
this manual may be reproduced in any form without the express written permission of Bristol
Babcock.

A Few Words about Bristol Babcock

For over 100 years, Bristol® has been providing innovative solutions for the measurement
and control industry. Our product lines range from simple analog chart recorders, to
sophisticated digital remote process controllers and flow computers, all the way to turnkey
SCADA syst em s . Over t he yea r s, we have b ec om e a l ea d i ng s up plier to t he el ec tronic gas
measurement, water purifi c a tion, and w a s tewater treatm ent indust r i es .

On off-shore oil platforms, on natural gas pipelines, and maybe even at your local water
company, there are Br i s tol Babcoc k i ns truments , controll er s , and sys tems running yea r - i n
and year-out to provide accurate and timely data to our customers.

Getting Additional Information

In addition to the information contained in this manual, you may receive additional
assistance in using this product from the following sources:

Contacting Bristol Babcock Directly

Bristol Babcock's world headquarters are located at 1100 Buckingham Street, Watertown,
Connecticut 06795, U.S.A. Our main phone numbers are:

[phone]
[phone] (FAX)

Regular office hours are Monday through Friday, 8:00AM to 4:30PM Eastern Time, excluding
holidays a nd scheduled f a c tory shut d ow ns . During ot her hour s, caller s may leave m es sages
using Bristol's voice mail system.

Telephone Support - Technical Questions

During regular business hours, Bristol Babcock's Application Support Group can provide
telephone support for y our technical q ues tions.

For technical questions regarding ACCOL, Open BSI, or UOI call [phone].

For technical questions regarding Bristol's OpenEnterprise product, call [phone] or
e-mail [e-mail]

For technical questions regarding Bristol's Enterprise Server® / Enterprise
Workstation® products, call [phone].

For technical questions regarding TeleFlow products call [phone].
For technical questions regarding Network 3000 hardware products call [phone].

For technical questions about ControlWave call [phone] or [phone].

You can e-mail the Application Support Group at: [e-mail]
The Applic a tion Support G r oup mainta ins an area on our w eb s i te for software upd a tes and

other technical information. Go to: www.bristolbabcock.com/services/techsupport/

i

For assistance in interfacing Bristol Babcock hardware to radios, contact Communication
Technologies in Orlando, FL at [phone] or [phone].

Telephone Support - Non-Technical Questions, Product Orders, etc.

Questions of a non- technical na ture (product orders, literature r eq uests, price and delivery
information, etc.) should be directed to the nearest sales office (listed below) or to your
Bristol-authorized sales representative.

Major U.S. Sales Offices Major Internati o nal Sales Offices:

Watertown, CT [phone] Bristol Babcock Ltd (UK): (441) 562-820-001
Birmingham, AL [phone] Bristol Babcock, Canada: [phone]
Ontario, CA [phone] Bristol Babcock Asia Pacific 61-8-9791-3654
Farmington, NM [phone] BBI, S.A. de C.V. (Mexic o) [phone]
Houston, TX [phone]
Richardson, TX [phone]

Please call the main Bristol Babcock number (860-945-2200) if you are unsure which office
covers your particular area.

Visit our Site on the World Wide Web

For general information about Bristol Babcock and its products, please visit our site on the
World Wide Web at: www.bristolbabcock.com

Training Courses

Bristol Babcock’s Training Department offers a wide variety of courses in Bristol hardware
and software at our Watertown, Connecticut headquarters, and at selected Bristol regional
offices, throughout the year. Cont act our T raining Depa rtment at ( 860) 945-2269 for course
information, enrollment, pricing, and schedules.

ii

Who Should Read This Manual?

This document is intended to be read by a system engineer or technician who is
configuring or trouble shooting Network 3000 communications. It assumes that
controllers have been i nstalled, ACCOL programming has been completed a nd the
controllers have running ACCOL loads, and that all network cabling has been
connected and tested.

NOTE: Although some of the information in this book is also applicable to
ControlWave-series controllers, they are not part of the Network 3000 product line,
and so will not be covered as part of this book.

This document assumes familiarity with the following subjects:

• Users must have some level of knowledge regarding how their Network 3000
hardware is to be used. In particular, they should be familiar with the data
update requirements for their particular process. Users should also have access
to whatever drawing or document specifies the addresses of each controller.

• Users must be familiar with using certain Open BSI Utilities programs such as
the Open BSI Setup Tool (Open BSI 2.3 or earlier) or NetView (Open BSI 3.0 or

newer), and DataView. For information on these subjects, see the Open BSI
Utilities Manual (Ver 2.x document# D5076 or Ver 3.x/4.x document# D5081).

• Users may need to alter signal values, and certain other parameters in their
ACCOL load as part of the Network tuning process. As such, some familiarity
with ACCOL signals is helpful. See An Introduction to ACCOL (document#
D4056) and the ACCOL II Reference Manual (document# D4044) for details.

• Users with Open BSI 2.3 (or earlier) may need to edit their Network Topology
files using the NETTOP program. This subject is discussed in the NETTOP and
NETBC User Manua l (document# D4057). Users with Open BSI 3.0 (or newer)
may need to edit their Network Definition (NETDEF) files using NetView, and
specify RTU configuration parameters using LocalView. NetView and LocalView
are discussed in the Open BSI Utilities Manual (Ver 3.x/4.x document# D5081).

iii

BLAN

iv

Table of Contents

Chapter 1 - Introduction............................................................................................ 1-1

BSAP Networks.......................................................................................... 1-1

IP Networks................................................................................................ 1-4

Converting NETTOP Files For Use Below An NHP.................................1-8

Overview of Configuration......................................................................... 1-9

Is this still unclear? ........................................................................ 1-10

Chapter 2 - Guidelines For Setting Up BSAP Networks..........................................2-1

A Quick Review of BSAP Network Terminology...................................... 2-1

Things to Be Aware of When Defining Your Network ............................. 2-3

Chapter 3 - Guidelines For Setting Up IP Networks................................................ 3-1

A Discussion of IP Network Terminology ................................................. 3-1

What is A Network Host PC (NHP)?.........................................................3-2

What is the Format of IP Addresses?........................................................ 3-6

Things to Be Aware of When Defining Your Network ........................... 3-11

Chapter 4 - Guidelines For Configuring the BSAP Data Link ................................ 4-1

Setting Up the Master Ports (including Port, Poll Period, #NDARRAY) 4-2

Setting Up the Slave Ports (including Port, Poll Period)....................... 4-14

Using Advanced Poll Period Parameters ................................................ 4-16

Setting Up Data Link Parameters At the Network Master................... 4-18

BSAP Local Line (used for special applications only)............................ 4-25

Configuration Issues For Modems........................................................... 4-26

Configuration Issues For Radios............................................................. 4-37

Chapter 5 - Guidelines For Configuring the IP Data Link...................................... 5-1

Declaring An IP Port in ACCOL Workbench............................................ 5-2

Specifying the Characteristics of the IP Port(s)........................................ 5-3

Defining the IP System Parameters.......................................................... 5-6

Chapter 6 - Application Notes ................................................................................... 6-1

Peer-to-Peer Communication (BSAP Master/Slave Modules).................. 6-1

Peer-to-Peer Communication (IP_Client/IP_Server Modules)................. 6-4

Setting Task Rates..................................................................................... 6-7

Setting Task Priorities............................................................................... 6-7

ACCOL Load Design Issues....................................................................... 6-8

Remote Process I/O..................................................................................... 6-8

Communication I/O Buffers....................................................................... 6-8

Report By Exception (RBE)........................................................................ 6-9

Open BSI Utilities (Version 2.3 or earlier).............................................. 6-10

Open BSI Utilities (Version 3.0 or newer)............................................... 6-11

Open BSI Scheduler and Open BSI Data Collector Array/Archive

Collection (All versions)....................................................................... 6-12

v

OpenEnterprise.......................................................................................... 6-12

Enterprise Server..................................................................................... 6-13

Intellution® FIX® .................................................................................... 6-13

Iconics Genesis ......................................................................................... 6-14

Chapter 7 - Trouble-Shooting Network 3000 Communications............................... 7-1

Trouble-Shooting Techniques.................................................................... 7-2

Frequently Asked Questions Concerning Communication Problems.... 7-22

Checklist for Diagnosing Communication Problems.............................. 7-30

Appendix A - Summary of Port Types.......................................................................A-1

Appendix B - Message Traffic....................................................................................B-1

Glossary .....................................................................................................................G-1

vi

Chapter 1 - Introduction

Introduction

Bristol Babcock’s Network 3000-series of remote process controllers (DPC 3330s,
RTU 3305s, 3530s, etc.) run ACCOL loads to measure or control some part of a
process. Besides performing these control duties, each Network 3000 controller is
designed to operate as a

node

The data tran smi tted thro ug h the ne twork i s sen t a ccordi n g to a p re- de fi n ed f orma t
called a network

communication protocol

protocols supported by Bristol devices.1
Protocol) is a Bristol Babcock proprietary protocol which is supported by all
Network 3000 devices2. IP (Internet Protocol) is a more industry-standard protocol,
and is supported by many types of computers world-wide. Currently the 386EX
Protected Mode DPC 3330/DPC 3335 controllers with Ethernet hardware installed
are the only Network 3000-series controllers which support the IP protocol. Bristol
Babcock’s ControlWave controller also supports IP, however, ControlWave
communications are not covered in this book.

in a data network.

. There are two basic types of network

BSAP

(Bristol Synchronous/Asynchronous

In addition to the choice between these two protocols, it is possible to ha ve separate
BSAP and IP networks as part of the same overall system, or to have a BSAP
network serve as a

sub-network

of an IP node.

BSAP Networks

At the top o f a BS AP netw ork i s a ho st compu ter, cal led th e
network master is usually a PC Workstation running

(HMI)

or

supervisory control and data acquisition (SCADA)

human-machi ne interface

network maste r

software such as

Intellution® FIX®, Iconics Genesis, or OpenEnterprise software, in combination
with Bri sto l B ab co ck ’s

Open Bristol System Interface (Open BSI)

software. The
HMI/SCADA software at the Network Master allows the operator to view what is
going on in the network through graphical displays, trends, or printed logs and
reports.3 Below the Network Master are the remote process controllers.

The controllers in the network are organized into a hierarchical structure of one or

1

In addition to these standard types, certain Bristol controllers support ’ custom’ protocols for
communication with various foreign devices, PLC’s etc. See the ACCOL II Custom Protocols Manual
(docu ment# D4066) fo r detail s.

2

Some non-Bristol devices can be made to communicate via BSAP, provided that special
communication software programs called communication drivers are created for them.

3

Pseudo Master devices can be connected to lower levels of the network to view data. These are

similar to Network Masters, however, they are not considered to be ‘nodes’ in the network, and so do
NOT appear in th e NET T O P/NETDEF file s.

. The

Networ k 3000 Co mmunicati o ns Config uration Guide Page 1-1

Chapter 1 - Introduction

more

levels

. A BSAP network can support up to six levels (not including the
Network Master referred to as level 0.) The number of levels required varies
depending upon the size and scope of your project. Each controller is assigned an

address

specified using either the
BSI
called its

to allow data to be routed correctly. Addresses and network structure are

or the Open

NetView

Node Routing Table

program.

Network Topology (NETTOP) Program

4

The network information stored internally by a node is

, and is updated whenever an ACCOL load is

downloaded, or a valid time synchronization message is sent.
The level of a given controller specifies how many intervening

s there are between it, and the network master. The first level controllers are

line

called

top-level nodes

because data must travel over only 1 communication line to

communication

reach the Network Master. A communication line can consist of a direct cable
connection, a radio or satellite link, or a dial-up modem connection. Each
communication line is configured independently with its own

period, timeout

, etc.

baud rate, poll

Each controller (n ode) serves as a
immediately below, and as a

slave

master

to the nodes conn ected to it on the level

to the node connected to it on the level

immediately above. A node can have many slaves bu t only one master. Each master

its slaves for data, which it retains in memory until it is polled by

polls

4

The choice of whether NETTOP or NetView is used to configure your network depends on the
software in your system. If you are using DOS-based ACCOL Tools (AIC, ABC, etc.), or Open BSI
Utilities versions lower than 3.0, then NETTOP must be used. If you have Open BSI Utilities 3.0 (or
newer) then NetView must be used. If you want to use a mixture of older DOS-based ACCOL Tools
with Open BSI 3.1 (or newer) you will need to use NetVie w , an d th e n u s e th e NETBCX utility to
generate network files compatible with the DOS-based tools.

Page 1-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

its master

.

Chapter 1 - Introduction

In this way, data flows from slave to master, slave to master, etc. until it reaches
the Network Master, where it is made accessible to the operator via HMI software.
From a given node, direct

peer-to-peer communication

Master/Slave modules) is only possible to its Master node, any connected slave
nodes, and any siblings (nodes on the same level which share the same master). If
communica tio n i s req ui red to a n y n od e n ot in th ese categ ori es, it mu st be ro ute d u p
using Master/Slave modules at each individual level of the network, until it reaches
either the Network Master, or a Master which is a sibling to another Master. The
message can then be routed down, again, using Master/Slave modules at each level,
until it reaches the desired node.

What Type of Data Gets Polled in a BSAP Network?

When a BSAP Master pol ls its slave nodes for data, i t passes message traffic f rom
other nodes in the network, and collects response messages from its slaves. The
type of data collected includes alarms, RBE signal changes, audit trail, or array
data, etc. The various methods of BSAP network data collection are described
below:

(using ACCOL

Data
Collection
Method:

Alarm Data
Collection

Report By
Exception
(RBE) Data
Collection

What is this used for? What sort of configuration activities are

required?

Collecting analog alarm
and logical alarm signal
data, when these signals
enter or le ave an alarm
state.

Collecting RBE logical
signals which change
status, or RBE an alo g
signals which have a
value change exceeding a
specified deadband.

No user configuration is required, o th e r than
declaring signals as alarms in the ACCOL load.
Alarms will automatically be sent to the top- le vel
node(s). Configuration may be required, however,
within the H MI/SCADA pack age in order to retriev e
the alarms.
Signals must be declared as RBE sig nals. An RBE
Module must be configured in each ACCOL load
containin g RBE signals, an d e xpanded memory is
required . Compatible RBE Manag e r software mu st be
available for th e p articular Netw o rk Master
HMI/SCAD A so f tware.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 1-3

Chapter 1 - Introduction

Data
Collection
Method:

Remote
Data Base
Access
(RDB) Data
Collection

Template
Data
Collection

Peer-to­Peer
Communica
tion

What is this used for? What sort of configuration activities are

required?

Collecting signal, array,
signal list, arc hive or
audit trail data. RD B
access is used by Open
BSI, UOI, and other
programs specifically
written to use RDB
commands. Often this
type of collection occurs
based on user request, i.e.
calling up a DataView
Window, etc.
Collecting sig nal data
from all global signals, in
the network, at specified
intervals, and storing it in
the Enterprise Server
Real Time Data Base.
Sending signal and/or
array data from one
controller, to another
controller, so the other
controller can perform
some processing on it.

Configuration varies depending upon the program
being used.

Users will need to assign signals to Scan Time Classes,
config ure the Real Time Data Base and Template Data
Collection, create a Serial CFE Port in Top-Level
Nodes, and reserve expanded memory in the top-level
node for te mp late data.

Users must configure Master/EMaster and Slave
Modules in their ACCOL loads. NOTE:
Master/E Master Modules and Slav e Modules sh o uld
NOT be con f used with M aste r and Slave
communication described previously. They are
different.

IP Networks

The 386EX Protected Mode DPC 3330 / DPC 3335 controllers with PES03/PEX03 or

newer

are not part of the Network 3000 product line.

IP

exchange information with one another. IP is
protocol used on the
millions of people around the world use for business, government, and educational
purposes. The ’IP’, in fact, stands for

IP allows Network 3000 controllers to communicate using
type of

5

Bristol Babcock controller. The details of how IP works are beyond the scope of this manual,
however. One book on the subject we recommend is Internetworking with TCP/IP, Volume I:
Principles, Protocols, and Architecture by Douglas E. Comer, published by Prentice Hall.

Page 1-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

firmware sup port IP. ControlWa ve controlle rs also supp ort IP, howe ver, they

is a communications protocol which allows computers on

also

the standard communication

Internet,

local area network (LAN)

This manual will discuss certain aspects of IP as part of communications configuration for your

a constantly changing set of computer networks that

5

Internet Protocol

.

originally developed by Xerox Corporation.

different

Ethernet

networks to

, a standard

Chapter 1 - Introduction

(Other types of LANs may be supported in the future, however, Ethernet is
currently the only choice available.)

A Bristol Babcock network using only IP nodes does NOT follow a hierarchical
structure. For certain applications, this is a significant advantage over BSAP,
because all nodes in a given section of the network are essentially on the same
level. This means that peer-to-peer communication is simplified; there is no need to
use more than a single pair of IP_Client/IP_Server modules6 to get a message from
one node to any other node because all nodes are ’siblings’ on the same level.

Information on the structure of the network is stored in a Network Defi nition File
(NETDEF) at an Open BSI workstation called the

abbreviated as

7

file.

. A program called

NHP

NetView

is used to create the NETDEF

Network Host PC

, which is

If an IP node or an Open BSI workstation needs to communicate with

another

IP
node or Open BSI workstation, and it doesn’t know the address of the IP Port for
that node or workstation, it obtains the necessary addresses and routing
information from the NETDEF file at the NHP.

6

IP_Client and IP_Serve r pr ov id e IP t h e same node-to-node d ata t r an s f e r c apabilitie s as the Master

and Slave modules do for BSAP networks.

7

Although any Open BSI 3.0 workstation can potentially serve as an NHP, the ’host function’ refers
to Open BSI 3.0 workstations which c on tain d e taile d r ou t in g in for mation (in th e ir NETDEF file) for
particular RTU’s, and therefore are ’hosting’ them. Open BSI 3.0 workstations which do NOT store
detailed routing information for a particular RTU are considered to be proxy workstations with
respect to that RTU because they initially establish communication with it by requesting access

through that RTU’s NHP.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 1-5

Chapter 1 - Introduction

The concept of the NHP is easier to understand if you consider an analogy to the
public telephone system. Most people remember a certain set of phone numbers for
people they call frequently, but occasionally, they need to call someone whose
number they don’t know, so they call directory assistance and ask for the correct
phone number. The NHP performs the exact same function as the directory
assistance operator; except instead of giving out phone numbers, it provides IP
address information, on request, for connections to any node in its section of the
network.

Page 1-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 1 - Introduction

What Type of Data Gets Transmitted in an IP Network?

The type of data transmitted in an IP network includes alarms, RBE signal
changes, audit trail, or array data, etc. The various methods of IP network data
collection are described below:

Data
Collection
Method:

Alarm Data
Collection

Report By
Exception
(RBE) Data
Collection

Remote Data
Base Access
(RDB) Data
Collection

Peer-to-Peer
Communication

What is this used
for?

Collecting analog alarm
and logical alarm signal
data, when these signals
enter or le ave an alarm
state.

Collecting RBE logical
signals which change
status, or RBE an alo g
signals which have a
value change exceeding
a specified dead band.

Collecting signal, array,
signal list, arc hive or
audit trail data. RD B
access is used by Open
BSI and other programs
specifically written to
use RDB commands.
Often this type of
collection occurs based
on user request, i.e.
calling up a DataView
Window, etc.
Sending signal and/or
array data from one
controller, to another
controller, so the other
controller can perform
some processing on it.

What sort of configuration activities are
required?

Signals must be de clared as alarms in the ACCOL
load. Up to four diff e r e nt destination IP addresses
may be specified fo r receiving alarm data from a
particular node. Additional configuration may be
required, however, within the HMI/SCADA package
in order to retrieve th e alarms.
Signals must be de clared as RBE signals in the
ACCOL Load. An RBE Module must be configu red
in each ACCO L lo ad containing RBE signals. Up to
four different destination IP addresses may be
specified for receiving RBE data from a par ticular
node. Compatible RBE Manager so f tware must be
available for th e p articular Netw o rk Master
HMI/SCAD A so f tware.
Configuration varies depending upon the program
being used.

Users must configure IP_Client and IP_Server
Modules in their ACCOL loads.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 1-7

Chapter 1 - Introduction

Converting NETTOP Files for Use Below an NHP

It is possible to add a Network Host PC (NHP) on top of an existing BS AP n etwork,
replacing the existing Network Master node. To preserve the n etwork configu ration
data, the existing NETTOP files can be used to generate a single NETDEF (*.NDF)
file, using the NETCONV utility.

Any BSAP nodes which are to exist underneath IP nodes (instead of NHPs) must be
explicitly added in NetView, or added manually to the NETDEF file, using a text
editor.

The figure, below, shows a typical configuration which mixes both IP nodes, and
BSAP network(s).

Only the single BSAP node underneath the IP node would have to be manually
defined in NetView; the other BSAP node definitions (under the NHP) would be
extracted from the existing NETTOP file, and added to a new NETDEF file, using
the NETCONV utility.

Page 1-8 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 1 - Introduction

Overview of Configuration

Whether you are using BSAP, IP, or a combination of the two, the goal of
configuring network co mmunications is to achieve maximum throughput of data. To
obtain this performance goal, however, the user must have an understanding of
how Network 3000 communications work. We can think of the Network 3000
architecture as being made up of three major parts. These p arts are the
the

network routing

Each of these parts is configured indep endently, however, they all work together to
allow communications to function. Failure to properly configure any of these parts
can degrade system performance (slow down communication throughput) or can
even prevent any communication from occurring. It is important, therefore, to
understand how these three parts of the Network 3000 architecture are configured,
and h ow they interact.

information, and the

application

s which use the data link.

data li nk

,

Data Link

The

data li nk

data messages from node to node in the network. It includes the communication
ports, as well as the connections between them such as cables, modems, or radios. It
simply mo ves data messages; withou t regard to co ntent of the messag es. Th ere are
numerous software configuration parameters which control how the data link
functions. These subjects will be discussed in detail in Chapters 4 and 5.

or communication line handles the actual physical transmission of

Network Routing Information

Network routing information is used to determine where data messages should be
sent within the network. This is primarily determined based on entries made in the

NETDEF

will be discussed in detail in Chapters 2 and 3.

(or

NETTOP)

file inclu ding node names, addresses, etc. These subjects

Applications

Application

send messages out on the data link, and wait for responses to come back.
Applications include Open BSI utilities such as DataView, associated SCADA or
HMI packages such as OpenEnterprise, Intellution® FIX® or Iconics Genesis, as
well as Master Module communications in the controller.

s are the programs or system functions which

use

the data link. Th ey

There are several things to be aware of when creating your ACCOL l oad, as well as
certain application level parameters, which must be configured. These subjects will
be discussed in detail in Chapter 6.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 1-9

Chapter 1 - Introduction

Is this still unclear?

If these concepts are still unclear to you, it might help to think of the
communications system as analogous to using a postal delivery service. Let’s say
you want to order some merchandise by mail. You place a letter (data request) in an
envelope, and you write an address on the envelope indicating where you want it
delivered. (This is equivalent to defining network routing information.) Next you go
to your local post office, and drop your letter in the outgoing mailbox.

The postma n re trie ves you r l etter, al on g w ith l ots of oth er p eo pl es’ l etters, an d th ey
get sent ou t throug h the po stal system to ma ny di fferent p ost offi ces (nod es) in the
network. The mechanism of transit could involve delivery trucks, trains, planes, etc.
(This transportation system is similar to the concept of the data link, except instead
of trucks and planes the data link uses modems, cables, and radios.)

Finally, if all goes well, the mail order company will pick up your letter, from its
local post office, and will respond by sending back a package, addressed to you,
containing whatever merchandise you ordered. The response might actually be sent
as one package, or several packages at different times, if not all of the requested
items were ready at one time. The post office then delivers the return package(s) to
you. Both you and the mail order company are users (applications) which make use
of the postal system (data link).

The actu al time it tak es to complete th e full transacti on may vary, so mewhat. Fo r
instance, while the postal system might be able to guarantee that, under normal
circumstances, your request will reach the post office nearest the mail order
company within three days, there’s no guarantee that the mail order company
(application) will pick it up on time, that it will understand your request, or that it
has the items you want in stock. This is entirely out of control of the postal system
(data link).

Of course, if the postal truck carrying your letter gets into an accident (data link
failure) your request might not get delivered at all, and so you will have to send
another one. Similarly, if you write the wrong address (network routing
information) on the outgoing letter, the postal system doesn’t know the diffe rence,
and will try to deliver it to the non-existent address. Eventually, you’ll receive a
note from the post office saying that your letter (data request) was undeliverable.

If this analogy didn’t clarify the distinction between the data link, application, and
network routing information, don’t worry. We’ll talk more about these subjects in
the next several chapters.

Page 1-10 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 2 - Guidelines For Setting Up BSAP Networks

In this section, we will discuss, briefly, the terminology used in BSAP network
configuration, and then go over some important items, which, if overlooked, can
cause problems. For full details on using the Network Topology (NETTOP) program
as well as the alternative NETBC and NETREV compilers, see the

NETBC User Manual

Version 3.0 (or new er) you MUST use the Network Wizard in NetView,
the NETTOP, NETBC, and NETREV programs. NetView is discussed in the

BSI Utilities Manual

What is BSAP?

BSAP stands for Bristol Sy nchronous / Asynchronous Protocol. All ACCOL-based
Network 3000-series controllers can communicate using BSAP.

Each node (RTU) in a BSAP network is assigned a 7-bit local address (from 1 to

127) and a 15-bit global address, based on its location in the network. The local
address you configure in the NETTOP or NETDEF files for a particular RTU must
match the loca l address hardware switch setting (or FLASH pa rameter) configured
at the RTU.

(document# D4057). If you are using Open BSI Utilities

(document# D5081).

NETTOP and

instead of

Open

A Quick Overview of BSAP Network Terminology

Networ k 3000 Co mmunicati o ns Config uration Guide Page 2-1

Chapter 2 - Guidelines For Setting Up BSAP Networks

Node
Name:
PC 0 Not applicable DPC1, DPC2,

Network
Level:

Local Address: This node is a

master of:

This node is a
slave of:
Not applicable

DPC3
DPC1 1 1 DPC4, DPC5 PC
DPC2 1 2 Not applicable PC
DPC3 1 3 DPC6 PC
DPC4 2 1 DPC7 DPC1
DPC5 2 2 Not applicable DPC1
DPC6 2 1 Not applicable DPC3
DPC7 3 1 Not applicable DPC4

At the top o f a BS AP n etw ork is a comp u ter ca ll ed th e

network maste r

. Typically,
this is a PC Workstation.1 Going o ut through the communication p ort (e.g. COM1:
or COM2) of the network master is a communication line. The communication lines
could be direct connections, as shown in the drawing, or they could use modems
(dial-up), or they could use radios. It doesn’t matter for purposes of this discussion.

Below the PC workstation are controllers (

node name
levels

. A node’s level indicates how many communication lines there are between

. The nodes are organized into a hierarchical structure of 1 or more

nodes

), each of which has a unique

it, and the network master. The number of levels varies depending upon the
requirements of your particular system, and how you decide to lay it out.

Any data message from one node in the network to another node in the network is
considered to be

local

than 1 communication line is involved, a message is considered
The nodes on level 1 are called

network master. Another term f or top leve l node is
nodes, in turn, are called
Level 2. The Level 2 nodes are

if there is only one communication line involved. If more

.

. The top level

top level nodes

master

nodes to the nodes immediately below them on

slaves

of the top-level nodes, and are masters to any

global

because they are directly below the

first level slave

nodes connected to them on Level 3, and so on. NOTE: Any slave node can only
have 1 master.

Up to six levels (1-6) are allowed in the network, not counting the Network Master
on Level 0.

Each slave node
ranges from 1 to 127.

1

If this BSAP network is a sub-network of an IP node, the Network Master would be the Network

Host PC (NHP), and the IP node would be the only level 1 node.

2

A node’s configured local address may be viewed on-line through the #NODEADR.. system signal.

of a given master

2

is assigned a unique

local address

, which

Page 2-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 2 - Guidelines For Setting Up BSAP Networks

Things to Be Aware of When Defining Your Network

Every node in the system, as well as software in the network master, and any
pseudo masters (to be discussed later), depends on the data defined in NETTOP (or
NetView ) to rou te mess ag es. It’s essential that you plan out your network carefully,

before you attempt to define it in the Network Topology Program (NETTOP)

NetView.
Ideally, you should define your NETTOP/NetView information once at the

beginning of your system configuration, and then leave it alone. Although a small
network may no t be a ma jor pro blem, if y ou ha ve a l arge co mpli cated netw ork, an d
attempt to modify its layout late in the configuration process, you are likely to
endure a heavy cost in time and effort. Such a change may require you to re­generate data bases, re-download ACCOL loads, etc.

The best way to plan it out is to draw a picture of your network. You can draw one
similar to the figure at the beginning of this chapter, except with simple boxes
representing each node. You should show the communication lines connecting the
nodes, the network levels, and each node’s local address.

, or in

The following pages discuss some items to consider when setting up your network:

General Considerations - Applicable to Both NETTOP & NetView Users:

Some HMI/SCADA packages (such as Iconics Genesis for DOS) DO NOT support

•

peer-to-peer communication (discussed later in this manual) between level 1
nodes. Therefore, if peer-to-peer communication is required between two such
nodes, they cannot both be on level 1, or a different HMI/SCADA package must
be used.

Try to define the maximum

•

expected size of your network,
allowing room for any planned
future expansion. Defining some
extra nodes at the end which are
currently non-existent will not
affect system performance,
provided polling for them is turned

3

off.

Figur

Defining Network Size in NETTOP

3

Polling will be discussed in a later chapter. Note that program size in the master node of non-

existent slaves will be affected because extra #NODE.nnn signals will be created.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 2-3

Chapter 2 - Guidelines For Setting Up BSAP Networks

Defining Network Size in NetView

Choose your node names carefully. NETTOP users have only four characters to

•

work with. Because most software in the network master computer (including
Open BSI, and SCADA/HMI packages) use node names in data base generation,
you must avoid changing the node names later on, or else you will have to re­generate these data bases, or re-do other configuration. If you are using
NetView, up to 16 characters may be used for the node name, however, be aware
that not all HMI/SCADA packages, nor earlier versions of ACCOL and Open BSI
tools support the longer node names, therefore, if your system includes a
mixture of older and newer software, you should not exceed the 4 character
limit.

Page 2-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 2 - Guidelines For Setting Up BSAP Networks

A master node can have (using any or all of its ports) a total of 127 slave nodes

•

on the level below it. If more than this number is needed, the user can consider
using a technique called expanded node addressing (also known as expanded
BSAP or EBSAP) which is discussed fully in the
(document# D4044). Expanded node addressing however, introduces certain
complexities into the network which must be dealt with carefully, and is not
appropriate for all systems. The user should fully understand the implications of
using expanded node addressing, before attempting to use it. In particular, if
you are defining an EBSAP communication line within Open BSI 3.0 (or newer)
the virtual nodes on that communication line must be on Level 1 of your
network, and the expanded addressin g slave nodes must be on Level 2 of your
network.

As you plan your network, you will need to decide whether or not you want to

•

use data concentrator nodes. Data concentrator nodes are master nodes which
take data from one or more slave nodes, using peer-to-peer communication
(discussed later in this manual) and perform some sort of calculations on that
data; they may or may not have their own local process I/O. Data concentrators
are useful, for example, if control decisions must be made upon data from
several slave nodes. In the illustration on page 2-1, there are 3 nodes which
could potentially serve as data concentrator nodes: DPC1, DPC3, and DPC4.
NOTE: If not configured for peer-to-peer communication, these nodes would
perform whatever local I/O and control they have been configured for, and
simply serve as communication relay devices for their slave nodes, without
making control or processing decisions based on slave node data.

ACCOL II Reference Manual

The local address for a

•

node is assigned using
jumpers/switches on the
unit (or f or some mod el s,
software switches.) Th at
same address must be
entered for the node in
NETTOP or NetView.
DO NOT use the local
address as a means of
identifying a station
name.

Assigning Local Address in NETTOP

For example, if you are designing a network which has a node at each of three
pump stations called pump station #13, #25 and #36, don’t set the addresses of
your three nodes to 13, 25, 36. Local addresses should be sequential, and should
always start with address 1. Assigning these values randomly, with gaps in

Networ k 3000 Co mmunicati o ns Config uration Guide Page 2-5

Chapter 2 - Guidelines For Setting Up BSAP Networks

between, or using your own numbering scheme, imposes unnecessary overhead
in the master node (unusable extra #NODE.nnn. signals are created). Start with
address 1, then use address 2, 3, etc.

Entering the Local Address in NetView

Each and every node4 in your network must be specified in NETTOP/NetView.

•

Be sure to specify the correct .ACL file base name in NetView or in the

Load File”

field in NETTOP. This must be the actual .ACL file running in the

“Node

controlle r. Alth ou gh not requ i red , i t mig h t he lp yo u to us e th e sa me n ame as th e
node name. That way, you won’t forget which file goes in which node. NOTE:
Entering an incorrect node load file name will prevent Open BSI, and certain
other HMI/SCADA packages from functioning correctly.

4

If your system includes redundant controllers, each redundant pair is considered to be a single

node, since only 1 of the 2 can be th e on -lin e u n it at an y on e time.

Page 2-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 2 - Guidelines For Setting Up BSAP Networks

Defining the ACCOL load file name and other parameters in NetView

Considerations For NETTOP Users or Users With Open BSI 2.3 (or earlier):

When you define the Network Master node, in NETTOP, the choice for the type

•

of Network Master is simply a textual description. It doesn’t matter what you
choose. The Off Scan/On Scan field is currently unused, and the Alarm Zone is
only applicab le f or certain HMI p ack ages su ch a s Enterp rise S erver, or th e old er
Trolltalk VAX system.

Additional PC’s (laptops running Open BSI or UOI, for example) are often

•

connected at lower-levels of the network, through a controller’s Pseudo Slave
Port, or Pseudo Slave with Alarms Port. These PC’s are called pseudo masters.
They are NOT considered to be part of the network and so they MUST NOT be
defined in NETTOP. Pseudo Masters must, however, have current identical
copies of the NETTOP files in order to perform global communication.

When you release a new network file using NETTOP, you actually generate

•

three files, named

NETFILE.DAT, RTUXREF.DAT,

and

GLADXREF.DAT

. These

Networ k 3000 Co mmunicati o ns Config uration Guide Page 2-7

Chapter 2 - Guidelines For Setting Up BSAP Networks

are your NETTOP files, which will be referenced by Open BSI, and other
HMI/SCADA software in order to route data through your network. In addition,
data from these files is used to create a structure called a
The node routing table is transferred to a node through a time synchronization
operation, or through a download.

These files are critical to the operation
of your system; they must exist on
your Network Master computer (the
operator workstation). If you have
other computers (Pseudo Masters)
connected to nodes, they must also
have identical copies of these NETTOP
files. On an Open BSI workstation
(Version 2.3 or earlier), the location
where the system should look for these
files must be specified in the Open BSI
Setup Tool.5 (See picture at right.)

Node Routing Table

.

If you are communicating with a portion of the network, i.e. you are not

•

connected to a top-level nod e, bu t are i nstead conn ected local ly to a Sl ave Port of
a lower level node, be sure you specify the correct node name when establishing
communications in preparation for a download. In Communication Setup Menus
for DOS-based tools, if you specify the local address instead of the n ode name,
prior to attempting a download, the wrong Node Routing Table will be sent.
Similarly, if you are using Open BSI 2.x for this same purpose, configure the
line as “Local” in the Communica tion Li ne Pa rameters d ialo g box a nd b e sure to
specify the correct node name in the Downloader. In both cases these problems
can be avoided by downloading globally from the top of the network.

5

If you are using UOI, or th e old e r D O S-based tools (on -lin e A IC, Toolkit, T askspy) th e loc ation
where these programs look for the NETTOP files is whatever directory serves as the default ACCOL
directory, usually C:\ACCOL. The default ACCOL directory for DOS-based tools may be changed

using the command ‘SET ACCOL=\directory_spec\’ where directory_spec is the DOS directory
containing the network files. This command can be entered directly at the DOS prompt, or may be
included in the A UTOEX EC.BAT file of you r PC.

Page 2-8 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 2 - Guidelines For Setting Up BSAP Networks

Considerations For NetView Users ONLY (Open BSI 3.0 or newer):

Additional PC’s (laptops running Open BSI, for example) are often connected at

•

lower levels of the network, through a controller’s Pseudo Slave Port, or Pseudo
Slave with Alarms Port. These PC’s are called pseudo masters. They are NOT
considered to be ‘nodes’ in the network and so they MUST NOT be defined in
NetView. If they are connected to BSAP nodes, and these PC's must perform
global communi cation, they must have cu rrent identical
file. If local communication only is required, then LocalView may be used
instead. If the PC's are instead connected to IP ports on IP nodes there is no
need for the NETDEF file since the PC can request it from the Network Host PC
(NHP).

In Open BSI 3.0 (or newer), you can add a

•

single BSAP network underneath the
Network Host PC (NHP). The NHP then
serves as the BSAP Network Master node.

copies of the NETDEF

Adding A BSAP network

If you have an existing BSAP network defined via NETTOP, and you plan to

•

upgrade your Open BSI to Version 3.0 (or
utility to convert the NETTOP files into a Network Definition File (*.NDF) and
then run the NetDef to Database utility to generate NETDEF database file(s)
compatible with Open BSI 3.1 (or newer). The BSAP network will then be
known to the Network Host PC (NHP). (Otherwise, you would have to re-enter
all of the NETTOP information within NetView.) Instructions for using
NetView, NETCONV, and the NetDef to Database utility are included in the

Open BSI Utilities Manual

(document# D5081).

newer

), you must run the NETCONV

Networ k 3000 Co mmunicati o ns Config uration Guide Page 2-9

Chapter 2 - Guidelines For Setting Up BSAP Networks

In Open BSI 3.0 (or newer) you can

•

only have one single BSAP network
directly underneath the Network Host
PC (NHP). Additional BSAP nodes
however, can be added underneath one
or more IP nodes, by the addition of
one or more BSAP sub-networks. In a
BSAP sub-network, the NHP is the
Network Master node, and the IP nod e
to which the BSAP sub-network is
added is
BSAP nodes can only exist on levels 2
through 6 of the BSAP sub-network. It
is important, therefore, that when you
specify a local address for the level 1
IP node, you take care not to define a
large number, since that woul d reduce
the number of bits available to
represent addresses for the BSAP
nodes on level 2 through 6.

the only level 1 node

. Purely

Adding A BSAP Sub-Network underneath an IP

node

Setting the Local Address For the IP node on Level 1

Page 2-10 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 3 - Guidelines For Setting Up IP Networks

In this section, we will discuss, briefly, the terminology used in IP network
configuration, and then go over some important items, which, if overlooked, can
cause problems. Full details on using the NetView and LocalView programs are
included in the

A Discussion of IP Network Terminology

Open BSI 3.0 was the first version of Open BSI to support IP. IP is a standard
communications protocol for transmitting data over a computer network. It also
allows computers on

also

IP is
changing set of computer networks that millions of people around the world use for
business, government, and educational purposes. The ’IP’, in fact, stands for

Internet Protocol

the standard communication protocol used on the

Open BSI Utilities Manual

different

1

.

networks to exchange information with one another.

(document# D5081).

Internet,

a constantly

Any Bristol Babcock Network 3000-series controller which supports IP
communications will be referred to as an

IP node

configured with at least one network connection called an

2

.

Each IP node must be

IP port

which will allow
communications via Ethernet. Information on defining IP ports is included in
Chapter 5.

Each IP port has an associated

IP address

. Because a particular IP node can have

more than one IP port, it can also have more than one IP address.
Each IP node also has an associated

sub-net mask

which is used in
communications routing decisions. These subjects will be explained later in this
chapter.

1

This section will discuss certain aspects of IP as part of communications configuration for your
Bristol Babcock controller (RTU). The details of how IP works are beyond the scope of this manual,
however. One book on the subject we recommend is Internetworking with TCP/IP, Volume I:
Principles, Protocols, and Architecture by Douglas E. Comer, published by Prentice Hall.

2

Currently, the only Network 3000 series controllers to support IP communication are the 386EX

Protected Mode versions of the DPC 3330 and D P C 3335 with PES03/PEX03 or newer fir mware.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-1

Chapter 3 - Guidelines For Setting Up IP Networks

There is no hierarchical stru cture enforced in a pure IP network; all nodes may be
on the same level.

An IP network can, however, have multiple sub-networks, including BSAP

networks

network levels. (For more information on BSAP networks, see Chapter 2.) There are
numerous other possible IP configurations including routers, multiple networks,
etc., however, these will not be discussed here.

Every IP network must have at least one

; the BSAP portions of which, must follow the standard BSAP concept of

Network Host PC (NHP)

.

sub-

What is A Network Host PC (NHP)?

A Network Host PC is any Open BSI 3.0 (or newer) workstation. Typically, this
workstation has some Network 3000-series remote process controllers (RTUs)
connected to it (it serves as the ’host’ for those RTUs). Those RTUs must be defined
in a Network Definition (NETDEF) File at this NHP. Any other NHP can only gain
access to these RTUs if access is granted by this NHP. NOTE: An Open BSI 3.0
Workstation without attached RTUs is still technically considered to be an NHP,
even though it isn’t ’hosting’ any RTUs.

Page 3-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 3 - Guidelines For Setting Up IP Networks

How Do NHPs Work?

In a BSAP or EBSAP network, the
Network Host PC (NHP) performs the
function of the network master node
or level 0 node; it polls all Level 1
controllers (RTUs) for data.

For an IP network (and also BSAP or
EBSAP networks) the Network Host
PC (NHP) serves as a central location
for obtaining address information
about RTUs and workstations in this
portion of the network. This
information is stored in the Network
Definition (NETDEF) File, generated
by NetView.

Any Open BSI 3.0 workstation, whether or not it has attached RTUs of its own, can
communicate with any RTU in the netw ork, provided it has been granted access to
the RTU, by that RTU’s NHP.

The concept of the NHP is easier to understand if you consider an analogy to the
public telephone system. Most people remember a certain set of phone numbers for
people they call frequently, but occasionally, they need to call someone whose
number they don’t know, so they call directory assistance and ask for the correct
phone number. The NHP performs the exact same function as the directory
assistance operator; except instead of giving out phone numbers, it provides address
information, on request, for connections to any node in its section of the network.

In addition, following our same analogy, the directory assistance operator can
generally do one of two things when you call for inf ormation. If you are unable to
call someone, one thing the directory assistance operator can do is establish a

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-3

Chapter 3 - Guidelines For Setting Up IP Networks

connection for you. This same concept applies with respect to NHPs. If a particular
Open BSI Workstation needs to communicate with an RTU associated with another
NHP, the other NHP can relay messages between its RTU and the inquiring
workstation. This is called proxy access, and is depicted in the figure, below:

Page 3-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 3 - Guidelines For Setting Up IP Networks

The second thing that can happen when you call for information is that the
directory assistance operator just gives you the number of the person you want to
call and says "call them yourself, directly!" This also applies in the case of NHPs.
An NHP can grant proxy direct access to a workstation req uesting access to one or
more of its RTUs. That w orkstation can then contact the RTU directly, without
having to pa ss messa ges through the RTU’s NHP. NOTE: Proxy direct access is
only possible in IP networks, and requires the definition in the Comm Line Wizard
(at the workstation requesting access) of an IP communication line for the proxy
RTUs. Proxy direct access is depicted in the figure, below:

What Are the Advantages of Using NHPs?

All any Open BSI 3.0 (or
to communicate with an RTU is the IP address of the NHP for that RTU, and the
RTU name. This simplifies network configuration because if the address of an RTU
should change, for any reason, only its NHP needs to be notified. Any other
workstation can find out from the NHP.

Also, because you can have multiple NHPs, each of which is responsible for a
portion of your network, your supervisory control is truly distributed among
multiple sites. Through proxy access, any Open BSI 3.0 (or

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-5

newer

) workstation in the network needs to know in order

newer

) workstation can

Chapter 3 - Guidelines For Setting Up IP Networks

communicate with RTUs belonging to any NHP in the network, provid ed that it
knows the name of the RTU, and the IP address of the NHP for that RTU.

What is the Format of IP Addresses?

Each network connection from an IP node has an

IP address

which is unique

within the ne twork. It is important to note that the IP address is ass ociated with

network conn ection

the
single IP node to have

(IP Port) on the node, NOT the node itself. This allows a

more than on e

IP port, and consequently, more than one IP

address.
IP addresses consist of 32

(1’s and 0’s) which are divided up into 4 groups of 8

bits

bits each. A period is used to separate each group. Each group of 8 bits is then
converted from binary to a decimal number from 0 to 255. The resulting IP address
is said to be in

dotted decimal

notation.

Each of the numbers in the address generally has a specific meaning. The IP
address is generally divided up into a
each node in the network, and a

local portion

network portion

of which some part

which must be common to

must be unique

to a particular node.

How is the Specific Meaning of Each Part of the Address Defined?

Addresses must be assigned to be consistent with whatever conventions have been
established for your system.3 In addition, there are certain rules to defini ng
addresses, which will be discussed later.

The specific meaning of each part of the address is defined in something called the

sub-net mask

. The sub-net mask is simp ly another set of 32 bits (w hich mus t also

be converted to dotted decimal notation). Each bit in the sub-net mask corresponds

3

For example, if this network has connections outside the plant (i.e. a connection to the real world­wide Internet), then the choice of this network number is assigned by an Internet governing body
called the Network Information Center (NIC) or whatever Internet service provider you are using.

Page 3-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 3 - Guidelines For Setting Up IP Networks

to a bit in the IP address. If a bit i n the su b-net ma sk is set to 1 (ON) , then the
corresponding bit
of the IP address. The network portion can be ignored (or ’
performing communications to
definition, all nodes in the same network h ave ide ntical network p ortions. Any bit
in the sub-net mask which is 0 (OFF) is considered to be part of the local addressing
scheme.

The figure, on the next page, shows the IP address an d corresponding sub-net mask
for an IP address of 120.0.210.1 and a sub-net mask of 255.0.0.0.

in the IP address

nodes within the same network,

is considered to be part of the

masked

because by

network portion

’) when

As we said before, a ’1’ in the sub-net mask indicates that the corresponding IP
address bit is part of the network portion of the address. Because the first part of
the IP address ’01111000.’ has a corresponding sub-net mask of ’11111111’ we know
that ’01111000’ (120 in decimal) is the network portion of the address.

The remaining parts of the IP address ’00000000.11010010.00000001’ have a
corresponding sub-net mask of ’00000000.00000000.00000000’. These bits are used
as part of the local communications addressing scheme.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-7

Chapter 3 - Guidelines For Setting Up IP Networks

RULES FOR CREATING A LOCAL ADDRESSING SCHEME:

When you are creating your IP address, the network portion of the
address must appear
CANNOT define an IP address as 0.200.14.1. The network portion must
appear first. This means that when creating the sub-net mask, the
masked portion (i.e. all 1’s) must appear first.

The organization of the remaining bits can follow any local
communications scheme you choose to devise, except that each group of
bits that represents something

For example, let’s say the first 16 bits have been ’masked out’ to defi ne the
network address, i.e. there is a sub-net mask of:

11111111 . 11111111 . 00000000 . 00000000

first

. For example, if the network portion is 200, you

must be contiguous

.

which in dotted decimal format is:

255 . 255 . 0 . 0

That leaves 16 bits (indicated by the 0’s) for devising a local
communications scheme.

You might want to use the fi rst 8 bits to indicate a section or area number
for a section of your network. 8 bits will allow up to 256 sections to be
defined. Another 8 bits (remaining out of the 16 available) can be used to
indicate a node number, allowing up to 256 IP controllers (RTUs) and
Open BSI workstations, in a given section.

Page 3-8 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 3 - Guidelines For Setting Up IP Networks

Sub-net Masks Determine Which Nodes Are Reachable From This Node:

So far, we have been talking about the mechanics of creating IP addresses and sub­net masks. The aspect we have not discussed is why IP addresses and subnet masks
are so important.

A node’s IP address, and its sub-net mask, define the range of acceptable addresses
with which the node can communicate. For example, if one node has an IP address
of 4.3.2.1 and another node has an IP address of 100.100.0.1, there is no common
network portion between the two addresses. For that reason, there is NO way these
two nodes can communicate with each other directly - - they are each part of

different

one or more

For two nodes to communicate directly, the network portion of their addresses (as
specified by the sub-net mask) must match exactly.

To illustrate this concept,
look at the figure, at right.
The network shown has one
Network Host PC (NHP)
called NHP1, and 3
controllers (RTUs) named
OAK_STREET,
ELM_STREET, AND
WALNUT_AVE.

networks. Any messages between these nod es would have to pass through

router

computers.

The table, on the next page, however, reveals a problem with the configured sub­net masks.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-9

Chapter 3 - Guidelines For Setting Up IP Networks

NODE NAME IP ADDRESS,

SUBNET MASK:

NHP1 IP ADR: 100.22.49.1

MASK: 255.255.255.0

WALNUT_AVE IP ADR: 100.22.49.178

MASK: 255.255.0.0

OAK_STREET IP ADR: 100.22.50.33

MASK: 255.255.0.0

ELM_STREET IP ADR: 100.22.51.14

MASK: 255.255.0.0

Based on their specified IP addresses and sub-net masks, OAK_STREET,
ELM_STREET, and WALNUT_AVE can all communicate with each other. They can
also send messages to NHP1.

MASK SAYS THIS NODE CAN
SEND MESSAGES TO ALL
NODES WITH ADDRESSES:

100.22.49.
where

255.

100.22.
where

0 to 255.

100.22.
where

0 to 255.

100.22.
where

0 to 255.

yyy

yyy

is an integer from 0 to

yyy.zzz

yyy

yyy.zzz

yyy

yyy.zzz

yyy

and

and

and

zzz

are integers from

zzz

are integers from

zzz

are integers from

Page 3-10 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 3 - Guidelines For Setting Up IP Networks

There is a problem, however. NHP1
has a sub-net mask which specifies
that it can only send message s to
nodes with addresses

100.22.49.

nnn

where
integer from 0 to 255. The only
node which it can send messages
to, therefore, is WALNUT_AVE.

To remedy this situation, NHP1’s
sub-net mask

should be changed to

255.255.0.0 so that it can also send
messages to OAK_STREET and
ELM_STREET. The corrected sub­net mask is reflecte d in the figure
at right.

nnn

is an

Things to Be Aware of When Defining Your Network

If you have a small network which will never be connected to the world-wide

•

Internet, your choice of IP addresses is largely unrestricted. Even if you have no
plans to connect your network to the global Internet, however, the Internet
Engineering Task Force

recommends

, as per,

RFC 1918

private networks should be assigned from the following ranges:

10.0.0.0 to 10.255.255.255

172.16.0.0 to 172.31.255.255

192.168.0.0 to 192.168.255.255

These particular ranges of Internet addresses have been set aside for private
networks. Any messages coming from these addresses can be recognized by most
Internet Service Providers (ISP) as coming from private networks, and so can be
filtered out. This helps avoid addressing conflicts should an accidental
connection occur between a private network, and the global Internet.

*

RFC 1918 - Rekht er, et a l , Best Current Practice memo - Address Allocation for Private Internets,
Internet Engineering Task Force, February, 1996. Please see http://www.ietf.org for complete text
of this memo.

*

that IP addresses for

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-11

Chapter 3 - Guidelines For Setting Up IP Networks

Once you have an outside connection, to another network, or the worldwide

•

Internet, your IP addresses MUST CONFORM to whatever conventions have
already been established. If you are intending to connect your Open BSI
network directly to the global world-wide Internet, you must obtain a range of IP
addresses from your Internet service provider (ISP) or from an Internet
governing body such as the Internet Assigned Nu mbers Authority (IANA).

If you have a small network, it may simplify your configuration to use the

•

sub-net mask throughout each sub-net.
The examples presented in this book generally involve small numbers of IP

•

nodes in order to better illustrate networking concepts. Other than the caveats
and rules discussed earlier, regardin g addressin g an d su b-n et masks, there i s no
restriction on u sers creatin g mu ch more ela bora te netw orks. Be a ware, h owe ver,

it is th e user’s responsibility

that
fully understand potential problems associated with large or unorthodox
network configurations. The examples and trouble-shooting notes in this book do
not attempt to encompass every possible user configuration.

We are discussing the Internet Protocol (IP) for use in communication between
Bristol Babcock controllers. The normal, intended application is for a ’closed
circuit’ internet (LAN) of controllers and worksta tions in a company plant or
industrial site. While, there is no built-in restriction against connecting an IP
network of Bristol nodes to the world-wide Internet, customers must be aware that
any external IP connection (no matter what brand of RTUs and software you are
using) poses potential risks.

well as default UDP/TCP socket numbers, to lessen the possibility that an
unaut horized intruder could gain access to your internal company process control
data.

access by plant operators, and should NOT be considered protection against
intentional malicious activity by a sophisticated intruder, i.e. professional ’hacker’.
Customers should consider purchasing commercially-available "firewall" software
to gain a further degree of protection against such malicious intrusion.

These Bristol Babcock security features are designed to prevent accidental

to configure and support such networks, and to

WARNING

Customers are urged to change default passwords, as

same

If you have a redundant pair of controllers; each must have a

•

for its IP Port.
If your IP network includes a BSAP sub-network underneath an IP RTU, then

•

the Network Host PC (NHP) must serve as
network, and the IP RTU must be the only BSAP level 1 node.

Currently, IP nodes do NOT include support for dial-up modems, or radios,

•

though support for these may be included in a future release.

Page 3-12 Networ k 3000 Co mmunicati o ns Config uration Guide

network maste r

unique

of the BSAP sub-

IP address

Chapter 3 - Guidelines For Setting Up IP Networks

The IP address of the Network Host PC (NHP) is initially defined in the System
Wizard of NetView.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-13

Chapter 3 - Guidelines For Setting Up IP Networks

Alarm and RBE (Report By Exception) data is sent to destination IP addresses

•

defined in the Network Wizard of NetView.

Page 3-14 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 3 - Guidelines For Setting Up IP Networks

The range of valid IP addresses, and the sub-net mask for the communication

•

line are defined in the Comm Line Wizard of NetView.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-15

Chapter 3 - Guidelines For Setting Up IP Networks

The IP address for an RTU’s IP port is defined in the RTU Wizard of NetView.

•

Page 3-16 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 3 - Guidelines For Setting Up IP Networks

In the ACCOL load, each IP node must be configured with an

•

Although the IP address

•

information is defined in
NetView, the same information
MUST be configured in the
FLASH memory of each
individual RTU, using
LocalView. LocalView is used to
assign an IP address for each
Ethernet IP Port on the IP node,
and to specify the IP address for
the RTU’s associated NHP.

Once LocalView has been used to configure this information, the RTU must be
reset (or downloaded with a new ACCOL load) for the new p arameters to be
activated. NOTE: For more information on IP address configuration, see

’Chapter 5 - Guidelines For Configuring the IP Data Link’

IP Port

.

.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 3-17

BLANK PAGE

Chapter 4 - Guidelines For Configuring the BSAP Data Link

The

data link

network. This includes the port and communication line itself, an d for purposes of
our discussion, also includes the software configuration for the ports and the
communication line. The functioning of the data link is completely independent of
the applications which use the link (HMI/SCADA p ackages, Master Modules, etc.)
In order to understand how the
how data moves through a BSAP network.

refers to the actual mechanism by which data flows through the

data li nk

operates, it is necessary to understand

At a specified interval, called the
all of its slave nodes. This request for data is called a
if a node on level 2, has a Master port defined, with ten slave nodes on level 3, this
node will sequentially ask each of its ten slaves for data. If it makes this request
once every second, that means that the poll period for this Master Port is 1 second.

Each slave node has a slave
port, through which it
receives poll messages from
its master, and it responds to
the poll messages with d ata.
The slave port also h as a p oll
period defined, however, in
this case,

a totally different meaning

For slave ports, the poll period refers to the amount of time the slave node will wait
to hear from its master. If the slave node does not receive a poll message from its
master withi n this period of time, it discard s data messag es for the ma ster (except
for alarms, which are saved in buffers), because the master is assumed to be ‘dead’.

the poll p eriod has

.

poll period

, each master node requests data from

message. So, for example,

poll

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-1

Chapter 4 - Guidelines For Configuring the BSAP Data Link

The actual character-by-character movement of data is governed by the

baud rate

The baud rate specified on a master node’s master p ort must match the ba ud rate
for each slave on that communication line.

When all of these data link parameters are configu red correctly, data moves from a
slave n ode, to i ts master, to the master’s master, etc., until it reaches the network
master computer, where it may viewed by the operator, exported to HMI/SCADA
packages, etc.

This section will discuss data link parameters which specify how this movement of
data will occur. These parameters include:

"

Proper configuration of the Master Port or Expanded Addressing Master Port(s)
if this is a Master node including: baud rate, high slave addresses, timeouts, as
well as setting the proper poll period for the port(s), and defining the node array
(#NDARRAY..)

"

Proper configuration of the Slave Port, if this is a Slave node, including the bau d
rate and the poll period for the slave.

"

Proper configuration of the Network Master (usually a PC workstation)
including baud rates, retries, etc.

.

"

Special considerations concerning radios, modems, etc.

1

Setting Up the Master Port(s)

Master Port configuration should be performed prior to configuring other po rt types.
It involves defining the port in the ACCOL source fi le, setting the poll period for the
port, and setting up the #NDARRAY to control polling of slave nodes.

Master Port Definition in the ACCOL source File

A controller’s Master Port(s) are configured within the ACCOL source file.

1

If you are using one or more Expanded Addressing Master Ports, the same instructions generally
apply. Specific differences will be noted in footnotes.

Page 4-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Choice of High Slave Ad dresses:
One of the first things to be aware of is that if a controller will have more than 1

master port,

ascending order and MUST NOT OVERLAP.

the range of addresses for each master port’s slave nodes must be in

For example, let’s say we
have a 386EX Protected
Mode DPC 3330 which is
a Master to 34 Slave
nodes.

If Ports B, C, and G of
this controller are to be
configured as Master
Ports (see figure, above,
and at right) then the

“High Slave Addr”

parameters for each port
must be configured
properly.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-3

Chapter 4 - Guidelines For Configuring the BSAP Data Link

For Port B, the

Addr”

parameter is 10, this

“High Slave

means that the range of slave
node addresses for Port B is 1
to 10.

Similarly, if Port C’s

“High Sl ave Addr”

slave node addresses is 11 through 18, and if Port G’s

parameter i s 18, then Po rt C’s range of

“High Slave Addr”

parameter is 34, then its range of slave node addresses is 19 through 34.
Note also, in this configuration, that once these ports have been set up, and slave

connections established, etc., the user CANNOT, during future expansion, convert
Ports BIP1, BIP2, A or D to Master Ports without changin g the local addresses for
all of the intervening slave nodes. Ports H, I, and J, however, could be used for this
purpose.

2

Choice of Baud Rate:

“Baud Rate”

is the character by
character rate at which d ata can
be sent out through this port.
The available baud rate depends
upon the type of port, as well as
the characteristics of the under­lying link (cable, radio, dial-up
modem, etc.)

The most critical a spect of set ti ng th e ba ud rat e is ma ki ng sure t ha t t he b au d ra te of
each Master Port matches the baud rate of its associated slave node’s slave ports.

That is, if the Master Port is set at 9,600 baud, the Slave Port in each slave node on
that port must also be set at 9,600 baud. Bristol Babcock controllers

do NOT

‘fall

back’ to lower baud rates if they are unable to communicate at the specifi ed rate.

3

NOTE: Not all ports support all baud rates.

ACCOL II Reference Manual

the

2

The same rules apply for Expanded Addressing Master (EAMaster) Ports, except that the “High
Slave Addr” includes virtual slave nodes, and a “High EASlave Addr” field appears for specifying
the high slave address of a virtual node’s slave nodes. See ‘Expanded Node Addressing’ in the
ACCOL II Reference Manual (document# D4044) for details .

3

Although some ports configured as Masters support synchronous communication, synchronous
communication is NOT supported on any Expanded Addressing Master (EAMaster) Ports.

(document# D4044) contains details on which

The

‘Communicati on Ports’

section of

Page 4-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

ports support which bau d rates, as well as whethe r synchronous commun ication is
supported.

Choice of the Timeout value
The

“Timeout”

field for the Master Port specifies the amount of time (in tenths of
seconds) for which the Master Port will wait for a particular slave node to begin
transmission of a respon se message to the Master’s poll. The default value is 5 (0.5
seconds). If the slave node does not begin respond ing within that period of time, a

response timeout

For example, if the

is said to have occurred.

“Timeout”

is set to 7, it means th at from the time the Master

4

issues a poll to a slave node, it expects the slave node to begin transmitting a
response within 0.7 seconds. If the slav e node does no t begin respond ing within 0.7
seconds a response timeout is logged. The Master will make two

additional

attempts (f o r a tota l of thre e a ttemp ts) to commu n i cate w ith th e u n res po ns iv e sl av e
node.5 If there is still no response, that slave node is considered to be ‘dead’ and is
added to the list of ‘dead nodes.’6 During each polling cycle, the system will choose

one

node f rom the lis t of dead nodes a nd try to communi cate with it, in ca se it has
been ‘revived’. To completely turn off polling for a node, its entry in the node array
must be turned off. (See

Node Polling’

later in this section.)

‘Setting up t he Node Array (#ND ARRAY) to Control Slav e

If your data li nk uses rad ios with RTS -CTS del ays (to all ow time to key the radio)
then the

“Timeout”

value must be

greater than

the RTS-CTS delay time. If a
repeater is being used, you must add timeout time equivalent to two additional
delays (one delay for transmit, another delay for receive.) If you are using a satellite
or cellular phone system to dial-up nodes, you will also typically require an
extended value for the

“Timeout”

.

If you need to have a dynamic

"Timeout"

value, which may be changed on-line, you
may make use of the advanced master polling parameters, discussed later in this
chapter. See

4

For information on seeing whether response timeouts are occurring, see ‘Chapter 7 -

Troubleshooting Communication s’ later in this manual.

5

If your firmware revision supports it, and you have configured a Master Poll Control Signal List,
the number of attempts can be changed by altering the DATA MSG ATTEMPTS list entry. See
'Configuring Advanced Master Polling Parameters' on page 4-10 for details.

6

For a technical description of the polling algorithm, and the handling of ‘dead nodes’ see the
Networ k 3000 Co mmunicati o ns Application Pr o grammer’s Reference (document# D4052).

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-5

’Configuring Advanced Master Polling Parameters’

.

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Setting the Master Port Poll Period

As discussed, previously, each port has an associ ated p oll p eriod . For a Ma ster Port,
the poll period should be set to the time it takes, under normal operating
conditions, for this Master Port to poll each and every slave node on its line for data

once

-

. Note, also, that Master Port poll periods should always be defined

defining Slave Port poll periods.

prior

to

The poll period is set using #POLLPER.

nnn

system signals in the ACCOL load . The
choice of which poll period signal to use depends upon which port you are
configuring. The table, below, shows the correspondence between #POLLPER.

nnn.

signals and their associated ports.

If you are configuring the
poll period for this port…

You must use this poll period

signal:
Port A #POLLPER.000.
Port B #POLLPER.001.
Port C #POLLPER.002.
Port D #POLLPER.003.
BIP 1 #POLLPER.004.
BIP 2 #POLLPER.005.
Port G #POLLPER.006.
Port H #POLLPER.007.
Port I #POLLPER.008.
Port J #POLLPER.009.

One way to specify the Master Port poll period, is to use the worst case poll period
calculation, discussed below, as a starting point, and then tune the poll period, on­line, to achieve the optimum system throughput.

Worst Case Poll Period Calculation:
A ‘worst case’ calculation of the poll period is as follows:

7

Master Poll Period = 7,650 * (number of nodes on the line) / (baud rate)

So, for example, if a master port has 12 slave nodes, and the baud rate used is

7

For a more accurate calculation of the poll period, which takes into account message size, delays,
etc., users can obtain the POLLPER.BAT program from the on-line bulletin board of Bristol
Babcock’s Application Support G r ou p. Bristol’s Application Support Group can be c on t ac t ed using
the phone numbers listed on page ii of this manual.

Page 4-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

9,600, then a ‘worst case’ calculation of the poll period is:

7,650 * (12) / (9600) = 9.56 seconds

If lin e turnaround time is a factor (due to modems, or radios), this must be taken
into account when calculating the poll period. For example, if a modem/radio
combination requires 0.2 seconds after RTS assertion before CTS is asserted, then

0.8 * (number of nodes on the line) shou ld be added to the poll period calculated by
the formula above. 0.8 was chosen because it represents 4 times the RTS/CTS delay;
this amount of time is needed because there can be up to four messages per data
transaction.

The ‘worst case’ poll pe riod can be entered
as the initial value for the poll period
signal when creatin g the ACCOL load. The
value can then be altered, on-line, using
DataView, based on the tuning technique
described below.

Technique

For Fine-Tuning the Poll Period

Once the system is ‘up and running’ with the initial ‘worst case’ poll period, the poll
period
This must be done carefully, however, because making the poll period too short may
have a negative impact on performance.

may

be able to be shortened based on observed performance of the network.

8

To perform this ‘fine tuning’, the system must be operating normally. All nodes on
the line must be on-line and communicating, and there must be no non-existent
nodes.

8

Making the poll period too short w ill c au s e the s ys t e m to go as fas t as it c an, but c an r e d u c e
throughput by eliminating preferred polling and other aspects of the polling algorithm.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-7

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Go to the Master node which has the
port being tuned, and observe the
transmit (TX) and receive (RX) LEDs
which correspond to the particular
Master Port. Under normal operating
conditions; these LEDs should both be
OFF simultaneously between 30% and
50% of the time.

If they are OFF more frequently than that, the poll period is unnecessarily long,
and there is too much ‘dead time’. Call up the associated poll period signal in
DataView, and enter a smaller value to shorten the p oll period.9 This process may
be repeated until the OFF time for the LEDs is between 30% and 50%.

If the LEDs are ON almost continuously, the poll period

is too short

and must be

lengthened.

IMPORTANT

The Master Port poll period tuning process, described
above, is only appropriate when all slave nodes on the
line are operating. Attempting to use this technique
when there is polling for non-existent controllers (out­of-service, not yet installed, etc.) will generate incorrect
results.

Measuring the Rate at Which Nodes Are

Actually

Being Polled:

One way to get a general idea of how often a particular slave node is polled is to
establish communications with that node (from the level of its Master) via
DataView (or Toolkit, if applicable) and call up the #TIME.001. or #TIME.007.
signals. These represent time in seconds; the rate at which they are ch anging shows
the rate at which polling is actually occurring.

If the rate at which the seconds values change is significantly slower than the
Master Port poll period you defined, you probably set too small a value for the poll
period (i.e. you are trying to poll your nodes faster than the system will allow.)

NOTE: Mak e sure the refresh rates for DataVie w are not se t longer th an the poll
period or this technique will NOT work.

9

Some dead time is always required. It is used for processing alarms, download messages, etc.

Page 4-8 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Setting Up the Node Array (#NDARRAY) To Control Slave Node Polling

Users should always define a node array to allow polling for slave nodes to be
turned ON/OFF as necessary.

Turning ON/OFF polling is necessary for many reasons: If certain slave node
addresses have been reserved for fu ture expansion, or if an existin g slave node h as
failed, or has been taken out of service for maintenance or repairs, polling for the
affected slaves MUST BE TURNED OFF via the user-defined node array. Failure to
configure and use the node array will result in the system believing that ‘dead’ or
non-existent nodes are alive, and so it will continue to try to contact them.
Unnecessary communication retries, as well as excess overhead in the master n ode,
will degrade system performance significantly.

Creating the Node Array:
The ACCOL programmer

must create a Read/Write
Logical Data Array in the
controller containing the
Master Ports. The number of
rows defined must be equal to
(or greater than) the highest
slave address among all the
Master Ports in this
controller.

For example, if the ACCOL program we are crea ting is for a controller which h as
twenty slave nodes on two Ma ster Ports, with Master Port A covering addresses 1 to
10, and Master Port B covering addresses 11 to 20, then a twenty row by one
column logical read/write array should be created.

The number of the logical array (1 to 255) to be used is specified by the system
signal #NDARRAY..

10

Normally, the operator or ACCOL logic should leave each element set to ON, so that
communication with slave nodes can occur. If, for whatever reason, one or more
slave nodes are taken out of service (failure, main tenance, repairs, communication
problems) or if there are certain node addresses reserved for future system

10

If you enter a negative value for #NDARRAY.., the absolute value of the number you enter refers to
a signal list, which allows advanced users to define the node array and set additional polling
parameters, See ‘Configuring Advanced Maste r Pollin g Parameters’, later in this chapter, for
details.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-9

Chapter 4 - Guidelines For Configuring the BSAP Data Link

expansion, then polling for all of these non-existent or failed nodes should be
turned off using the node array. This prevents unnecessary communication
attempts b y the maste r to a non-existent node.

In the figure, shown on the previous page, the third slave node has
been struck by lightning and so has failed. The master controller
(which contains the node array) will continue attempts to
communicate with it, until, the third element in the node array
(corresponding to address 3) is changed from ON to OFF by the
operator, as shown at right.

For a full description of how to configure the node array11, see the #NDARRAY
portion of the

‘System Signals’

(document# D4044).

section in the

ACCOL II Reference Manual

Configuring Advanced Master Polling Parameters

newer only)

(PES03, PEX03, PLS 03, PL X03, LS 501, R MS 04, o r

Advanced users may choose to utilize certain additional parameters, which provide
greater control over Master/EAMaster Port polling. In order to use these advanced
parameters, a negative number must be entered on the #N DARRAY.. system signal.
The absolute value of this number is used to designate a signal list which will serve
as the Master Poll Control Signal List, instead of directly referencing the node
array.

11

If your Master node includes Expanded Addressing Master (EAMaster) Port(s) the #NDARRAY
includes virtual nodes as well as real nodes. Also, an additional node array, similar to #NDARRAY
must be created for each virtual node, to control polling for its slave nodes. See the ‘EAStatus
Module’ and ‘Expanded Node Addressing’ sections of the ACCOL II Reference Manual (document#
D4044) for details .

Page 4-10 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

The Master Poll Control Signal List contains the following items; the choice of
signal names used is entirely that of the ACCOL programmer.

NODE ARRAY number An analog signal containing the number of the node array

used to turn ON/OFF polling of Slave nodes. (This is
normally specified by the #NDARRAY.. system signal,
however, it is already used to reference the Master Poll
Control Signal List.) For instructions on configuring the
node array, see

SET SIZE An analog signal containing the number of configuration

signals per Master Port in
configuration signals need to be used; the number used is
determined based on the value of SET SIZE. If SET SIZE
is 7, all seven configuration signals must be configured
for each port (RESPONSE TIMEOUT, POLL MSG
ATTEMPTS, DATA MSG ATTEMPTS, IDLE POLL
ENABLE, DATA CARRIER DETECT, RASCL LINK,
RASCL SELECT); if SET SIZE is 2, only the first two
configu ration sign als (RESPONS E TIMEOUT, and POLL
MSG ATTEMPTS) must be configured, etc.

‘Creating the Node Array’

this

list. Not all of the

on page 4-9.

The remaining signals in the Master Poll Control Signal List are the configuration
signals. Each Master/EAMaster Port should be configured with its own set of
configuration signals, and they should appear in the list in the following order,
based on which p orts are defi ned (BIP1, BIP2, Port A, Port B, Port C, Port D, Port
G, Port H, Port I, Port J).

RESPONSE TIMEOUT this is the first configuration signal, and is available if

SET SIZE is greater than or equal to 1. The RESPONSE
TIMEOUT allows dynamic user control over the Master
Port response timeout. When configured, this signal
disables the response timeout value configured with the
port, and allows it to be specified dynamically via this
signal. In this con figuration signal, the respon se timeout
is the amount of time (in milliseconds,
seconds) that this Master Port will wait for a given Slave
node to begin transmission of a response to a poll
message. This value can range from 1 to 65,535
milliseconds. If the value is specified outside of this
range, it will be ignored, and the standard response
timeout, discussed on page 4-5 will be used

not

tenths of

instead

.

POLL MSG ATTEMPTS this analog signal specifies that number of times the

Master Port will attempt to send a poll message to its

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-11

Chapter 4 - Guidelines For Configuring the BSAP Data Link

slave nodes. The default is 1 attempt. With this signal,
the user can specify from 1 to 255 attempts. If a value
outside the 1 to 255 range is specified, the default will be
used. POLL MSG ATTEMPTS is useful, for example,
when using a system with radios. Since the window of
time in w hich the ra dio i s ON an d ‘listening’ may be very
short, it may be useful to issue several poll messages in
order to increase the chances of the radio being on when
the poll is sent.

DATA MSG ATTEMPTS this analog signal specifies that number of times the

Master Port will attempt to send a data message to its
slave n odes. The default is 3 attempts. With this signal,
the user can specify from 1 to 255 attempts. If a value
outside of the 1 to 255 range is specified, the default will
be used. DATA MSG ATTEMPTS is useful, for example,
when using a system with radios. Since the window of
time in w hich the ra dio i s ON an d ‘listening’ may be very
short, it may be useful to issu e several data messages in
order to increase the chances of the radio being on when
the data is sen t.

IDLE POLL ENABLE is a logica l signal wh ich, when se t ON, causes a dditional

polling to occur in a system where slaves dial into the
Master. For this to function, the DIAL_UP_ACK enable
signal in the Enhanced Slave Dial Control List must be
ON in the Slave nodes (see page 4-34.) This additional
polling occurs after all normal polling, and polling to see
if ‘dead’ nodes have been revived has been completed.
Under normal conditions, any extra time would leave the
port idle. When the IDLE POLLING ENABLE signal is
ON, however, the Master will issue a poll to Node 0 (a
non-existent node; there is no Node 0). Even though Node
0 is non-existent, a node attempting to dial in can respond
to the po ll via the DIAL_UP_ACK, and transmit its data.

Another usage for IDLE POLL ENABLE is in a radio
system where radios at the slaves are periodically turned
on to listen for messages from the Master. IDLE POLL
ENABLE allows radios to have a shorter listening time,
because messages are sent more freq uen tly, and the slave
nodes can therefore quickly determine whether or not a
message is intended for them, and if it is not they can
shut off the radios to conserve power.

Page 4-12 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

DATA CARRIER
DETECT ENABLE is a logical signal which, when ON, causes the Master

Port to hold off sending poll messages until the modem’s
carrier detect is ON. This is useful because issuing a poll
to a Hayes-compatible modem while it is still ‘negotiating’
the line would cause the modem to hang up, thereby
terminating the connection. DATA CARRIER DETECT
ENABLE can prevent this problem from occurring.

RASCL LINK reports which RASCL link is currently in use.12 If a logi cal

signal is used, OFF (0) indicates that RASCL Link 1 is in
use, and ON (1) indicates that RASCL Link 2 is in use. If
an analog signal is used, '1' indicates Link 1 is in use, and
'2' indicates Link 2 is in use. This signal is only applicable
if a baud rate of 1MEG RASCL is configured for this port.

RASCL SELECT is an analog signal which allows user control over which

RASCL link (1 or 2) is currently in use. Entering '1'
selects link 1; entering '2' sel ects link 2. Any other value
allows automatic link control. Note that the signal will
automati cally reve rt to automatic link control after a
single execution. If the user wants to keep the selection
on a particular link, this signal must be control inhibited,
thereby keeping RASCL in manual mode. This signal is
only applicable if a baud rate of 1MEG RASCL is
configured for this port.

An example Master Poll Control Signal List is shown below. It uses only the
RESPONSE TIMEOUT configuration signal for each port (because SET SIZE
equals 1).

#NDARRAY = - 7
SET.SIZE. = 1

*LIST 7
10 NDARRAY..
20 SET.SIZE
30 PORTA.RESP.TIMO
40 PORTC.RESP.TIMO
50 PORTD.RESP.TIMO

12

RASCL stands for Redundant Automatic Switching Communication Link, and is available on
certain ports and versions of the DPC 3330 and DPC 3335.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-13

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Setting Up the Slave Port(s)

13

Slave Port configuration should be performed
node has been configured. Configuration of the Slave Port includes defining the
port in the ACCOL source file, and setting the poll period for the port.

Slave Port Definition in the ACCOL Source File

A particular ACCOL source file can have
because, a particular Slave controller can only have one Master.

The first parameter to define
for the Slave Port is the

Rate”

. The value chosen must

“Baud

match the baud rate value for
the Master Port, in the
associated Master node.

after

the Master node of this Slave

one and only one

Slave Port. This is

14

So, for example, if node DPC3 is a Master to node DPC5, and DPC3’s Master Port
has a baud rate of 9,600, then DPC5’s Slave Port baud rate must also be 9,600.

Setting the Slave Port Poll Period

The Slave Port poll period

Master Port,

discussed previously. For a Slave Port, the poll period refers to a

has a tot ally d ifferen t meani ng th an the poll peri od on a

period of time that the Slave Node will wait to receive a poll message from its
Master Node. If this period of time expires without a poll request from the Master,
the Slave Node assumes its Master is ‘dead’ and so will discard all data messages
(other than alarms) collected up to that time, for the Master.

13

There are two other types of ports which can be substituted for Slave ports under certain
circumstances. Serial CFE Ports (instead of Slave Ports) should be used if this is a top-level node in
an Enterprise Server system. Another alternative is VSAT Slave Ports, however, these ports require
a corresponding VSAT Master which is currently only supported by Open BSI workstations. For
more information on Serial CFE and VSAT Slave Ports, see the ’Communication Ports’ section of

the ACCOL II Reference Manual (document# D4044).

14

One Pseudo Slave with Alarms Port, and several Pseudo Slave Ports may be defined for use by
Pseudo Master devices, but these ports do NOT serve the same function as the Slave Port. See
Appendix A for summary information on port types.

Page 4-14 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Under most ci rcumstan ces, the Sl ave Port p oll peri od shou ld be set to at least 3 to 5
times the rate at w hich th e Master node’s Master Port is actually polling this slave.

If the poll perio d for the Master Port wa s set accurate ly, thi s would b e 3 to 5 time s
the Master Port poll period. To get an idea of how fast polling is actually occurring,

Measuring the Rate At Which Nodes Are Actually Being Polled’

see ’
So, for example, if the Master Port poll period in the Master n ode is 20 secon ds, the

Slave Port poll period in the Slave node should be set to 100 seconds. The reason
the poll period is set this way is to allow for any communication difficulties or
retries which the Master node encountered while conducting its poll of all of the
other nodes.

on page 4-8.

If you set the Slave Port poll period to
#NODE.
destined for the Master node will be discarded. (A count of the number of messages
discarded can be viewed as the ‘Messages Aborted for Transmit Queu e’ statistic in
the Portstatus Module , or in the Open BSI Remote Commu nication Statisti cs Tool.)

The poll period is set using #POLLPER.
choice of which poll period signal to use depends upon which port you are
configuring. The table, on page 4-6, shows the correspondence between
#POLLPER.

There is one instance in which the Slave Port poll period shou ld be set to a small
value. This occurs if the Slave node has a critical control scheme which requires
timely updates of data from elsewhere in the network. In this case, the ACCOL
program must include logic which monitors the Slave Port’s #LINE.
identify a communication loss, and perform some sort of conditional logic based on
that occurrence. In this case, the Sla ve Port poll period w ould be set ‘tightly’ sin ce
the fresh acquisition of data is critical.

nnn

. alarms to be generated, in the Master node, and data messages

nnn.

signals and their associated ports.

too short a period of time

nnn

system signals in the ACCOL load . The

, it will cause

nnn

signal to

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-15

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Using Advanced Poll Period Parameters

(PES03, PEX03, PLS03, PLX03, LS501, RMS04, or
newer firmwa re only)

Advanced poll period parameters are available to:

"

Compensate for the turn-around time of messages in a 186-based 33XX
controller.

"

Provide quicker response from Slave nodes.

The advanced poll period parameters are accessible by placing a negative value on
the #POLLPER.

nnn

signal corresponding to the port you want to configure, and
setting up a signal list with signals representing the advanced parameters. The
absolute value of the #POLLPER.

nnn

signal value is the number of the signal list.

For example, if Port C is to be configured using advanced parameters, and the
parameters are to be in s ignal li st 7, then a value of ’-7’ should be assign ed to the
#POLLPER.002. signal.

The parameters must be entered in the signal list in the order in which they are
presented below. Each signal must be analog; the choice of signal names is entirely
up to the user’s discretion.

POLL PERIOD This is whatever the poll period for this port should be,

depending upon whether it is a Master Port or a Slave
Port.

386 DELAY TIME At higher baud rates, 186-based controllers sometimes

have problems responding quickly enough to a message
from a 386EX-based controller. This signal specifies a
delay time (in seconds) to allow a 386EX-based controller
to compensate for the message turn-around time in a 186­based controller. This may be used in either a Master Port
(386EX node is master to a 186 slave node) or on a S lave
Port (386EX node is a slave to a 186 master node). This
delay can range from 0 to 65.535 seconds with resolution
in milliseconds. A typical value for 386 DELAY TIME
would be 0.010 seconds (10 milliseconds).

IMMEDIATE RESPONSE
DELAY This parameter may only be used on Slave Ports; a

Master Port will ignore it. The IMMEDIATE RESPONSE
DELAY specifies a period of time (in seconds) that a Slave
node will wait, before responding to a request for data
from its Master. This delay can

actually speed up

Page 4-16 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

communication throughput

node time to get data ready for the Master, thereby
reducing th e number of ’ACK NO DATA’ messages to its
Master, and the subsequent additional data requests from
the Master. (An ’ACK NO DATA’ is simply the Slave
telling the Master that it received the request, but it has
no data yet to give.) This value can range from 0 to 2.55
seconds (with a resolution in milliseconds).

An example is shown below:

because it gives the Slave

Notes For Remote Process I/O Users

If you are using Remote I/O Racks (RIO 3331) connected to a DPC 3330, DPC 3335
or RTU 3310 host, an RIOR Port must be created. Configuration information on
these ports is included in the

Reference Manu al

RIOR Port, see the discussion on poll periods in the
same manual.

(document# D4044). For details on setting the poll period for an

’Communication Ports’

section of the

’System Sign als’

ACCOL II

section of the

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-17

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Setting Up Data Link Parameters At the Network Master

The previous sections of this chapter dealt entirely with setting up ports and polling
between controllers. This section of the chapter deals with ports and polling at the
Network Master device.

In a pure BSAP network, the Network Master device is the host computer (usually
a PC workstation) at the top of the network (Level 0). Its responsibility is to collect
data from the

top-level nodes

collection, many of the same types of configuration issues for controller Master
Ports apply. The Network Master has to perform polling, it has to do this at a
certain baud rate, etc. This section will highlight certain aspects of the
configuration of Network Master communications for the two Bristol Babcock
supplied Network Master software packages - - Open BSI Utilities (either Version

2.x or 3.x and

newer

), and the Universal Operator Interface (UOI). For full details

on installing and configuring these programs, please see the

Manual

D5074), and the

(document# D5076 or D5081), the

UOI Operat or Manual

(controllers on Level 1). In order to perform this

Open BSI Utilities

UOI Configuration Manual

(document#

(document# D5075).

Open BSI (Version 2.3 or earlier 2.x) Data Link Parameters

Open BSI is configured via the Open BS I Setup Tool. System Startup parameters
are entered in the System Startup dialog box, then the user configures a
communication line using the Commu nication Line Para meters dialog box.

The figure, on the next page, reproduces the System Startup dialog box for the
Open BSI Setup Tool. Of the various parameters in this dialog box, the ones which
have a direct affect on the data lin k are the

Retries”. “Max RTU Address”

is similar to the ‘High Slave Addr’ parameter for a

“Max RTU Address”

and the

“RTU

Master Port, except it encompasses all top-lev el (l evel 1) nod es. (Th e add ress range
for each commun ication line is defined in a different dialog box. )

You may remember that when a Master Port polls a slave node, and receives no
response within a specified timeout period, it will make two additional attempts for
a total of three attempts.

15

For the

“RTU Retries”

parameter in Open BSI, you

have the option of changing the number of attempts made.

15

It is possible to change the number of retries made, if your firmware supports advanced master

polling parameters. See ’Configuring Advanced Maste r Pollin g Parameters’ on page 4-10.

Page 4-18 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-19

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Once the Sy stem S tar tup p ara meters h a ve bee n d ef i ne d, the Acti ve Commu n ica tio n
Lines dialog box must be activated and a new communication line must be
configured via the Communication Line Parameters dialog box. All parameters
within this dialog box are related to configuration of the data link. Many serve
similar functions to the parameters used when defining a Master Port. (See the
figure below.)

Page 4-20 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Open BSI (Version 3.0 or newer) Data Link Parameters

In Open BSI Version 3.0 (or
Network Master of the BSAP network. Open BSI Version 3.0 (
parameters are confi gured in the Comm Line Wizard of NetView:

newer

), the Network Host PC (NHP) serves as the

and newer

) data link

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-21

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Adjustme nts to the link- level ti meout pe riod, an d dia ling p arameters ma y be made
via the

[Advanced Parameters]

push button.

Page 4-22 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

UOI/TMS Data Link Parameters

If you are using UOI or TMS
software, or a DOS-based version
of the ACCOL Tools, a common
set of Communication Setup
Menus is used to configure the
data link. Optionally, much of
the information normally
entered in these menus may be
specified, off-line, in the
ATOOLS.INI file, which is
automatically read at system startup.

Users must choose either local communication with the attached node, or global
communication to another node which exists in the NETTOP files. If local
communication is used, the node address and/or node name must be entered, as
well as the Expanded Node Addressing Group Number (if applicable). If global
communication is used, only the node name (as defined in NETTOP) needs to be
entered.

Once the decision concerning local or global communication has been made, the
user must e nter the actu al data li nk paramete rs includi ng the ba ud rate, timeout,
etc.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-23

Chapter 4 - Guidelines For Configuring the BSAP Data Link

POLL_RATE Parameter i n ATOOLS.INI
The default rate at wh ich UOI/TMS or the DOS-based ACCOL Tools requ est data

from their attached top-level nodes is 0.3 seconds. This value may be modified by
changing the POLL_RATE parameter value in the ATOOLS.INI file. Valid rates
which may be entered are: 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, or 1.0. For full details on the
ATOOLS.INI file, see the

Page 4-24 Networ k 3000 Co mmunicati o ns Config uration Guide

ACCOL II Reference Manual

(document# D4044).

Chapter 4 - Guidelines For Configuring the BSAP Data Link

BSAP Local Lines (Used for special applications only - Open BSI Version 3.1 and
newer)

Beginning with Open BSI Version 3.1 (and newer), Open BSI Workstations can
include a

There are two common uses for the BSAP local line:

Users running Open BSI on a laptop/notebook PC want to plug into the pseudo

•

slave port of a Network 3000 controller which is at a lower level of the BSAP
network (network level 2, 3, 4, 5, or 6) to view data from various network nodes.
This is useful during testing and debugging or any time a technician is visiting a
portion of the network that is geographically distant from the Network Host PC.

A communication line has failed, so operators need to establish a temporary,

•

backup communication line. For example, if the regular hard-wired
communication cable becomes disconnected from an RTU, but the RTU also has
a Pseudo Slave Port connected to a modem, and auto-answer is configured, a
BSAP local line could be configured at the NHP to dial into that RTU and obtain
data, until the normal hard-wired connection is restored.

BSAP local line

.

IMPORTANT: Communication traffic from a particular workstation to a particular
RTU can only occur via one communication line at any one time. Therefore, if a
BSAP Local Line is active (as a temporary backup line) and then the regular
communication line is re-activated, RTUs reachable via the BSAP local line will be
inaccessible via the regular line as long as the BSAP local line is active.

For more information on BSAP local lines, see

Architectures’

as well as
same manual.

’ Comm Line Wizard : Step 2 of 2 (Loca l BSAP Line)’

in Chapter 1 of the

Open BSI Utilities Manual

’Overview of Supported Network

(document# D5081),

in Chapter 6 of the

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-25

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Configuration Issues For Modems

Modems may be used to al low d ata to b e tran sferred over tel eph one l ine s. Data can
be transferred either:

"

between two Network 3000-series controllers

"

between a Network Master (PC) and a Network 3000-series controller

A pair of modems is required for each setup; i.e. there must be one modem on each
end of the telephone line. Depending upon the Network 3000 series controller(s)
you are using, they may have an internal modem already installed. In other cases,
external modems must be purchased.

-or-

The telephone line can be either a

. A private line means there is a single dedicated phone connection between the

line

private line

or

public switched network

two modems, eliminating any need for dialing. A public switched network line
means that the mod em must dial a nu mber using the public te lephone system.

Which Modems Are Supported?

Bristol Babcock currently offers three differen t modems for use with our Network
3000 product lines. These are shown in the table, below:

Modem: Baud Rate: Supports

Dialing?

For details on hardware set up of this
modem please see the Bristol Manual
listed below:

BBI 1200 bps

1200 YES CI-1200
Modem
BBI 1200 bps -

1200 NO CI-1200-PL
PL Modem
BBI 9600 bps -

9600 YES CI-9600A
PSTN Modem

Note: Older Bristol modems such as the SNM are no longer sold by Bristol, and so will not be
discussed.

Because of the large number of modems available on the market, it is NOT p ossible
for Bristol Babcock to test and support non-Bristol Babcock modems. Bristol
Babcock’s Application Support Group, however, has successfully tested current
Bristol equipment with U.S. Robotics modems.

The table, on the next page, shows how modem characteristics should be set when
connecting a U.S. Robotics modem to an Open BSI workstation.

Page 4-26 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

The first column de scribes what modem settings are required.
The second column shows corresponding initialization codes which would be

included in the initialization string for US Robotics modems. If you are starting
with a modem which is at its factory default settings, the E1, &C1 and &D2 codes
shown in the table are unnecessary.

If you have a different type of modem,

the manual accompanying that particular modem

A modem connected to an Open BSI Workstation must
have the following settings:

use the equivalent initialization code from

; not the code shown here:

Corresponding
US Robotics Code:

Turn ON local echo of codes. E1

NOTE: E1 is a factory
default

Disable ARQ result codes. &A0
Set modem’s serial port rate to va riable; to follow the

&B0
connection rate.
Set Carrier Detect (CD) to normal operation. &C1

NOTE: &C1 is a factory

default

Set Data Terminal Ready (DTR) to normal. &D2

NOTE: &D2 is a

factory default

Disable transmit data (TD) Flow Control. &H0
Disable data compression. &K0
Disable error control. NOTE: This setting is NOT

&M0
required if you have an identically configured modem
at the RTU end.
Ignore RTS for receive data (RD). &R1

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-27

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Configuration Tips

If you a re tryin g to con fi gure mo dems f or use w ith B ristol equi pmen t, the fol lowi ng
configuration tips are offered, however there is NO GUARANTEE that these will
allow compatibility with modems other than those listed under the

Are Supported?’

section.

If your Network Master is a PC running Bristol DOS-based software tools such as
UOI, TMS, on- line AIC, Toolkit, Taskspy, etc, which will be using a modem to dial
out to a particular Network 3000-series node:

"

The ATOOLS.INI file may be used to configure modem initialization strings.
Modems should be set to their factory defaults, typically using an ‘AT@F’
command. The user should then examine the factory default profile of the
modem. (For information on ATOOLS.INI, see the

ACCOL II Reference Manu a l,

document# D4044.)

‘Which Modems

"

If the factory default profile of the modem uses

data compressi on

this

DISABLED.

"

If the factory default profile of the modem uses

error correction

this

must be

DISABLED.

"

Flow control should be set to RTS/CTS.

"

Result codes should be enabled and set to NUMERIC.

"

All devices on the line must share the same baud rate

. That is, the baud rate
specified in the Communication Setup Menu (or ATOOLS.INI file) for the
Bristol software program (Toolkit, UOI, etc.) must be same baud rate as the
modems on both ends of the line, and the Slave/Pseudo Slave ports in the
Network 3000-series controllers. Bristol Babcock controllers do NOT
dynamically chan ge their bau d rates.

If you are configuring a Network 3000-series controller to dial

another

Network

3000-series controller, or an Open BSI Workstation:

"

The ‘Auto-dial’ feature must be configured. It allows controllers to dial out.
Note, however, th at the auto-dial feature is not truly automatic. For examp le,
if a message to a slave node must be transmitted, that, in itself, will not
trigger the port to dial the sla ve. Dia ling i s activa ted based on entri es in th e
Dial Control Signal List.

must be

"

Master Ports, Slave Ports, and Pseudo Slave Ports can be configured for
dialing. If Slave or Pseudo Slave Ports are used, however, dialing will only

Page 4-28 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

establish the connection; the Master must still poll the slaves in order for
data to be transmitted.

"

Any modem used must go to command mode when Data Terminal Ready
(DTR) is OFF for 50 milliseconds. The modem must also use a Carriage
Return code of 13 (0DH).

"

The ports and modems at both ends must be set to the same baud rate.

Normally, the standard Dial Control Signal List must be configured. If, however,
you are using Slave auto-dialing (i.e. the Slaves dial into the Master) and your
Network Master and controllers support it, you should consider using the Enhanced
Slave Dial Control Signal List, instead, which can significantly improve throughput
if you have a large number of Slaves.

"

The number of the Dial Control Signal list (or Enhanced Slave Dial Control
Signal List) is specified for a particular port using the port’s corresponding
#DIAL.

nnn

system signal. The table below shows this correspondence

between ports and #DIAL signals.

If you are configuring auto­dialing for this port….

You will need to
use this system
signal to specify
the Dial Control

List
Port A #DIAL.000.
Port B #DIAL.001.
Port C #DIAL.002.
Port D #DIAL.003.
BIP 1 #DIAL.004.
BIP 2 #DIAL.005.
Port G #DIAL.006.
Port H #DIAL.007.
Port I #DIAL.009.
Port J #DIAL.010.

A positive value for the #DIAL.

nnn.

signal specifies that the Dial Control Signal list
will be used.
A negative value for the #DIAL.

nnn.

signal specifies that the Enhanced Slave Dial

Control Signal List will be used.

"

Auto-dialing is described in the

ACCOL II Reference Manual

(document# D4044). Highlights from that section

‘Auto-Dial Modem Interface’

section of the

are presented below.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-29

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Standard Dial Control Signal List
The standard Dial Control Signal

List allows dialing to be activated,
reports the status of dialing, and
specifies the phone numbers, and
any special commands which are
sent to the modem.

For example, if #DIAL.000. is set to 1, a Dial Control Signal List must be created,
similar to the list shown below. The choice of signal names is entirely up to the
user.

*LIST 1
10 PORTA.ENABLE. DIAL
20 PORTA.STATUS.DIAL
30 PORTA.PHONE.SEL
40 PHONENUM.CALL.RTU1
50 PHONENUM.CALL.RTU2
60 PHONENUM.CALL.RTU3

The standard Dial Control Signal list has the following elements:
DIAL ENABLE signal if a logical signal, this signal enables dialing when ON,

and disables dialing when OFF. For a Slave or Pseudo
Slave Port, this should be defined as an analog signal, in
which case, it indica tes the max imum conn ection ti me, in
seconds. Once a connection has been established, a timer
starts which terminates the connection when the specified
time has elap sed. Setti ng th is si gna l to 0 o r OFF di sabl es
dialing or terminates the connection, if it is in progress.
Typically, ACCOL logic would be used to turn ON this
signal and activate dialing.

DIAL STATUS signal is an analog signal which reports error and status

information concerning auto-dial operations.

Possible values for this signal are shown in the table, below:

Error/Status Code:

Description

:

2 Dialing or Special Command completed

successfully
1 Dialing or Special Command in progress
0 Phone hung up

-1 Dialing complete but line was Busy

Page 4-30 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Error/Status Code:

-2 Dialing failed (no dial tone when required,

-3 Dialing not possible

-4 Special Command failed (“ERROR”

-5 Invalid dial control signal list

-6 Invalid Phone #/Command Select (value=0

-7 Invalid Phone # / Command (signal not a

-8 Invalid Phone # / Command length (S tring

PHONE#/COMMAND
SELECT signal specifies which command or phone number should

Description

no answer, or no response from modem) or

Special Command failed (no response from

modem or unrecognized response from

modem)

response from modem)

or > number of entries in Phon e#/Command

list)

String)

is NULL or > 32 characters)

currently be used, from those specified beginning with
entry 4 in this signal list.

:

PHONE#/COMMAND
signals 1 to

If you are specif ying a speci al command for a Hayes-compati ble modem, you must
precede it with an exclamation point ‘!’ in the first character position. The
remainder of the string is sent, and automatically terminated with a carriage
return (13 or 0DH, not shown below):

n

If you are using a phone number it can include:

0-9 any digit from 0 to 9
W wait for 4 seconds, then proceed
T change to Touch Tone dialing
P change to pulse dialing (default is Tone). Pulse remains in effect

for only a single dialing sequence.

- dashes may be added for readability; they are NOT sent to the
modem.

Example: T-9-555-1212

each of these string signals consists of a phone n umber or
(for Hayes-compatible modems) a special modem
command.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-31

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Example: !AT &D2 &R1 SO=1 &W0

Enhanced Slave Dial Control Signal List

newer firmware ONLY)

The Enhanced Slave Dial Control Signal List may be used

(PES03, PEX03, PLS03, P LX03, LS501, RMS04, or

instead of

the stand ard
Dial Control List, for Slave Port dialing only. In addition to the types of entries in
the standard list, it allows the user to specify a delay in disconnecting if the slave
runs out of data to send, and a dial-up acknowledgment signal.

When standard Slave dialing is used, a slave node dials into its Master. The Master

has no way of knowing which slave is dialing in

, therefore, it begins polling

all

of
its slaves, one by one, until it polls the slave dialing in, which can then answer.
This is equivalent to the situation of hearing a ‘knock’ on your door, but the person
knocking cannot speak until you call out

their

name. The only way you can find out
who it is, then, is to call out the names of everyone you know, until you reach the
name of the person doing the knocking, who can then answer, as in ‘Is that you
Alice?’, ‘Is that you Bob?’, ‘Is that you Carol?’, and continuing this until you say ‘Is
that you Margaret?’ and Margaret answers ‘Yes it’s me!’ Depending upon the
number of acquaintances you have, this could take a long time.

Enhanced Slave dialing is a much more efficient method. In this situation, when
the Master begins to poll its first slave node, the slave dialing in will answer (even
if it is not the first slave) and the Master will then accept

its

data immediately,
rather than polling its other slaves. To follow the previous analogy, when you hear
a knock on the door, and start asking ‘Is that you Alice?’, the person knocking will
answer ‘No! it’s not Alice, it’s Fred' (or whatever their name is) . Now you can say 'Hi
Fred, what do you have for me?' and Fred responds with data.

This enhanced method, which uses the DIAL_UP_ACK enable signal, greatly
increases throughput in a system with many slave nodes, by eliminating
extraneous polling to nodes that aren’t dialing in.
The figure, at right, shows
the organization of the
Enhanced Slave Dial Control
Signal List. It is similar to
the Dial Control Signal List
except it has some additional
signals which include the
Empty Transmit Queue
disconnect time signal, and
the DIAL_UP_ACK enable
signal.

Page 4-32 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

For example, if #DIAL.003. is set to ‘-5’, an Enhanced Slave Dial Control Signal List
must be created, similar to the list shown below. The choice of signal names is
entirely up to the user.

*LIST 5
10 PORTD.ENABL E.DIAL
20 PORTD.ST ATUS.DIAL
30 POSITION.OFSEL.SIG
40 MAXEMPTY.QUEUE.TIME
50 PORTD.DIALUP.ACK
60 PORTD.PHONE.SEL
70 PHONENUM.CALL.RTU1
80 PHONENUM.CALL.RTU2
90 PHONENUM.CALL.RTU3

The Enhanced Slave Dial Control Signal List has the following elements:
DIAL ENABLE signal (see description under Dial Control Signal List)
DIAL STATUS signal (see description under Dial Control Signal List).
POSITION of SELECT this analog signal specifies the position in the list of the

PHONE#/COMMAND SELECT signal. (The position of
this signal has been made variable to allow for special
purpos e si gn al s to be a d de d to th e l i st, su ch as th e E mpty
Transmit Queue Disconnect signal, the DIAL_UP_ACK
Enable signal, and any signals related to future
enhancements.)

In List 5, shown on the next page, POSITIO N.OFSEL.SIG
is the POSITION of SELECT signal.

*LIST 5
10 PORTD.ENABLE.DIAL
20 PORTD.STATUS.DIAL
30 POSITION.OFSEL.SIG
40 MAXEMPTY.QUEUE.TIME
50 PORTD.DIALUP .ACK
60 PORTD.PHONE.SEL
70 PHONENUM.CALL.RTU1
80 PHONENUM.CALL.RTU2
90 PHONENUM.CALL.RTU3

In this case, it should be set to 6 because there are two
special purpose signals (

PORTD.DIALUP.ACK

) which pushes the

MAXEMPTY.QUEUE.TIME

and

PHONE#/COMMAND SELECT signal
(PORTD.PHONE.SEL) to position 6.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-33

Chapter 4 - Guidelines For Configuring the BSAP Data Link

If there are no special purpose signals to be included in
the list, this signal should be set to ‘4’.

EMPTY TRANSMIT QUEUE
DISCONNECT TIME this is an analog signal which specifies how long, in

seconds, a connection should be maintained after the
slave node’s transmit queue becomes empty (i.e. the slave
has no more data messa ges to send to its mas ter). If this
signal is not included, is of the wrong type, or its entry is
less tha n or eq u al to z ero, th e d ef au l t di sco nn ect ti me w il l
be 10 seconds.

DIAL_UP_ACK Enable this is a logical signal. When set ON, it allows the Slave

node to acknowledge the first poll message issued by its
Master, whether or not the poll is for this slave. This
allows the slave node dialing in to immediately send data
to the master, ra ther than wa it for the master to p oll all
of its nodes until it reaches the correct slave node. This
feature significantly increases throughput on Slave dial­up.

PHONE#/COMMAND
SELECT signal (see description under Dial Control Signal List).

PHONE#/COMMAND
signals 1 to

If your N etwor k Master is a PC running Open BSI (2.x) which will use a modem to
dial Top-Level Nodes:

You must be running the Windows 95/NT version of Open BSI. Windows 3.1 users
CANNOT use Open BSI to dial.

Open BSI can be configured to send an initialization string to the modem. Any
other modem configuration must be performed using any modem configuration
software provided by the modem manufacturer, or if applicable, through switch
settings on the modem.

n

(see description under Dial Control Signal List).

Page 4-34 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Dial parameters are configured using the Dial
Parameters dialog box, which is accessible from
the
Communication Line Parameters dialog box,
when the

The
Open BSI workstation (PC) must operate
consistently with the modem’s carrier detect
setting (i.e. it CANNOT be left constantly on.)

The modem’s connectio n baud rate must be constant, and must be eq ual to the baud
rate selected for the workstation in the Open BSI Setup Tool’s Communication Line
Parameters dialog box, and th e baud rate selected for the controller’s slave port.

[Dial Parms]

“Dial”

carrier detect

option is selected for the line.

push button of the

line from your modem to the

The modem’s transmit and receive
setting s must be consistent with those at the RTU's modem.

The modem should be set to ignore the Ready to Send (RTS) signal from the
workstation.

Local echo of codes should be turned ON.
When configuring the dial string in the Open

BSI Setup Too l’s RTU Dial Strings dialog box,
include a ‘DT’ or ‘DP’ modem command i n the
string.

Open BSI immediately precedes the dial string with the ‘AT’ modem command.

flow control

settings, and

data compression

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-35

Chapter 4 - Guidelines For Configuring the BSAP Data Link

If your N etwork Master is a PC running Open BSI (3.0 or newer) which will use a
modem to dial Top-Level Nodes:

Open BSI can be configured to send an initialization string to the modem. Any
other modem configuration must be performed using any modem configuration
software provided by the modem manufacturer, or if applicable, through switch
settings on the modem.

Dial parameters are configured using the Dial
Parameters dialog box, which is accessible from
the

[Dial Parameters]

Line - Advanced Parameters dialog box, when the

“Dial”

The
Open BSI workstation (PC) must operate
consistently with the modem’s carrier detect
setting (i.e. it CANNOT be left constantly on.)

option is selected for the line.

carrier detect

line from your modem to the

push bu tton o f th e B S AP

The modem’s connectio n baud rate must be constant, and must be eq ual to the baud
rate selected for the workstation in the Comm Line Wizard (NetView) and the bau d
rate selected for the controller’s slave port.

The modem’s transmit and receive
setting s must be consistent with those at the RTU's modem.

The modem should be set to ignore the Ready to Send (RTS) signal from the
workstation.

Local echo of codes should be turned ON.
When configuring the dial string in the

Advanced RTU Parameters dialog box
(accessible from the third page of the RTU
Wizard in NetView), include a ‘DT’ or ‘DP’
modem command in the string.

Open BSI immediately precedes the dial string with the ‘AT’ modem command.

flow control

settings, and

data compression

Page 4-36 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Configuration Issues For Radios

Bristol Babcock Network 3000 equipment has been used successfully with several
different brands and types of radios.

The configuration of the radio interface varies depending upon the type of radio,
particu larly with respect to cabling and timing. A full discussion of radio
configuration is beyond the scope of this book, however, here are some items to be
aware of:

Some radios have fixed speed internal modems and some synthes ized. Most

•

radios today are configurable in several different ways including frequency,
baud rate, and pre and post data delays. As far as cabling, all radios use the full
RS-232 compliment except the Spread Spectrum type including Microwave Data
Systems (9810) and Freewave brands.

Spread Spectrum radios only use TxD, RxD, and Signal Ground. The RTU at

•

each of these sites must have either an internal jumper on the connector
between RTS and CTS on the RTU end or the loopback switches on the
communications engine board enabled.

The MDS licensed radios (2310, 4310, 4710, 9710, etc.) and DataRadio (formally

•

EF Johnson) use TxD, RxD, RTS, CTS, DSR, DCD, DTR, and Signal Ground and
no loopback enabled. The master radio companions of these radios also use the
full compliment. Timing parameters of these radios are different too. MDS
radios are faster as far as key speed and receive speed than DataRadio are.

If there is a repeater in the system, some configuration parameters can change.

•

A tail-end-link setup in a system is another scenario which will include a modem
sharer in addition to two radios working together to send the data further down
the road s imilar to a repeater. The modem sharer is necessary in the data line if
there is a co-located RTU at the site.

For addition al techn ical assi stance in using radi os with Bris tol equ ipment, contact
our Communication Technologies division in Orlando, FL at the number listed on
page ii.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-37

Chapter 4 - Guidelines For Configuring the BSAP Data Link

Radio Tip - If messages are truncated wh en using RTS/C T S, try this…

If you are using RTS/CTS with radios, and encounter a problem where the Open
BSI workstation can transmit, but RTUs are unable to respond, it could be related
to PC port configuration in Windows 95 which results in messages being truncated.

If this problem is occurring, go to
then

"Settings"

call up the Windows 95 control panel.
Next double-click on the ’System’ icon.
The System Properties dialog box will
appear. Click on the ’Device Manager’
tab.

Click on the plus sign ’+’ next to the
’Ports (COM & LPT)’ selection. A list of
ports will appear. Click once on the port
used for Open BSI communications
(typically COM1 or COM2), then click on
the

[Properties]

and

"Control Pan el"

push button.

"Start"

to

Page 4-38 Networ k 3000 Co mmunicati o ns Config uration Guide

Chapter 4 - Guidelines For Configuring the BSAP Data Link

The Communication Port Properties
dialog box will appear; click on the
’Port Settings’ tab, then click on the

[Advanced]

the Advanced Port Settings dialog
box.

push button to call up

In the Advanced Port
Settings dialog box, drag the
’Transmit Buffer’ slide bar to
the low end of its range, and
click on the
button.

Then choose the
button in both the
Communication Port
Properties dialog box, and in
the Device Manager page, to
exit the control panel and
save the settings.

Reboot your PC for the new settings to take effect.

[OK]

push

[OK]

push

Networ k 3000 Co mmunicati o ns Config uration Guide Page 4-39

BLANK PAGE

Chapter 5 - Guidelines For Configuring the IP Data Link

The

data link

refers to the actual mechanism by which data flows through the
network. This includes the port and communication line itself, an d for purposes of
our discussion, also includes the software configuration for the ports and the
communication line. The functioning of the data link is completely independent of
the applications which use the link (HMI/SCADA packages, IP_Client/IP_Server
Modules, etc.)

A Bristol Babcock network using the IP protocol operates like any other IP
network.1 Data messages from a node are divided up into small pieces called

packets

. The packets are tran smitted thro ugh the communica tion li ne (suc h as an
Ethernet cable) and are then re-assembled into the full data message when they
reach the destination node. Currently, the Ethernet port is the only type of IP
connection supported by Bristol Babcock networks.

Just as many BSAP network terms do NOT apply in IP networks, many of the
concepts associated with BSAP data link configuration (polling, etc.) do NOT apply
in IP data link configuration. (These terms still apply, however, within a BSAP sub­network underneath an IP node.)

The main configuration activities involved in IP data link configuration involve
setting up communication ports. These include:

"

Setting up the Network Host PC (NHP) for this network. (This was discussed as
part of Chap ter 3.)

"

Defining a Master port in any IP node which serves as a

network maste r

BSAP sub-network. (If applicable, see Chapter 4; if you are not including a
BSAP sub-network, this activity is unnecessary).

"

Declaring one or more IP ports in the ACCOL load of each IP node, using
ACCOL Workbench.

"

Specifying the characteristics of the IP port(s) in the RTU’s FLASH memory
using the

"RTU Comm Config"

feature of LocalView.

to a

1

The specific type of IP protocol Bristol Babcock uses is called the User Datagram Protocol (UDP).

Networ k 3000 Co mmunicati o ns Config uration Guide Page 5-1

Chapter 5 - Guidelines For Configuring the IP Data Link

Declaring an IP Port in ACCOL Workbench

An IP Port is declared in the *COMMUNICATION PORTS section of the ACCOL
source file.

The 386EX Protected Mode units contain two Built-In Ports (BIP)s, four to eight
serial ports (Ports A through D, and, optionally, G through J) as well as an optional
Ethernet Port. Any of these ports can be configured as the IP port.

IMPORTANT

Although ACCOL Workbench will allow you to configure serial ports
BIP 1, BIP 2, Port A through D, and G thro ugh J as Internet (IP)
ports, PLS03/PLX03/PES03/PEX03 firmware WILL ONLY SUPPORT
IP COMMUNICATION VIA THE ETHERNET PORT; the serial ports
will be available for IP communication in a

future

firmware release.

NOTE: Althoug h an IP node can contain more than one IP Port, once
it has an IP Port defined, it CANNOT include any of the following port
types: Slave, VSAT Slave, or Serial CFE.

Examples of syntax for the ACCOL source file are shown below:

BIP_1 IP
BIP_2 IP
PORT_A IP
PORT_B IP
PORT_C IP
PORT_D IP
PORT_G IP
PORT_H IP
PORT_I IP
PORT_J IP
ETHRNT IP

No other configu ration is necessary within the ACCOL source file. IP Port baud rate
and other characteristics must be defined using the
of LocalView.

"RTU Comm Config"

feature

Page 5-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 5 - Guidelines For Configuring the IP Data Link

Specifying the Characteristics of the IP Port(s)

The characteristics of the IP ports are specified
using the
LocalView

"RTU Comm Config"

2

. (NOTE: LocalView must be used for
initial IP configuration; once global
communications have been established, the
same feature may be used in NetView to
view/modify the IP port chara cteristics.)

feature of

2

LocalView always uses BSAP to communicate, even though you are configuring IP port parameters
with it. A full description of the RTU Comm Config feature is included in the Open BSI Utilities
Manual ( document# D5081).

Networ k 3000 Co mmunicati o ns Config uration Guide Page 5-3

Chapter 5 - Guidelines For Configuring the IP Data Link

Your IP port can either be the Ethernet Port, or one of the serial ports. If you are
using the Ethernet Port as your IP port, please read

Interface’
the Serial Ports’

for the system. These are described under

The IP port chara cteristics, soft switches, etc. which you set or change using

LocalView’s

RTU has been reset or downloaded with a new ACCOL load. Once initial

configuration has been completed using LocalView, and global communications

have been established, RTU parameters may then be altered in the same way,

using NetView. For more information on using LocalView and NetView, see the

Open BSI Utilities Manual

. If you will be using serial port(s) for your IP port, please read

. In either case you must also conf igure the g eneral IP para meters

’Defining the IP System Parameters’.

IMPORTANT

"RTU Comm Config"

(document# D5081).

feature do not come into effect until

’Setting Up the Ethernet

’ Setting Up

after

the

Setting Up the Ethernet Interface

Ethernet parameters are only configured if you are using the Ethern et Port as you r
IP Port; they d o NOT apply if you are only using serial IP ports.

Page 5-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 5 - Guidelines For Configuring the IP Data Link

Setting Up the Serial Ports

Be sure to select an Internet Protocol (such as ’PPP_HDLC’ or ’User Mode’ if you
have created your own Internet Protocol implementation) for the port; choosing
either ’BSAP’ or ’EBSAP’ will disable the Internet parameters.

(NOT suppo rted in PES03/P EX03 PLS03/PLX 03 f irmware)

Networ k 3000 Co mmunicati o ns Config uration Guide Page 5-5

Chapter 5 - Guidelines For Configuring the IP Data Link

Defining the IP System Parameters

Once your IP ports have been defined (either Ethernet or Serial) you must also
define certain IP System Parameters.

Page 5-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 6 - Application Notes

As discussed earlier in this manual, an
makes use of the data link to transmit messages. Among the commonly used
applications are elements of the ACCOL load itself such as Master Modu les used for
peer-to-peer communication. Open BSI Utilities or DOS-based tools such as UOI
are applications because they make use of the data link to collect data from
controllers in the network. HMI/SCADA packages such as OpenEnterprise,
Intellution

This section will discuss various applications, and provide configuration hints and

tips.

nonetheless important when trying to improve system performance.

®

FIX®, and Iconics Genesis are also applications.

Not all information p resented is strictly related to communication s

Peer-to-Peer Communication

Peer-to-peer communication

one Network 3000 controller to another. This transfer is performed using Master
Modules (or EMaster Modules) and Slave Modules.

is the transfer of signal list or data array data from

application

NOTE:

(BSAP Master/Slave Modules)

is a program or function that

, but it is

Peer-to-peer communication is typically used in situations where one controller
must make some logical decision based on data from another controller.

Another reason to use peer-to-peer communication is if greater user control is
required over when communications occur. The user can create ACCOL logic to
activate the Master Module only at certain times or when certain data is ready.

IMPORTANT

Do NOT confuse Master/EMaster Modules and Slave Modules with Master
nodes/Ports and Slave nodes/Ports. They are NOT th e same things.

The following are certain things the user should be aware of when setting up peer­to-peer communication:

Users must designate which signal list or data array must be transferred. Si gnal

•

lists on both ends must have the same number of entries, and data arrays on
both ends must have identical row/column dimensions. Peer-to-peer
communication transmits data v alues only; it does not, for example, transmit the
signal names, inhibit/enable status information, etc.

A given controller can pa rticip ate in peer- to-peer commu n icati on w ith i ts Master

•

node, any of its slave nodes, an d any sibling nodes (n odes that share the same
Master).

Networ k 3000 Co mmunicati o ns Config uration Guide Page 6-1

Chapter 6 - Application Notes

It’s recommended th at Ma ster Mod u le (s) be sep ara ted in to a de di cate d ta sk; thi s

•

allows easy tuning of the task rate, with out affecting other modules.
The EMaster Module is necessary for peer-to-peer communication with

•

Expanded Addressing S lave nodes (slaves of ‘virtual nodes’) but can also be used
in standard BSAP networks to communicate with the slave o f a slave node.

For maximum throughput, it is recommended that Master Modules (but NOT

•

EMaster Modules) be placed in
Slave Modules. That is, the Master Modules should be placed in the Slave
and the Slave Modules shou ld be placed in the Master
for this is to distribute the Master Modules among the slave nodes; thereby
eliminating the situation of having several Master Modules competing for a
single node’s resources (CPU time, buffers, etc.) If, instea d, the Master Modu les
are put i n the Master n ode, the recommen ded task ra te for the ta sk contai ning
them is

half

the poll period for the correspon ding Master Port.

lower-level

nodes with resp ect to th eir asso ciated

nodes

. The main reason

nodes

It is recommended that

•

declared in the ACCOL load
Master/EMaster Modules function differently from other modules in that their

•

at least

one additional communication I/O buffer be

for each Master/EMaster Module

in the load.

communication transactions may still be underway when execution of the
ACCOL Task proceeds to the next module in sequence. Because of this, if an
ACCOL module (say a Calculator) follows a Master Module because its
calculation s depends on data bei ng collected b y the Master Module, appropriate
ACCOL logic must be added to ensure that the data collection has been
completed

before

execution of the Calculator;

using OLD DATA.

This ACCOL logic would consist of a WAIT FOR (DONE)

otherwise the Calculator will be

statement after the Master Module, with “DONE” being defined as a logical
signal, on the STATUS_1 terminal of the Master Module. This sign al is turned
ON when commun ications are comp lete, and is O FF when communi cations are
in progress or off. DO NOT INCLUDE A TIMEOUT VALUE in the WAIT
statement; the Master Module has its own built-in timeout of two minutes,
which may be modified in ACCOL.1 This same WAIT FO R logic shou ld be used
in situations where multiple Master Modules are in the same ACCOL load;
execution of a Master Module should not be started until communications have
been completed by the previous Master Module.

It is recommended that the STATUS_2 signal on the Master Module be made an

•

analog alarm signal, with a high alarm limit of 2, and a low alarm limit of -1;
whenever this signal is in an ‘alarm state’, the user will know that peer-to-peer

1

To make this timeout modification, see instructions under the MODE terminal description in the
‘Master/EMaster’ section of ACCOL II Reference Manual (document# D4044).

Page 6-2 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 6 - Application Notes

*A-ARRAY 3 RW (4,4)

data is suspected of containing errors. See the

ACCOL II Reference Manual

(document# D4044) for details on interpreting

‘Master/EMaster’

particular error v alues.

The figure, below, shows a typical peer-to-peer communication setup:

Network

Level n

NOTE that MASTER Modules
in this AC COL load are in a
task all by themselv es. Th is
arra nge ment a llo ws ’tunin g’
via the Task Ra te.

PO IN T te rm in a l spe c if ies
which SLAVE Module in
the A CCOL load to a ccess

*TASK 0
10 * SLAVE
POINT 1
ENABLE ENABLE.SLAVE.MOD
INTYPE 1
OUTTYPE ;ANALOG_SIGNAL_OR_VALUE
INLIST 3
OUTLIST ;ANALOG_SIGNAL_OR_VALUE
STATUS_1 SLAVE.COMM.DONE
STATUS_2 SLAVE.STATUS.CODE

NOTE that SLAVE Module
ha s been placed in the
Master n ode to improv e
throughput. Note a lso that
Slave Modules sh o uld always
be placed in Task 0.

’0’ o n RE MOTE te rminal indicates
tha t peer-to- pee r comm u n ication
will b e with a n ode’s M aster no de.

po s itiv e number indic a t e s
pe er-to-peer com munic a t ion
will be with one of t h is nod e’s
slaves which has a local
address of
negative num ber indicate s
peer-to-pe er com munication will
be with a sibling node w hich has
a lo cal add ress of . ( A siblin g
no de is a no d e wh ich is a sla ve to
the s a me Maste r nod e as this no de).

’n’

’n’

’-n’

’ n’

section of the

Network
Level n+1

Networ k 3000 Co mmunicati o ns Config uration Guide Page 6-3

*A-ARRAY 5 RW (4,4)

*TASK 5 RATE: 1.000000 PRI: 1 REDUN: 0
10 * MASTER
REMOTE 0
POINT 1
MODE 0
INTYPE ;ANALOG_SIGNAL_OR_VALUE
OUTTYPE 1
INDEX ;ANALOG_SIGNAL_OR_VALUE
INLIST ;ANALOG_SIGNAL_OR_VALUE
OUTLIST 5
STATUS_1 MASTER1.COMM.DONE
STATUS_2 MASTER1.STAT.CODE
20 * WAIT FOR (MASTER1.COMM.DONE) 0.1
30 * MASTER
REMOTE -1
POINT 2
MODE 1
. .
. .

WAIT statem en t is used
to prevent an other MAST E R
Mo dule fro m being s tarted
before the first one finishes
its transaction.

Chapter 6 - Application Notes

Peer-to-Peer Communication

Peer-to-peer communications using IP_Client / IP_Server modules allows data from
ACCOL structures such as signal li sts and data arrays to be transferred from one
IP-capable remote process controller to another IP-capable remote process
controller. Peer-to-peer communication is typically used in situations where one
controller must make some logical decision based on data from another controller.

Another reason to use peer-to-peer communication is if greater user control is
required over when communications occur. The user can create ACCOL logic to
activate the IP_Client Module only at certain times or when certain data is ready.

NOTE

Currently, the only IP-capable Network 3000 RTU’s are 386EX
Protected Mode versions of the DPC 3330 or DPC 3335, with
PES03/PEX03 or newer firmware. To perform peer-to-peer
communication in non-IP RTU’s you must use BSAP
Master/EMaster/Slave modules.

The IP_Client Module initiates all such communication transactions by contacting
an IP_Server Mod ule in another IP-capable controller (the target RTU). The
IP_Client Modul e requests either read- only, or read-w rite access to various arrays,
signal lists, etc. in the target RTU. The IP_Server can grant those requests for
access, or deny them, based on how its module terminals are configured.

(IP_Client / IP_Server Modules)

The IP_Client Module can requ est access to entire structures ( an entire data array,
or an entire signal list) or it can request access to only a set of contiguous signals in
the signal list, or contiguous elements in a data array based on the entries in its
SERVR_INDEX and CLNT_COU NT terminals.

The IP_Server Modu le i n th e targ et RTU can restrict a ccess to the stru ctures u nd er
its control using its LIST_DB, AARRAY_DB and LARRAY_DB terminals. It can
also restrict access to IP_Clients whose node names or addresses appear in the
KNOWN_IP_NODES list; any unlisted nodes will be denied access by the
IP_Server.

An example of IP_Client/Server communication is as follows:
There are two remote process controllers, with IP node addresses of 120.0.0.1, and

120.0.0.3, respectively (see figure on page 6-6). The user wants to read the first
three entries of Signal List #5 in the 120.0.0.3 unit (the server), into positions 8, 9,
and 10 of Signal List #3 of the 120.0.0.1 unit (the client). This example outlines the
configuration required to perform this data tra nsfer.

Page 6-4 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Chapter 6 - Application Notes

Configuring the IP_Client Module:

When configuring the IP_Client Module, the target RTU for which access is desired
must be identified on the REMOTE terminal signal. In this case, the target RTU
will be identified by its address of 120.0.0.3.

Because read access to that RTU is required, th e ACCESS_MODE signal must be
set to 1 (DBREAD).

Because the type of structure being read is a signal list, the STRUCT_TYPE signal
must be se t to 1 (Signal List).

Since data must be read from Signal List #5 in the target RTU, the
SERVR_STRUCT_NO terminal signa l must be set to 5. Since only the first 3 entries
in Signal List #5 are to be read, then the SERVR_INDEX terminal si gnal must be
set to 1 (i.e. start with the first position in the list), an d the CLNT_COUNT
terminal signal must be set to 3 (i.e . read on ly the first 3 entri es of the list).

In addition, the signal list data being copied in from the target RTU is to be stored
in positions 8, 9, and 10 of Signal List #3. Therefore, the CLNT_STRCT_NO must
be set to 3 (for Signal List #3), and CLNT_INDEX must be set to 8 (because position
8 is the first position where data must be stored).

NOTE: Always check for successful completion via the STATUS terminal(s) before
attempting to use peer-to -peer data.

Configuring the IP_Server Module

IP_Server configuration is some what simpler. If the LIS T_DB and
KNOWN_IP_NODES terminals were left unwired, there would be almost no
configuration. To illustrate how the IP_Server Module allows restrictions to be
placed on access, these terminals will be wired.

The LIST_DB terminal will specify an analog data array which will identify which
signal lists are to be made accessible to IP_Clients. In this case, we must include
Signal List #5 in that array, and specify that it will be available for read access.

The KNOWN_IP_NODES terminal will specify a signal list which will contain
signals that identify the RTU’s which will be granted access via IP_Client requests.
In this case, we must identify the RTU with address 120.0.0.1.

NOTE: Always check for successful completion via the STATUS terminal(s) before
attempting to use peer-to -peer data.

Networ k 3000 Co mmunicati o ns Config uration Guide Page 6-5

Chapter 6 - Application Notes

Page 6-6 Networ k 3000 Co mmunicati o ns Configuratio n Guide

Setting Task Rates

Chapter 6 - Application Notes

The

task rate

specifies how often (in seconds) the controller will attempt to start
execution of an ACCOL task .2 This determines how fast the modules in a particular
ACCOL task are executed, and how fast data is processed by that task. If the task
includes process I/O modules (ANIN, ANOUT, DIGIN, DIGOUT etc.) it also affects
how fast data is collected from (and sent to) process instrumentation.

To maximize throughpu t, tasks should be set to execute as quick ly as possible,

long as t hey do NOT resul t in task

slippage

. Slippage is a condition in which a task

so

cannot run at its scheduled time because it is still attempting to complete the
previous execution. Slippage is reported via the #RCNT.

nnn

is the tas k number).

nnn

system signals (wh ere

If peer-to-peer communication is being configured in a Master Node (and Master
Modules are placed in the Master node) the task rate for the task containing the
Master Module(s) should be set to one half the poll p eriod of the Master port.

Task rates can be tuned dynamically (via DataView, Toolkit, etc.) using the
#RATE.

nnn

. system signal (where

nnn

is the tas k number.)
Setting the task rate to 0 stops execution of the task.
The amount of time it takes for a task to execute can be measured using the

#RTTIME.000. and #RTTIME.001. system signals. See the
of the

ACCOL II Reference Manual

(document# D4044) for details.

‘System Signal s’

section

Setting Task Priorities

In ACCOL, multiple tasks execu te concurrently. If two tasks are scheduled to run
at the same time, the one with the higher priority will be executed ahead of the
lower priority task. Priority is set using the #PRI.

nnn.

system signals (whe re

the task number) and can range from 1 to 64.
The aspect of task priority most relevant to network communications is to make

sure that your ACCOL tasks are NOT given a priority higher than system-level
tasks which handle communications.3 To ensure this, do NOT assign an ACCOL
task a priority gre ater than 32.

2

For more information on task rates, see ‘An Introduction to ACCOL’ (docu m ent# D4056) fo r detail s.

3

For a list of system tasks priorities, see the ‘Task’ section of the ACCOL II Reference Manual
(docum ent# D4044). F o r more in f o rmation on task pri o rity see ‘An Introduction to ACCOL’
(docu ment# D4056).

nnn

is

Networ k 3000 Co mmunicati o ns Config uration Guide Page 6-7

Chapter 6 - Application Notes

ACCOL Load Design Issues

There are numerous factors in volved in the design of your ACCOL load, wh ich can
potentially affect communications throughput.

For example, if you are doi ng time-criti cal calcula tions, for exampl e, massive array
manipu lati ons) at the s ame time w hen commun icatio ns are to occur, a conf lict over
system resources can result, causing the calculations to be delayed. If possible,
schedu le the calculations to occur at some time othe r than when d ata collection i s
scheduled to occur.

As another example, if there are many signals in your system which change
infrequently or are of lesser importance, you may want to consider collecting them
via Report By Exception (RBE), wh ich is su pported by OpenEnterprise, Enterprise
Server or Intellution® FIX®. RBE reduces the amount of da ta flowing on the data
link at any one time.

A full list of these sorts of design issues is beyond the scope of this document.

Remote Process I/O

Remote process I/O

3330/DPC 3335 or RTU 3310 model controller in order to allow I/O to be physically
separate from the controller itself. Users should make sure the I/O point count from
all of the boards in the RIO 3331s
fully populated host. Numerous other factors can affect RIO system operation as
well. Users are urged to test their desired system configuration to verify that it
provides adequate performance.

One technique which may
process I/ O is to group the remote input modules (RANIN, RDIGIN, etc.) together in
the task. Because all remote I/O d ata from all racks is requested simultan eously,
keeping the remote I/O modules close together in the task may allow a single
request for all data to satisfy all the modules, with out causing an additional request
due to stale data.

involves connecting RIO 3331 Remote I/O Racks to a DPC

and

in the host unit, does NOT exceed that of a

slightly

improve throughput in a unit using remote

Communication I/O Buffers

Buffers

The system automatically allocates a certain number of buffers in your Network
3000 controller; additional buffers should be added by the user, depending upon the
requirements of your application.

are pre-allocated blocks of memory which are used to hold data messages.

Page 6-8 Networ k 3000 Co mmunicati o ns Configuratio n Guide