Professional Documents
Culture Documents
T-034 - Networking PDF
T-034 - Networking PDF
T-034 - Networking PDF
English
Original Instructions
4-2013
A044E523 (Issue 1)
Table of Contents
1. INTRODUCTION............................................................................................................................
1.1 Overview .................................................................................................................................
1.2 About This Manual..................................................................................................................
1.3 Application Manuals................................................................................................................
1.4 Safety......................................................................................................................................
1
1
2
3
3
5
5
5
8
8
9
9
9
10
11
11
11
11
11
12
12
13
14
14
15
15
17
17
19
19
19
20
20
21
23
23
23
24
24
24
A044E523 (Issue 1)
Table of Contents
4-2013
3.4.3 Bridge.........................................................................................................................
3.4.4 Routers.......................................................................................................................
3.4.5 Gateways ...................................................................................................................
3.5 Web Servers .........................................................................................................................
24
25
25
26
29
29
29
30
30
31
31
31
32
32
5. MEDIA CONSIDERATIONS.........................................................................................................
5.1 Overview ...............................................................................................................................
5.2 Cable Concepts ....................................................................................................................
5.2.1 Attenuation .................................................................................................................
5.2.2 Characteristic Impedance and Network Termination .................................................
5.2.3 Electromagnetic Interference .....................................................................................
5.2.4 Crosstalk ....................................................................................................................
5.2.5 Shielding ....................................................................................................................
5.2.6 Insulating Jacket - Plenum vs. PVC ..........................................................................
5.3 Cable Types .........................................................................................................................
5.3.1 Twisted Pair Cable.....................................................................................................
5.3.2 Fiber-Optic Cable .......................................................................................................
5.3.3 Coaxial Cable.............................................................................................................
5.4 Twisted Pair Cable Standards ..............................................................................................
5.4.1 NEMA Level IV Cable ................................................................................................
5.5 Physical Layer Protocols .....................................................................................................
5.5.1 Ethernet......................................................................................................................
5.5.2 Modbus over Serial Line ............................................................................................
5.5.3 PCCNet ......................................................................................................................
5.5.4 LonWorks ...................................................................................................................
5.5.5 Wireless Media...........................................................................................................
35
35
35
35
35
36
36
37
38
38
39
40
40
41
41
41
41
42
43
43
44
6. PROTOCOLS...............................................................................................................................
6.1 Overview ...............................................................................................................................
6.2 Protocols ...............................................................................................................................
6.2.1 Proprietary vs. Open Protocols ..................................................................................
6.3 The OSI Reference Model ....................................................................................................
6.3.1 Layer 1 - Physical Layer ............................................................................................
6.3.2 Layer 2 - Data Link Layer ..........................................................................................
6.3.3 Layer 3 - The Network Layer .....................................................................................
6.3.4 Layer 4 - Transport Layer ..........................................................................................
47
47
47
47
48
48
50
52
55
ii
A044E523 (Issue 1)
4-2013
Table of Contents
57
57
58
59
59
59
60
60
60
60
61
61
61
61
62
62
62
63
63
63
64
64
7. MONITORING SYSTEMS............................................................................................................
7.1 Overview ...............................................................................................................................
7.2 Purpose of Monitoring...........................................................................................................
7.2.1 System Reliability.......................................................................................................
7.2.2 Operating Cost ...........................................................................................................
7.2.3 Agency Requirements................................................................................................
7.3 Monitoring System Functions................................................................................................
7.3.1 Status Display ............................................................................................................
7.3.2 Alarm Notification .......................................................................................................
7.3.3 Data Logging..............................................................................................................
7.3.4 Reporting....................................................................................................................
7.4 Monitoring System Architecture ............................................................................................
7.4.1 SCADA Systems ........................................................................................................
7.5 Web-based Monitoring Systems ...........................................................................................
7.5.1 Web Server Location .................................................................................................
7.6 Which Monitoring System to Choose....................................................................................
65
65
65
65
65
66
66
66
69
69
72
72
73
75
77
78
79
79
79
80
80
81
6.4
6.5
6.6
6.7
6.8
A044E523 (Issue 1)
iii
Table of Contents
4-2013
9. APPENDIX ...................................................................................................................................
9.1 Glossary................................................................................................................................
9.2 Acronyms ..............................................................................................................................
9.3 Codes and Standards ...........................................................................................................
9.3.1 Related Product Standards........................................................................................
9.3.2 Modification of Products.............................................................................................
iv
85
85
93
96
96
97
A044E523 (Issue 1)
WARRANTY
This manual is published solely for information purposes and should not be considered all inclusive. If
further information is required, consult Cummins Power Generation. Sale of product shown or described in
this literature is subject to terms and conditions outlined in appropriate Cummins Power Generation selling
policies or other contractual agreement between the parties. This literature is not intended to and does not
enlarge or add to any such contract. The sole source governing the rights and remedies of any purchaser of
this equipment is the contract between the purchaser and Cummins Power Generation.
NO WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING WARRANTIES OF FITNESS FOR A
PARTICULAR PURPOSE OR MERCHANTABILITY, OR WARRANTIES ARISING FROM COURSE OF
DEALING OR USAGE OF TRADE, ARE MADE REGARDING THE INFORMATION,
RECOMMENDATIONS AND DESCRIPTIONS CONTAINED HEREIN. Each customer is responsible for the
design and functioning of its building systems. We cannot ensure that the specifications of Cummins Power
Generation products are the proper and sufficient ones for your purposes. You must satisfy yourself on that
point.
In no event will Cummins Power Generation be responsible to the purchaser or user in contract, in tort
(including negligence), strict liability or otherwise for any special, indirect, incidental or consequential
damage or loss whatsoever, including but not limited to damage or loss of use of equipment, plant or power
system, cost of capital, loss of power, additional expenses in the use of existing power facilities, or claims
against the purchaser or user by its customers resulting from the use of the information, recommendations
and descriptions contained herein.
A044E523 (Issue 1)
WARRANTY
4-2013
vi
A044E523 (Issue 1)
Introduction
1.1
Overview
The purpose of this Application Manual is to educate engineers, system integrators, distributors,
and interested users in the fundamentals of networks, as they apply and are used in on-site
power generation systems.
Communication networks have long been used to make equipment and processes operate more
reliably and efficiently in the following ways:
Networks improve system reliability by communicating fault or potential fault conditions to
equipment or personnel that can take appropriate action to prevent failures or equipment
damage.
Networks improve serviceability of a system as technicians are able to assess a systems
service requirements remotely and in some cases allow for download of software upgrades
over the internet.
Networks minimize operating costs by making service more efficient and simplifying status
indication and reporting requirements.
Networks minimize installation costs by reducing point to point wiring and the associated
wiring errors.
Networks enable more efficient operation of a system as operating and performance
information enables a system controller to cycle equipment on or off and make other
adjustments of operating parameters.
Networks allow building management systems to monitor and control equipment. This
includes heating and cooling, lighting, transportation and security systems as well as power
generation and distribution equipment.
Data provided over a network is often used for trend analysis which can optimize system
performance over time or be used to diagnose problems.
Networks will play an expanded role in power generation, distribution and transmission systems
moving forward. Demand for power is growing in excess of existing generation capacity at the
same time tolerance for interruptions in our power supply is decreasing. Developing additional
centralized generation capacity is capital intensive and fraught with political difficulties as fuel
sources are perceived as non-renewable, harmful to the environment and pose health and
safety risks. This results in a demand for distributed rather than centralized power generation
sources and for improvements to power transmission systems in terms of supplying energy
needs more efficiently and more effective fault isolation and fault tolerance systems.
As power generation systems migrate from centralized to distributed generation and control, the
communications infrastructure will need to become more comprehensive and standardized so
that equipment from multiple suppliers will be able to communicate with each other seamlessly.
Seamless communication is the enabling technology that allows the potential advantages of
distributed power generation to be realized. Examples of this include
Switching loads off of the grid when there is a danger of a utility transformer becoming
overloaded or to do maintenance on utility distribution equipment
Exporting power to the grid from distributed sources to serve loads that have been
interrupted due to a fault in the distribution system.
A044E523 (Issue 1)
1. Introduction
4-2013
Supplementing alternative energy sources with diesel gensets to mitigate the intermittent
nature of the alternative source
Implementing load demand schemes across multiple distributed sources in order to use the
available capacity most efficiently.
Even equipment that is not connected to a grid will be expected to communicate with other
equipment for applications ranging from simple remote alarm notification to full monitoring
control and analysis. Networking capability is becoming an enabling technology for power
generation equipment across the entire range of power generation applications.
1.2
A044E523 (Issue 1)
4-2013
1. Introduction
1.3
Application Manuals
Every standby generator set installation will require power transfer equipment, either transfer
switches or paralleling switchgear. The proper system for the job and its proper application are
crucial to reliable and safe operation. The following Cummins Power Generation application
manuals address related aspects of standby and emergency power systems. Because these
manuals cover aspects requiring decisions that must be taken into consideration early in the
design process, they should be reviewed along with this manual.
Application Manual T-011 - Automatic Transfer Switches. Many applications utilize multiple
power sources to enhance electric power system reliability. These often include both utility
(mains) service and generator set service to critical loads. T-011 covers the various types of
power transfer systems available and considerations for their use and application. Careful
consideration of the power switching system at the start of a project will enable a designer to
offer the most economically viable and most reliable service to the facility user.
Application Manual T-016 - Paralleling and Paralleling Switch Gear. Paralleling equipment
makes two or more generator sets perform as one large set. This can be economically
advantageous, especially when the total load is greater than 1000 kW. The decision whether to
parallel sets must be made in the early stages of design, especially if space and the need for
future expansion are critical factors.
Application Manual T030 Liquid Cooled Generator Sets. Generator sets may operate as prime
power sources or provide emergency power in the event of utility power failure. They may also
be used to reduce the cost of electricity where the local utility rate structure and policy make
that a viable option. Because of their important role, generator sets must be specified and
applied in such a way as to provide reliable electrical power of the quality and capacity required.
T-030 provides guidance to system and facility designers in the selection of appropriate
equipment for a specific facility, and the design of the facility, so that these common system
needs are fulfilled.
1.4
Safety
Safety should be a primary concern of the facility design engineer. Safety involves two aspects:
safe operation of the generator set itself (and its accessories) and reliable operation of the
system. Reliable operation of the system is related to safety because equipment affecting life
and health is often dependent on the generator set such as hospital life-support systems,
emergency egress lighting, building ventilators, elevators, fire pumps, security and
communications.
Refer to the Codes and Standards section for information on applicable electrical and fire codes
around the world. Standards, and the codes that reference them, are periodically updated,
requiring continual review. Compliance with all applicable codes is the responsibility of the
facility design engineer. For example, some areas may require a certificate-of-need, zoning
permit, building permit or other site-specific certificate. Be sure to check with all local
governmental authorities early in the planning process.
A044E523 (Issue 1)
1. Introduction
NOTE:
4-2013
A044E523 (Issue 1)
Networking Basics
2.1
Overview
A network can be defined as a collection of devices or nodes that communicate with one
another over a common medium (i.e. wire, fiber, wireless) where information is exchanged via
sets of rules known as protocols. In this chapter we will define some of the key terms and
concepts associated with networks and discuss how networks have evolved historically.
2.2
FIGURE 1.
As you can imagine, this type of system monitoring is pretty limited because it gets expensive
and complex as the number of conditions to be indicated grows and as the distance between
the generator and the remote monitoring point increases, or if the number of monitoring points
increases. While a system of this type could provide simple control commands, such as a
remote start capability, it can not handle analog information (for example indication of a range of
temperatures) but rather just provides discrete yes/no type of information.
A044E523 (Issue 1)
2. Networking Basics
4-2013
Even a simple generator set application could include some pretty complicated wiring, as
indicated in Figure 2, because every wire would have a specific purpose and would need to be
connected in exactly the correct location.
FIGURE 2.
Note that even for this installation, there are inputs (a mechanism to tell the system the status of
a specific piece of information), outputs (a mechanism to provide information to a remote person
or device), media (wiring between the devices), and a power supply to operate the system (in
this case DC battery power).
Network-based communication was developed to deal with some of the weaknesses of
traditional communication links between devices. There are a number of similarities between
network-based communications and traditionally designed systems.
Figure 3 shows an annunciator panel, with the same outward functionality as shown in Figure
1, which utilizes a design with network-based communication. In this example there are still
switch inputs (only one is shown for simplicity), and these switch inputs, instead of directly
carrying current to a lamp, provide an input to an input/output (I/O) chip. The I/O chip connects
to a microprocessor (uP) that reads data off the chip and then transfers it through media (in this
case a twisted pair of wires) to a matching microprocessor in the annunciator. This second
microprocessor operates an output switch in the annunciator I/O module, and this output switch
can operate the lamp when required.
A044E523 (Issue 1)
4-2013
2. Networking Basics
FIGURE 3.
Note that a power supply is still required to operate the remote annunciator, but it does not
necessarily need to come from the generator set control. There are still inputs, outputs, media,
and with this design, the microprocessors must communicate with each other in the same
language, or protocol. They also must accept the same type of media to transmit messages.
The major advantages seen in the network-based design versus the more traditional designs is
a huge simplification in the installation of the generator set and, because of that, a greatly
reduced installation cost in both dollars and time. A major advantage that may not be clearly
visible is that since all the operation of the equipment is managed by software in the
microprocessors, changes can be made with software manipulation rather than with hard wire
changes, again providing lower cost and also a greater degree of flexibility in design. Changes
and additions are also easy to accomplish with this design.
As can be seen in Figure 4 vs. Figure 2, there are many less connections to make (just control
power to each component and the media that interconnects themnote in this illustration the
twisted pair cable is represented by a dashed line).
A044E523 (Issue 1)
2. Networking Basics
4-2013
FIGURE 4.
A network, then, can be loosely defined as a series of devices (in our illustration a transfer
control, generator set control, and annunciator) that communicate with each other over a
common media (twisted pair cable, wireless, fiber optic cable) where information is exchanged
via an established language (set of rules) which is called protocol. The devices in a network
have inputs and outputs (I/O) that provide the necessary functionality for each system. Points
with network connections are sometimes called nodes.
2.3
Common Terminology
In this section we will define some of the more commonly used networking terms that are used
in this manual. See the Appendix, Chapter 9, for a complete glossary of terms.
2.3.1
Protocols
A protocol can be defined as a set of rules used mutually by two or more devices or software
applications to communicate. Protocols specify all of the characteristics of a communication
network from the physical and electrical qualities to the languages used to communicate.
A044E523 (Issue 1)
4-2013
2.3.2
2. Networking Basics
Topology
Topology refers to the physical shape and interconnection scheme of the network. The three
most common topologies used in control networks are known as "bus", "ring", and "star". A bus
topology is one in which all of the devices are connected in a line. A ring topolgy is similar to a
bus topology except the devices on the ends are connected forming a closed loop. A star
topology is one in which one device is in the center and connects to all of the other devices like
the hub and spokes of a wheel.
2.3.3
Bus
The terms media, bus, and channel are often used interchangeably to describe the physical
carrier of the communication. While the terms are related they do have separate meanings.
A bus refers to the "backbone" of a communication channel. When a bus topology is used all of
the nodes on the network are physically connected to the bus.
2.3.4
Channel
A channel refers to an uninterrupted physical communication path. Most PowerCommand
networks consist of a single channel. More sophisticated networks will have multiple channels. A
single channel consists of only one protocol and one type of media. Devices such as routers
and gateways can combine separate channels consisting of different protocols and media into a
single network. Routers and gateways will be discussed in detail in Chapter 3.
Figure 5 illustrates the concept of a channel and a bus. Each of the routers separate the
network into two channels: a Lon channel and an Ethernet channel. Each of the channels is
connected in a bus topology with each device on the channel connected to a common pair of
wires. This pair of wires is referred to as the bus.
A044E523 (Issue 1)
2. Networking Basics
4-2013
FIGURE 5.
2.3.5
Media
Media is the generic term for the physical carrier of the signal, whether it is copper wire, fiber
optic cable or, as in the case of wireless communication, air.
10
A044E523 (Issue 1)
4-2013
2.3.6
2. Networking Basics
Node
The term "node" is a generic term referring to any device connected to a network. A node could
be a generator set control, an annunciator, a router or a web server, or any device on the
network.
2.3.7
Bandwidth
Bandwidth refers to the speed of communication over the network. It is the amount of data that
can be transmitted in a fixed amount of time, usually expressed as bits per second (bps or
bit/s). The term bandwidth is also used to express the finite amount of time in which certain
communication may occur, so phrases like thats not an efficient use of bandwidth refers not
only to the speed of the propagation of the data but also to the total amount of time that a
transmitting device uses to access the communication channel.
2.4
Evolution of Networks
2.4.1
Discrete Communication
Prior to the introduction of microprocessors to control systems, devices communicated over
discrete wires with a separate wire representing each signal. Discrete wiring is not inherently
bad and is necessary when signals need to transmit significant levels of power such as when
activating a relay. Switchgear applications still include very large wiring harnesses. However,
when signals are used just to convey information, communication networks offer materials and
labor cost savings and advantages in scalability over bundles of discrete wire. As
microprocessors became widespread in control devices, network communication became a
viable and economical alternative to communication via discrete wires.
2.4.2
Serial Communication
As personal computers (PCs) became common, a demand arose for PCs to be able to
communicate with other devices, such as printers and modems. Simple serial communication
protocols were developed to support communications between two devices. As communication
equipment became available PCs and Programmable Logic Controllers (PLCs) began to use
serial communication to monitor and control industrial networks.
Serial communication can be defined as the process of sequentially sending data one bit at a
time over a common communication channel. This allows a single wire or pair of wires to
transmit many different pieces of information. Data is transmitted by a series of electrical pulses
and translated into a series of binary digits or bits, each of which can have one of two values,
commonly represented as 0 or 1.
Establishing timing between sending and receiving devices is necessary for the receiving device
to recognize where one character (meaning a number or a letter) ends and the next begins
within a string of pulses. There are two methods for achieving this, known as synchronous serial
communication and asynchronous serial communication. With synchronous communication a
separate clock signal is transmitted between the two devices and a new message begins based
on a pulse of the clock signal.
A044E523 (Issue 1)
11
2. Networking Basics
4-2013
FIGURE 6.
A parity bit is often used as a simple form of error checking. When a parity bit is used the
sending device will set that bit so that it is transmitting either an even or odd number of 1s in the
packet based on a predefined setting in the devices for even or odd parity. For example, if the
packet consists of 7 data bits and even parity is selected, then if the 7 data bits are 1010001
(3 1s) then the parity bit should be a 1 so that there are an even number of 1s. If the parity bit
is 0, the receiving device would recognize that there has been an error in communication.
2.5
Early Protocols
One of the earliest forms of serial communication is Morse code transmitted over telegraph
wires. The signals were simple electrical pulses transmitted over the wires, a short pulse
represented by a dot and a longer pulse represented by a dash. Each letter of the alphabet
was represented by a series of dots and dashes so a message could be broken down to this
serial data, transmitted, received, and re-assembled by the receiver, thus Morse code can be
considered one of the first communication protocols.
2.5.1
RS-232
As communication equipment became more widely used, there was a need to develop a
standard communication protocol so that equipment from several different manufacturers could
communicate with each other. The standard protocol was developed by the Electronic Industries
Association (EIA) and was known as RS-232. The RS-232 standard defines communication
between two specific types of equipment known as Data Terminal Equipment (DTE, typically a
PC) and Data Communications Equipment (DCE, typically a modem) over a distance of no
more than 50 feet (15 meters). RS-232 also specified the pin designations on the 9- and 25-pin
connectors that were most common at the time. RS-232 defines separate transmit and receive
wiring paths with the transmit pin on one device physically connected to the receive pin on the
other device. This means that both devices can transmit at the same time. This is known as full-
12
A044E523 (Issue 1)
4-2013
2. Networking Basics
duplex communication (see Figure 7). Although it was designed for a very narrow scope of
applications (communication between only two devices no more than 50 feet (15 meters) apart)
RS-232 is still commonly used in those applications today due to its simplicity and relatively low
cost to develop and deploy. Cummins Power Generations proprietary Mon protocol, which is
used by the InPower service tool to connect to controls, is built on the RS-232 physical layer.
FIGURE 7.
In the 1990s, control devices with microprocessors and networking capability were becoming
more popular. Networks consisted of several devices spread over large distances. A
communication channel that is shared by multiple devices presents several challenges beyond
what is required for communication between only two devices. It is necessary for signals to
travel over greater distances without distortion. The speed of communication becomes more
important. There is a need for identifying which device a message is intended for. When there
are many devices connected to a common communication channel it is not practical to have
individual transmit and receive lines for each device. This means that multiple devices cannot
transmit messages at the same time so there needs to be some method to determine which
device can access the channel. Protocols that support multiple devices on a network need to
address these concerns of distance, bandwidth, addressing, and media access.
2.5.2
RS-485
RS-485 was developed as the standard for connecting more than two devices on a network
over longer distances. Unlike RS-232, which has separate circuits for transmitting and receiving
signals, RS-485 specifies a single circuit for both transmitting and receiving signals, known as a
transceiver. The RS-485 transceiver is designed in such a way that multiple transceivers can be
connected to the network without affecting the one transceiver that is transmitting a signal (see
Figure 8). RS-485 was designed so that signals can travel long distances without being
attenuated by resistance and inductance in the wires and have a high level of noise immunity.
FIGURE 8.
RS-232 allowed two devices to send messages at the same time by having separate send and
receive paths between the two devices. Since RS-485 uses only a single transceiver for both
transmitting and receiving data, this is not possible. At any given time a device may either
transmit or receive data but the device cannot perform both operations. This is known as halfduplex communication. There is also a 4 wire implementation available on the market which
supports full-duplex communication. These devices consist of two separate RS-485
transceivers, so that one can transmit and the other can receive at the same time.
A044E523 (Issue 1)
13
2. Networking Basics
4-2013
Cummins Power Generations PCCNet protocol, used to communicate between a generator set
and its accessories, such as human machine interfaces (HMIs) and annunciators, is built on the
RS-485 physical layer. As the RS-485 standard defines only the physical layer of
communication, it defines distance and bandwidth. Addressing and controlling access to the
communication channel is outside of the scope of the RS-485 standard and is defined by higher
level protocols.
2.6
Addressing
Each device on a network must have a unique identifier often known as an address. Many
protocols will define both a logical and a physical address. The logical address is assigned
when the network is commissioned; if the same device is used in a different network in the
future, it will likely be assigned a different logical address. Also, if a device fails and needs to be
replaced, the new device will most likely be assigned the same logical address as the device
that it replaced.
Some protocols, such as Ethernet and LonWorks, will also specify a physical address. The
physical address is assigned by a standards body, is unique in the world, and is permanently
assigned to a specific device.
2.7
14
A044E523 (Issue 1)
4-2013
2.8
2. Networking Basics
Determinism
When networks are used to transmit time critical data for controllers, there is a need for the
communication to be deterministic, meaning that the amount of time it takes for a certain
message to travel between devices must be predetermined and consistent. In this case,
protocols will include some priority arbitration scheme so that critical messages will pre-empt
noncritical messages and devices transmitting these messages will get priority in accessing the
bus.
2.9
A044E523 (Issue 1)
15
2. Networking Basics
4-2013
FIGURE 9.
SWITCHED ETHERNET
In a switched Ethernet network the device will send messages to an Ethernet switch. The
Ethernet switch will check the message for errors, and then forward it to the intended recipient.
The message will also include a prioritization code so in the event that multiple devices are
attempting to communicate with the same recipient, the higher priority message will be
transmitted first.
A full duplex switched ethernet allows devices to transmit and receive data at the same time and
eliminates collisions by having only one device connected to each port of the switch. These two
features enable deterministic communication which makes Ethernet suitable for industrial control
networks.
Notice that Figure 9 shows two small networks with Ethernet switches connected by routers.
Routers are similar to switches in that they both manage network traffic. The difference is that
switches direct traffic within a local network while routers direct traffic between networks.
Routers enable networks using different types of media or different protocols to interact with
each other. The most common example of this is a control network using routers to connect to a
remote monitoring system over the internet. Switches and routers will be discussed in detail in
Chapter 3.
16
A044E523 (Issue 1)
4-2013
2.10
2. Networking Basics
2.11
A044E523 (Issue 1)
17
2. Networking Basics
4-2013
3. A CAN network for communicating between the genset controller and the Engine Control
Module (ECM). Can is a commonly used protocol in the automotive industry used to
communicate between the engines and other control systems in a vehicle. It is the standard
protocol used by Cummins ECM's in all kinds of applications and in a genset is used only
for communicating between the genset control and the ECM.
18
A044E523 (Issue 1)
Network Devices
3.1
Overview
In this chapter we will identify the types of devices that are commonly found in a power
generation system network. We will refer to Cummins Power Generations products as
examples.
We will discuss five types of devices: I/O devices, intelligent controllers, network interface
cards, network management devices (hubs, switches, routers, gateways), and web servers.
3.2
Input/Output Devices
Input/Output Devices allow controllers to get information from the physical world and to control
physical components. I/O devices convert data between physical states or values and logical
signals communicated over a network. For example, the position of a switch, a physical state, is
converted to a logical value and communicated to a controller, or a controller will send a logical
command to activate a relay, or change its physical state. Defining input and output
requirements for system components is a key requirement in the overall system design.
An input device will convert some physical parameter, such as a position of a switch or a
voltage level, to a logical value, which will be communicated over a network. An output device
will receive a logical signal over a network and change the state of an output, which will result in
some physical change to a piece of equipment, such as grounding a relay coil, which will cause
the relay contacts to close. An output from one device can be an input to another device.
Examples of Cummins Power Generations network I/O devices include the LonWorks DIM (see
Figure 10), CCM modules, and the Aux101.
FIGURE 10.
I/O can be either digital (discrete) or analog. Digital refers to signals that can be in one of two
states, for example on or off, or open or closed. Analog I/O refers to a value that is continuously
variable. Voltage, amperage, and temperature are examples of analog signals.
A044E523 (Issue 1)
19
3. Network Devices
3.2.1
4-2013
Digital Inputs
Digital inputs, also known as discrete or switch inputs, are used to indicate switch positions in a
power system. Oil pressure switches, temperature switches, status of generator run/off/auto
switches, and auxiliary position indicating switches on circuit breakers and transfer switches are
the most common uses for digital inputs. Digital inputs are inexpensive and simple to implement
compared to analog inputs so it is also common to use digital switches for analog parameters
such as oil pressure or coolant temperature, which will activate when the parameter has
become higher or lower than some threshold value. This method does have a disadvantage with
respect to using an analog input in that with a switch a control is not able to recognize a failed
switch. With an analog signal the control is able to differentiate a failed sensor from a parameter
which has passed some fault threshold because a failed sensor will yield an "out of range"
value.
Digital inputs are typically activated by a contact closure between the input line and either
ground or a control voltage reference provided by the control device. Inputs that are activated by
a closure to ground are called active low; those that are activated by closure to a control
voltage are called active high.
In some cases, a pull up or pull down resistor (see Figure 11) is required on a digital input to
positively connect the input when it is in its passive state. For example, if a customer activates
an input by switching it to the control voltage (active high), a pull down resistor is required
between the input and ground so that the I/O device will read the input as a low value when
the customers switch is open. This is the case with the digital inputs on Cummins Power
Generations CCM boards when they are used as Active high inputs. This is not a requirement
with the Aux101, DIM devices, or CCM inputs when they are used in an active low
configuration. In these cases, the pull up or pull down function is implemented on the control
boards.
FIGURE 11.
3.2.2
Digital Outputs
Digital outputs are used to start or stop generators, operate equipment, such as fans or louvers,
close motor starting contacts, and indicate conditions based on some event that occurs in a
control system. For example, with the Aux 101, generator events are mapped to relays using
the service tool. With some generator set controls the Human Machine Interface (HMI) is also
capable of mapping events to output relays. With the CCM and DIM, the signal is initiated by
some controller and mapped to outputs on the CCM or DIM using configuration software.
20
A044E523 (Issue 1)
4-2013
3. Network Devices
A common method for implementing a digital output is to use a form C contact (see Figure
12). A form C contact consists of one normally open and one normally closed contact and a
common line between the two contacts, which ensures that one contact is always open and one
is always closed to the common line and activating the output will change the state of both
contacts. The digital outputs on Cummins CCM, DIM, and Aux 101 are form C contacts.
FIGURE 12.
FORM C CONTACTS
The amount of current that a digital output circuit can source or sink is a key consideration in
determining how to apply the device. If the current rating of the digital output is not sufficient to
operate the piece of equipment it is connected to, an external pilot relay will be required.
3.2.3
Analog Inputs
Analog inputs measure continuously variable signals. The parameter to be measured is
converted to a direct current (DC) voltage and read by an input device. The DC voltage is
converted to a numerical value using an analog to digital (A/D) converter and that numerical
value is used by a control or communicated over a network. The sensing circuit will be scaled
so that the maximum current or voltage that the input device can accept will correspond to the
maximum value that the signal can have. For example, 0-5V DC is a common standard for
analog input devices. 0V would correspond to the minimum value and 5V would correspond to
the maximum value that the sensed parameter can have. Analog inputs are used to measure
generator output voltage and current, temperatures of oil, exhaust, ambient air, and alternator
windings, oil pressure, fuel level and voltage, and speed bias lines.
Some sensors, such as temperature and pressure senders, have a resistance that is
proportional to the sensed parameter. A fixed current is passed through the resistive element
which will create a voltage for the I/O device to read. With Cummins Power Generations Control
Communication Module (CCM) and Aux 101 boards, the current source is provided on the
board. Since oil pressure and different temperature senders all have very different
characteristics, they will require different levels of current to generate voltages in the appropriate
range to be measured by I/O devices. With the CCM there are different current sources for each
analog input and two of the inputs have two different levels of current available, which can be
selected using an on-board switch. With each of the Aux 101 inputs, the operator uses the HMI
to select what parameter is measured and the current source for that input is configured
appropriately.
Some controls can read voltage and speed bias inputs. These are simply customer supplied
potentiometers (variable resistors) through which a fixed current is passed to create a voltage
which can be read by the A/D converter.
Control and communication devices can read phase voltages and load current. Phase voltages
are stepped down through transformers to a level that is suitable for rectification to create a DC
voltage and then read by the A/D converter. Load current is passed through a current
transformer (CT) for sensing and the output current of the CT is run though a burden resistor
that creates a voltage that can be measured. With Cummins Power Generations equipment
there are typically two sets of CTs in the sensing circuit, one set right at the generator output or
load terminals of a transfer switch and a second set of smaller CTs mounted on a circuit board.
A044E523 (Issue 1)
21
3. Network Devices
4-2013
There are also devices on the market that create a current which is proportional to a measured
parameter. A common standard is 4 20 mA devices. To monitor this with a CCM, the current
is passed through a known resistance (off of the CCM board) and the CCM can read the voltage
across the resistor.
To make use of the values created by the A/D circuit, it is necessary to know how the
parameters are scaled. Scaling will depend on the physical properties of the sensing circuit and
the resolution of the A/D converter. For example, a temperature sensing circuit will convert
temperature to voltage depending on the resistance/temperature characteristic of the sender
and the current source. The A/D converter will convert that voltage to a number of counts
based on its resolution. For instance, a 10 bit A/D converter will convert the measured voltage to
a number of counts between 0 and 1024 (2^10). When the conversions between the measured
parameter and analog voltage, and between analog voltage and A/D counts are known, the
conversion between the measured parameter and counts can be calculated.
In many cases, the I/O device, or the control that gathers input from the I/O device, will make
this calculation or there will be a tool to make this calculation. In other cases, it will be easier to
establish the conversion through a test if it is known that the relationship between the measured
parameter and the analog voltage is linear. Simply reading the A/D output for 2 known inputs
will allow a simple calculation of the conversion factor. For example, consider a temperature
sender. Through test it is determined that at a temperature of 197 F the sender had a voltage
of 3.23V and at 172 F the voltage was 2.9 V. With these 2 points we can determine the sender
gain and the offset for the sender. The gain defines how much the voltage changes for each
degree of temperature change. It is simply the ratio of the difference in temperature to the
difference in voltage. In this example it is (197-172)/(3.23-2.92) = 80.65 F/V. This means that
for every 80.65 degrees that the temperature changes, the voltage will change by 1 volt. The
fact that the gain turned out to be a positive number indicates that as the temperature increases
the voltage will increase. Some types of sensors have a negative gain, meaning that as the
sensed parameter increases the corresponding voltage decreases.
Some senders do not have an output of 0 volts when the parameter being sensed is at 0. In
those cases we need to calculate an offset to fully define the sender. To continue the example
above, we will calculate the offset for that sender, which is the temperature that would
correspond to 0 volts. We can calculate this using the sender gain and one point. We'll use the
2.92V, 172 degrees point. The equation is 172 - 2.92*80.65 = -63.5. This means that 0 volts
would correspond to a temperature of -63.5 degrees. A plot of the linear relationship between
temperature and voltage is shown in Figure 13. Note that different analog devices have
different calibration tools and methods. This is just one example.
FIGURE 13.
22
SENDER CHARACTERISTIC
A044E523 (Issue 1)
4-2013
3.2.4
3. Network Devices
Analog Outputs
There are analog output devices available on the market which are typically controlled by
programmable logic controllers (PLCs). These devices commonly source a 4 20 mA current as
their output.
In general, analog outputs are not very common in network products. When a process needs a
controlled, variable analog voltage it typically uses a customized controller rather than an offthe-shelf networked product. While there are standard digital to analog converters (D/A)
available for integration into a circuit board, most analog outputs are generated by using a
pulse-width modulated signal (PWM) which is filtered to yield a steady analog signal. Cummins
Power Generation does not produce any analog output devices for use on a network but uses
them in system control products that interface with 3rd party equipment.
3.3
Intelligent Devices
Intelligent devices have a microprocessor and are capable of making decisions based on
information received over the network and interpreting and processing the information it
receives from its inputs before sending the information over the network. Generator and transfer
switch controls and HMIs are examples of intelligent devices. To illustrate, an HMI will translate
a series of button clicks by an operator (digital inputs) and send the information to a controller,
which it will then process and generate a digital output to start the generator. Intelligent devices
are often capable of operating as a standalone piece of equipment; they do not always need the
network communications to function.
3.3.1
FIGURE 14.
A044E523 (Issue 1)
23
3. Network Devices
3.4
4-2013
3.4.1
Hub/Repeater
Hubs and repeaters are simple devices that repeat all of the network signals to all network
devices to which they are connected. They do not process data or make any type of decision
regarding where to send the data.
A repeater has two ports. It connects two segments of a network. Since they are able to amplify
a weak signal, repeaters are able to extend a network over large physical distances.
A hub has multiple ports and allows devices to attach to a network in the star configuration.
(See Section 4.2 on page 29 for discussion on network topology.) A hub receives data and
broadcasts the data to all of its ports.
Hubs and repeaters are becoming less common as they are not able to meet demands for
faster networks with more devices. More often than not, they are replaced by Ethernet switches.
3.4.2
Switch
A switch is similar to a hub except instead of broadcasting all the messages it sees, it checks to
see that there are no errors in the message packet and if there are no errors, it will transmit the
message only to its intended recipient.
By sending only good messages to their intended recipient, switches increase the speed of the
network by eliminating a lot of unnecessary and erroneous traffic. Switches can also improve
network speed by eliminating collisions.
Ethernet switches divide a network into segments called collision domains. A collision domain
is defined as a set of network devices whose packets can collide with each other. As discussed
in Chapter 2, if two devices on a network try to talk at the same time there is a collision. When
a collision occurs, all the data is lost and each device will delay its message before attempting
to transmit again.
Because a switch manages traffic between ports, there will be no collisions between devices on
separate ports. Only devices that are connected to the same port of a switch can experience
collisions. Devices connected to the same port on a switch are said to be in the same collision
domain. To completely eliminate the possibility of collisions, only a single device is connected to
each port of a switch. There is only one single device in each collision domain.
To maximize speed of network communication, full duplex switching is employed. Full duplex
communication allows devices to both transmit and receive messages at the same time. This is
accomplished by each device having separate transmit and receive lines. By using full duplex
communication with only one device on each collision domain the speed of communication is
optimized. It is limited only by the characteristics of the transceiver and the media.
3.4.3
Bridge
A bridge is similar to a switch except that a bridge has only two ports where a switch may have
many ports. A bridge can connect two network segments separated by large distances or two
segments that are communicating over different types of physical media.
24
A044E523 (Issue 1)
4-2013
3.4.4
3. Network Devices
Routers
Routers are similar to switches in that they both manage network traffic. The difference is that
switches direct traffic within a local network; whereas, routers direct traffic between
networks. Where a switch communicates between network segments that are physically
connected to that switch, routers can connect two devices that are physically separated by great
distances with multiple routers in between them. Routers operate based on a devices logical
address rather than physical location. To illustrate, one can connect his/her laptop to their
internet service providers server from their home, their office, an airport or their favorite coffee
shop. The physical location does not matter; it is the IP (Internet Protocol) address, the logical
address, that matters.
Routers enable networks that use different media, access, and addressing protocols to
communicate with each other. (They can translate between different Physical, Data Link, and
Network Layer protocols, and layers 1, 2 and 3 of the OSI model. The OSI model will be
discussed further in Chapter 6.) They do not translate between higher level languages
(Application Layer), such as LonWorks and Modbus.
A common example of this is in a campus application where generators, transfer switches, and
switchgear are all located in an equipment room and are to be monitored in a control room a
great distance away. There is an Ethernet connection over fiber optic cable between the two
rooms. The equipment communicates via LonWorks over twisted pair wire. In each of the two
rooms there would be a router with one channel connected to the twisted pair wire and the
second channel connected to the fiber. The router does not translate the high level language
(LonWorks), it just manages the transmission of data from the twisted pair network, over the
fiber optic cable, and back to the twisted pair network at the other end. See Figure 15.
FIGURE 15.
3.4.5
Gateways
A gateway is a device that enables two network segments of different protocols to communicate
with each other. A gateway manages all layers of the OSI protocol stack from the physical
media to the high level language. Cummins Power Generations ModLon Gateway (see Figure
16), which converts LonWorks to Modbus data for Cummins products, is a good example of a
gateway. It should be noted that gateways can slow down communication and in some cases
can compromise reliability as they introduce single points of failure into the system.
A044E523 (Issue 1)
25
3. Network Devices
4-2013
FIGURE 16.
MODLON GATEWAY
In some cases, gateways that connect serial protocols, such as RS-232, to Ethernet are known
as Device Servers. They are typically used to connect devices such as printers and fax
machines which have only a serial connection to an office LAN so that the equipment can be a
shared resource.
3.5
Web Servers
Web servers are devices that reside on a network and serve web pages consisting of data
acquired from equipment on that network. In most cases web servers will be capable of sending
emails based on alarm conditions. Typically, with a web server in a power generation
application, the owner of the system needs to provide internet access at the site where the
generator is and will need to provide the appropriate IP and email server addresses to connect
to the internet.
Cummins PowerCommand 500 and 550 and iWatch 100 are examples of web servers (see
Figure 17). The PowerCommand 500/550 and the iWatch 100 communicate with generator and
transfer switch controls over LonWorks or Modbus and serve web pages that display the current
status of the equipment. Operators can view this data with their own web browser. If operators
have the proper security access they can also start the generator set or initiate a test of the
transfer switch through the web page. Web servers will also send text messages or emails on
alarm conditions so that service personnel can be alerted that there has been some event in
their power system. Cummins Power Generations DMC200 and DMC300 digital master
controllers also have integral web servers. Components in some equipment may also have webservers. For example, it's physically possible for a generator set control to have an on-board
web server. However, it is more practical to have a web server for a system than a single
component in a multiple component system.
26
A044E523 (Issue 1)
4-2013
3. Network Devices
FIGURE 17.
A044E523 (Issue 1)
27
3. Network Devices
4-2013
28
A044E523 (Issue 1)
Network Characteristics
4.1
Overview
This chapter defines physical network characteristics, topology, bandwidth, and distance. It
describes the interaction between these characteristics and provides a baseline for comparison
of various media and protocols.
Selection of topology, bandwidth, and distance impose tradeoffs on a system. Typically,
bandwidth and distance conflict with each other. Increasing the physical distance between
devices decreases the maximum speed of data transmission. Topology choices will affect the
other capabilities as well, with the simplest topology allowing for the greatest distance and
bandwidth, while more complex topologies will impose limitations.
4.2
Network Topology
Topology refers to the physical shape and interconnection scheme of the network. The three
most common topologies used in control networks are known as bus, ring, and star. Some
of the more complex topologies such as tree and mesh topologies are combinations of the
three basic topologies. Echelon Corporation, creators of the LonWorks network platform, uses
the term Free Topology to position LonWorks as flexible enough to be configured in any
topology. Some topology examples are shown below (Figure 18).
FIGURE 18.
A044E523 (Issue 1)
29
4. Network Characteristics
4.2.1
4-2013
Bus Topology
With a bus topology, all of the devices are connected to a common backbone or bus. A bus
topology is the simplest and least expensive to install. A bus topology will allow for the greatest
distance and bandwidth. With Cummins Power Generation networks, we always recommend
using a bus topology except where redundant communication paths are required.
Bus topology is very similar to a daisy chain topology. The terms bus and daisy chain are
often used interchangeably but there is a difference. With a bus topology, each device is
connected to the bus, and with a daisy chain, each device is connected directly to the adjacent
device. With a bus network, each device is connected to the bus through a drop or a stub. A
daisy chain can be thought of as a multi-drop bus network with a drop length of 0.
Bus topology networks often require termination circuits at the end of the bus. The equipment
manufacturer should be able to define when and what type of terminator is required. See
Chapter 5 for more information on termination.
Bus topology advantages include:
Maximizes distance and bandwidth,
Easy to install and troubleshoot,
Simplifies expansion,
Inexpensive Minimizes the amount of cable required.
Bus topology disadvantages include:
A bus topology doesn't allow for a redundant communication path. A single break in the
communication path will isolate devices from the network and may result in complete
network failure; however, generally both sections of network continue to function. Some
networks are intelligent enough to detect failure to communicate and annunciate that at the
component level, mitigating the issue. When a network can detect a failure in the bus it has
advantages over traditionally wired systems, because they cannot detect when
interconnecting wiring has failed.
All network traffic exists on the same wire, potentially creating collisions and reducing
bandwidth. Media access strategies can minimize this effect.
4.2.2
Ring Topology
With a ring topology, devices are connected to each other in a closed loop. The closed loop
provides redundancy in the communication path as a single break in the path does not disable
communication. It is a relatively simple topology to troubleshoot and it is what Cummins
recommends in applications where redundancy is required.
Ring topology advantages include:
Redundant paths make the network reliable. A single break in the communication path
does not disable the network.
Relatively simple to troubleshoot.
Ring topology disadvantages include:
Installation can be expensive and difficult depending on the physical location of all of the
devices. A ring topology may require long runs of cable and conduit to close the loop.
The long cable runs and the requirement to connect each node to two other nodes may
limit the applications in which a ring topology is practical.
30
A044E523 (Issue 1)
4-2013
4.2.3
4. Network Characteristics
Star Topology
A star topology network includes a central device to which all of the other devices are
connected. This central device is most often one of the traffic management devices discussed in
Chapter 3 (switch, router, hub, etc.), but may also be a master controller like a PLC. When the
central device is a router, it is typically connected to other routers and the local star network is
part of a larger tree or mesh network.
Star topology advantages include:
Each device has its own path which eliminates the possibility of collisions.
A star topology is more robust than a bus because a break in a communication path only
results in the loss of one device.
Star topology disadvantages include:
The central device represents a single point of failure.
Distance is limited compared to a bus topology.
Bandwidth will be reduced because messages need to go through the central device
before reaching their final destination. An exception to this may be when the central device
is a full-duplex Ethernet switch. The fact that the switch eliminates the possibility of
collisions and enables all devices to transmit and receive messages at the same time will
in most cases more than compensate for the extra device in the communication path.
4.2.4
Other Topologies
The three basic topologies discussed above are appropriate for most control networks. In some
information networks however, there are occasions when, due to the number of devices on a
network or the devices being physically separated by a long distance, a more complex topology
is required. In most cases, these topologies are combinations of simple networks which use the
three basic topologies and are separated by routers or switches. The two most common of
these topologies are tree and mesh topologies.
4.3
Bandwidth
Bandwidth is the amount of data that can be transmitted in a fixed amount of time. It is usually
expressed in bits per second. It is a function of the physical properties of the media and the
method used to transmit data. (These are characteristics of Physical Layer protocols as defined
by the OSI model. This is discussed in Chapter 6.)
A044E523 (Issue 1)
31
4. Network Characteristics
4-2013
The term bandwidth is often used interchangeably with network speed, but bandwidth is actually
a measure of network capacity. It is a measure of how fast data can travel through some media
rather than how fast data is actually travelling. A common analogy is to think of data flowing
through a pipe. Bandwidth is analogous to the size of the pipe. Increasing the bandwidth allows
more data to flow during some period of time but does not necessarily cause more data to flow.
Along with bandwidth, the other key factor that affects the speed of a network is latency or
delay. The term latency is defined as the amount of time it takes a packet of information to
travel from its source to its destination. Topology is the key factor affecting latency in control
networks as it takes a finite amount of time for routers and switches to process and forward
data. In Wide Area Networks, distance can significantly affect latency, particularly with cellular or
satellite communication as data travels at a finite rate. Latency is not always predictable in Wide
Area Networks as routers may not always send data to its destination over the same path.
4.4
4.5
Distance
Like bandwidth, a networks maximum distance is determined by the media and by the
properties and methods of the device that is transmitting the data. Resistance in a cable will
attenuate signals that are transmitted over it. As a cable gets longer, resistance, and therefore
attenuation, increase until the signal is attenuated so much that the receiving device is not able
to interpret it.
32
A044E523 (Issue 1)
4-2013
4. Network Characteristics
Properties of the transmitting device will also affect the maximum network distance. For
example, RS-232 and RS-485 networks can both communicate over twisted pair wires, but RS485 devices communicate in a manner that minimizes the effects of attenuation (and minimizes
the effects of interference) so that it can communicate over a much longer distance.
With some protocols, such as CAN and RS-485, the maximum speed and maximum distance
have a trade off. This tradeoff is caused by the capacitance of the cable. Capacitance delays
voltage transitions between states. If transmission of a second bit begins before the first bit
reaches its intended state, communication will fail. Data cannot be transmitted faster than it can
be propagated so a long distance may limit the maximum speed.
Network topology can also affect the maximum distance. For instance, with LonWorks, a bus
topology which is terminated (grounded through a resistance) at both ends has a higher
maximum distance than a LonWorks network with any other topology. This is due to the
terminated bus allowing signals to have better immunity to interference.
Some protocols, such as Ethernet and LonWorks, have routers or repeaters available which
allow extending the physical length of the network. These devices will re-transmit the signals
that they receive so distance limitations imposed by the media will apply to each segment of the
network connected to the device rather than to the entire network. LonWorks has a limit to the
number of routers that can exist on a single network, so that sets the maximum total length of
the network. Ethernet does not have any restriction to the number of devices on the entire
network so there is no practical limit to the distance of an Ethernet network.
A044E523 (Issue 1)
33
4. Network Characteristics
4-2013
34
A044E523 (Issue 1)
Media Considerations
5.1
Overview
This chapter discusses media over which communication signals travel. General application
considerations such as termination, grounding, and shielding are covered, as well as commonly
used media and how and where they are used. This chapter will focus on wired media because
the vast majority of industrial networks are wired. Some basic wireless concepts will be
discussed briefly at the end of the chapter.
5.2
Cable Concepts
We will begin by discussing the properties of cable that allow or inhibit successful transmission
of data.
5.2.1
Attenuation
Attenuation is a measure of the decrease in signal strength along the length of the media due to
resistance in the media and outside interference. Attenuation is typically expressed in decibels
(dB) as the ratio of output to input signal level and is the primary reason that network standards
specify maximum cable lengths.
5.2.2
FIGURE 19.
The series elements raise the effective impedance of the cable and cause the current to lag the
voltage. The shunt elements reduce the effective impedance and cause the current to lead the
voltage, so the characteristic impedance is the result of those two counteracting effects.
A044E523 (Issue 1)
35
5. Media Considerations
4-2013
With long lengths of cable the electrical characteristics of the cable can result in transmission
line effects, such as standing waves and signal reflections, which could distort the signal. A
cable is generally considered electrically "long" when the length of cable exceeds 1/4
wavelength of the signal. Wavelength can be approximated by dividing the speed of light by the
frequency of the signal. Therefore as cable length and frequency of the signal increase the
transmission effects become more prominent. Transmission line effects can be minimized by
placing a terminating circuit in the network. For most networks using twisted pair wire, the
terminating circuit is simply a resistor that matches the characteristic impedance of the cable
placed at each end of the network. This is known as terminating the network (see Figure 20).
NOTE:
LonWorks network termination requirements are more complex than this and
are discussed in the LonWorks Section 5.5.4.
FIGURE 20.
NETWORK TERMINATION
When connecting network segments together, such as when expanding an existing network
using different cable, it is necessary for the cables on the two segments to have the same
characteristic impedance. In situations where this is not possible, such as when expanding an
existing network using different cable, a balance resistor or Balun may be placed between the
two cable segments to give the entire cable run a consistent characteristic impedance.
5.2.3
Electromagnetic Interference
Electromagnetic interference (EMI) affects electrical signals whether traveling in a cable or
being broadcast through the atmosphere. EMI can disrupt the signal or degrade it sufficiently so
that it becomes unintelligible. EMI is produced by switching electromagnetic fields. Motors,
alternators, variable frequency drives (VFDs) and other motor control circuits, fluorescent
lighting and communications equipment are all sources of EMI. Shielded and twisted pair cable
is used to improve the cables immunity to EMI. Fiber optic cable, because it uses light rather
than electrical pulses to communicate, is not affected by EMI.
5.2.4
Crosstalk
Crosstalk is interference from adjacent cables that are close together and running side by side.
It results from signals of one cable being magnetically coupled to another cable and can be a
concern when there are multiple pairs of cable within a single insulating jacket. Conductors are
twisted about each other in twisted pair cable to counter this problem. As the cables twist about
each other, the polarity of the signal reverses, nullifying the crosstalk to some extent but not
eliminating it fully. The measurement for crosstalk is called NEXT (Near End Crosstalk) and is
expressed in dB as a measure of the signal loss in coupling from one pair to another. Cable
specifications specify a minimum level of crosstalk coupling loss for various frequencies.
36
A044E523 (Issue 1)
4-2013
5.2.5
5. Media Considerations
Shielding
Shielded cables are used to combat EMI. The shield can reflect EMI energy but its primary
purpose is to conduct EMI energy to ground. In order for the shield to work, it must be grounded
at only one point. Failure to make a good ground connection will limit the ability of the cable to
shield the conductors from EMI. Grounding the shield at more than one point will allow EMI
energy to circulate on the shield and potentially allow coupling energy onto the conductors. Note
that grounding here refers to the control ground rather than an equipment ground. Many
controls that use shielded cable for communication have a terminal for connecting the shield.
A shield will affect the characteristic impedance of the cable so it is not appropriate to replace
unshielded cable with shielded cable without considering the network termination circuits.
There are two types of shielding typically used for cables: foil (Figure 21) and braid (Figure 22).
Foil shielding uses a thin layer of aluminum, typically attached to a carrier, such as polyester, to
add strength and ruggedness. A braid is a woven mesh of bare or tinned copper wires. Because
copper has higher conductivity than aluminum and the braid has more bulk for conducting noise,
the braid is more effective as a shield but it adds size and cost to the cable. A braid is also more
durable than foil in applications where the cable will be flexed.
FIGURE 21.
A044E523 (Issue 1)
37
5. Media Considerations
4-2013
FIGURE 22.
For very noisy environments, multiple shielding layers are often used. Most common is using
both a foil and a braid. In multiconductor cables, individual pairs are sometimes shielded with
foil to provide crosstalk protection between the pairs, while the overall cable is shielded with foil,
braid, or both.
5.2.6
5.3
Cable Types
There are three types of cable typically used in control networks: twisted pair cable, fiber optic
cable, and coaxial cable. Each will be discussed in this section.
38
A044E523 (Issue 1)
4-2013
5.3.1
5. Media Considerations
FIGURE 23.
FIGURE 24.
In relatively high EMI environments, such as environments with motors and drives, it is often
necessary to use shielded twisted pair (STP) cable. The shield gives the cable better immunity
to EMI but is more expensive and more difficult to work with than UTP, and if the shield is not
grounded properly the shielding effects are negated.
A044E523 (Issue 1)
39
5. Media Considerations
5.3.2
4-2013
Fiber-Optic Cable
Fiber-optic cabling employs one or more optical fibers consisting of a central glass or plastic
fiber surrounded by a glass cladding and a plastic outer sheath (see Figure 25). Light-emitting
diodes (LEDs) or lasers send light through the fiber to a detector, which then converts the light
into electrical signals. Because light is being transmitted rather than electrical pulses, fiber optic
cable is immune to EMI and does not have the same attenuation effects as twisted pair cable.
Fiber optic cable has the ability to transmit over much longer distances and at much greater
speed than copper cable. Fiber optic cable and connectors are more expensive than copper
cable and installation requires special tools and a higher skill level than installation of copper
cables.
FIGURE 25.
FIBER-OPTIC CABLE
There are two common types of fiber cables, single mode and multimode.
Single mode cable has only one light beam in it. The cable has a small diameter core and is
made of glass. A laser is usually the light source and the cable can extend several
miles/kilometers.
Multimode cable simultaneously transmits several light beams. The core is wider than what is
used in single-mode and typically made of plastic. Instead of lasers, the light sources are
usually LEDs. Multimode cable has greater capacity and is less expensive than single-mode
cable; however, distance is limited to a few hundred feet/meters rather than
miles/kilometers. For this reason, multimode cable is usually used inside buildings and single
mode cable is used between buildings and over long distances.
5.3.3
Coaxial Cable
Coaxial cable was the first major cable used for Local Area Networks (LANs). It has now fallen
out of favor and is replaced with cable that is more flexible and less expensive, such as
unshielded twisted pair, or by cable which can offer greater distances or more capacity, such as
fiber optic cable.
Coaxial cable consists of two cylindrical conductors, one placed concentric within the other,
separated by insulating material called the dielectric (see Figure 26). The outer conductor is
typically a copper wire mesh which provides EMI shielding and is covered with an abrasion
resistant jacket.
FIGURE 26.
COAXIAL CABLE
In general, there are two types of coaxial cabling known as thin coaxial and thick coaxial. Thin
coaxial cable or Thinnet is less expensive and easier to work with and install than thick coaxial
cable. Thinnet cable has a maximum segment length of 607 feet (185 meters).
40
A044E523 (Issue 1)
4-2013
5. Media Considerations
Thick coaxial cable or Thicknet has an extra protective plastic cover that helps keep moisture
away from the center conductor. Thicknet cable does not bend easily and is more difficult to
install than Thinnet cable. It has a maximum segment length of 1640 feet (500 meters).
5.4
Cat 1
1 Mbps
Telephone
Cat 2
4 Mbps
Telephone
Cat 3
16 Mbps
10Base T Ethernet
Cat 4
20 Mbps
Token Ring
Cat 5, 5e
5.4.1
1000 Mbps
Cat 6
10,000 Mbps
Gigabit Ethernet
Cat 7
10,000 Mbps
Gigabit Ethernet
5.5
5.5.1
Ethernet
Institute of Electrical and Electronics Engineers (IEEE) 802.3 defines the physical layer of wired
Ethernet. There are several subsections to this standard which define requirements for specific
standards. A listing of some of the more commonly used cable standards is shown in Table 2
below. Nomenclature of the cable standards defined by IEEE 802.3 are of the form
10BaseT. The number 10 in this example refers to the maximum bandwidth supported by the
A044E523 (Issue 1)
41
5. Media Considerations
4-2013
media, 10 Mbps in this example. The T refers to the type of cable, twisted pair in this
case. The word Base indicates that the transmission is baseband rather than
broadband. Baseband communication uses the complete bandwidth of the cable system, so
only one signal is ever found on the cable at any time. Baseband communication is most
commonly used in LANs and control networks. Broadband transmission, by contrast, is analog
rather than digital and supports multiple different signals using the same cable simultaneously,
each signal operating at different frequencies. Broadband communication is not used within
control networks and will not be discussed further in this document.
TABLE 2.
5.5.2
Standard
Bandwidth
Medium
Max Length
10Base5
10Mbps
500 m
10Base2
10Mbps
185 m
10Base-T
10Mbps
100 m
10Base-FL
10Mbps
Optical Fiber
2000 m
100Base-TX
100Mbps
100 m
100Base-FX
100Mbps
Optical Fiber
2000 m
100Base-T4
100Mbps
100 m
550 m
1000Base-LX
1Gps
Multi-Mode fiber
1000Base-LX
1Gps
5 km
1000Base-LX10
1Gps
10 km
1000Base-T
1Gps
100 m
42
A044E523 (Issue 1)
4-2013
5.5.3
5. Media Considerations
PCCNet
PCCNet is a proprietary Cummins Power Generation protocol which uses an RS-485 physical
layer. Cummins specifies Belden 9729 shielded twisted pair or equivalent cable for PCCNet
networks. CPG controls include a terminating resistor that matches the characteristic impedance
of the 9729 cable so PCCNet networks do not require an external terminator. Maximum total
network length is 4000 feet (1219 meters). As with all networks using shielded cable, it is
required to ground the shield at one point in the network. Connectors on CPG controls that are
used for PCCNet communication include a terminal for connecting the shield to ground.
5.5.4
LonWorks
CPGs LonWorks networks use stranded, unshielded, twisted pair cable. The following cables
are qualified for use with FTT-10 networks:
NEMA Level IV cable (Cummins P/N 334-1350 [PVC] or 334-1351 [Plenum])
EIA/TIA Category 5 (CAT5) (stranded only)
The maximum network length depends on which cable and topology are used. Table 3 lists the
maximum network length for each cable when a free topology is used. In this instance free
topology can be defined as any topology other than a bus. Table 4 lists maximum network
length when a bus topology is used. A LonWorks bus network may use stubs up to 10 feet (3
meters) long.
TABLE 3.
NEMA Level IV
400
500
EIA/TIA Cat 5
250
450
TABLE 4.
NEMA Level IV
1400
EIA/TIA Cat 5
900
CPG LonWorks network segments require termination for proper data transmission
performance. Free topology and bus topology networks differ in their termination requirements.
Free topology segments only require one terminator per segment. This terminator can be placed
anywhere in the segment but is recommended to be placed near the middle of the segment. All
CPG Lonworks cards (cards for 3100, 3201, and 2100 genset controls, OTPC, BTPC, OHPC,
and CHPC ATS controls), Control Communications Modules (CCM), Digital I/O Modules (DIM)
and Lonworks Annunciators have an on-board free topology terminator. It is recommended that
this be used on a free topology segment. These boards have a terminator switch. One switch on
the network device must be in the terminate position and all others left unterminated. The
ModLon gateway also has a terminator switch. The ModLon terminator is configured for a bus
topology.
To utilize the maximum bus length allowed by a bus topology network, the network must be
terminated at both ends of the bus. Cummins Power Generation offers a terminator for this
purpose (p/n 0300-5729).
A044E523 (Issue 1)
43
5. Media Considerations
5.5.5
4-2013
Wireless Media
Wireless communication has become prevalent in society creating expectations that industrial
control networks will have wireless capability. Wireless communication is not appropriate for real
time critical control functions on a generator set (such as excitation and fuel control), but the
growing expectation is that people will be able to monitor and to execute basic control functions
such as start, stop, or test from a wireless HMI or handheld device. Service technicians in all
kinds of industries are expected to communicate with control equipment using wireless
handheld devices. As of 2012, wireless communication is not commonly used internally in power
generation networks except to connect to a remote monitoring system.
The most prevalent forms of wireless communication are summarized below. Details of the
enabling technologies for wireless communication are beyond the scope of this document. With
the exception of cellular communication, Cummins Power Generation does not offer any of this
equipment as standard product.
5.5.5.1 Wi-Fi
IEEE 802.11 defines the media access and physical layers for wireless LANs. It is the dominant
standard for wireless internet access. Wi-Fi is the commonly used term for networks that meet
this standard.
Power generation networks can be connected to a wireless LAN using a wireless
adapter. There are devices available that will connect a wired Ethernet network (communicating
using Modbus/TCP) or a serial network (communicating using Modbus RTU over RS-485) to a
wireless LAN so that the network can be monitored over the internet.
5.5.5.2 Bluetooth
IEEE 802.15 defines the media access and physical layers for wireless personal area networks
(WPAN), defined as ad hoc networks transmitting up to 33 feet (10 meters). The technology
was originally developed by an organization called the Bluetooth Special Interest
Group. Bluetooth is the common term for this technology.
5.5.5.4 Cellular
Cellular technology has been used to connect power generation equipment with remote
monitoring stations for many years. Most major generator manufacturers offer a cellular
monitoring product. The monitor communicates with the genset control using either discrete
hardwired I/O signals or a serial link and sends data over a cellular network to a monitoring
center.
44
A044E523 (Issue 1)
4-2013
5. Media Considerations
There are two cellular communication technologies: Global System for Mobile communications
(GSM) and Code Division Multiple Access (CDMA). The appropriate technology to use depends
almost entirely on location. CDMA is the dominant technology in the US, GSM is dominant
everywhere else. There are some exceptions to this, as some U.S. carriers support GSM
networks. Work with the cellular service provider at the location of the power system to
determine the appropriate cellular technology.
Cummins Power Generations PowerCommand 500 and 550 Remote Monitoring Systems are
available with either CDMA or GSM capabilities. CDMA or GSM must be specified when
ordering the product.
5.5.5.5 Satellite
Satellite communication is often used when high cost equipment is in a mobile environment
such as rental or used in remote environments where cell phone service is not
available. Because satellites communicate over RF signals, the antenna must have line of
sight visibility to the sky. Satellite service providers charge substantially higher rates than
cellular providers but they offer true global coverage.
A044E523 (Issue 1)
45
5. Media Considerations
4-2013
46
A044E523 (Issue 1)
Protocols
6.1
Overview
This chapter defines and describes protocols using the Open Systems Interconnection (OSI)
model as a framework for the discussion. Protocols used by Cummins Power Generations
(CPG) products as well other commonly used protocols are defined in the context of the OSI
model and how and where they are typically used is discussed. For protocols that are not used
by standard CPG products, a description of how CPG products may interact with equipment
using that protocol is included.
6.2
Protocols
A protocol can be defined as a set of rules used mutually by two or more devices or software
applications to communicate. Protocols specify all of the characteristics of a communication
network from the physical and electrical qualities to the languages used to communicate.
6.2.1
A044E523 (Issue 1)
47
6. Protocols
6.3
4-2013
FIGURE 27.
6.3.1
48
A044E523 (Issue 1)
4-2013
6. Protocols
Physical Medium: Specifies what carries the signal, such as copper wire, fiber optic cable,
radio waves, etc.
Signal Characteristics: Describes how the signal works electrically. Defines what constitutes a
"1" or a "0", whether they are voltage levels, pulses, or rising or falling edges between pulses.
Bit Encoding: Defines how data bits get encoded for transmission. Specifies parameters such
as whether the most or least significant bit is transmitted first and whether the bit stream needs
to be modified to keep the transmitting and receiving devices in sync with each other. For
example if there were a long string of consecutive zeros, there would be no transitions of the
signal and the receiving device could lose track of how many bits have been received.
Bit Rate (Bandwidth): Defines how many bits per second the media will accommodate.
Specifies "the size of the pipe".
Distance Limit: Defines how far two devices can be from each other. This is the wire length for
wired protocols and the physical distance between devices for wireless protocols.
Duplex: Defines whether the nodes can both transmit and receive simultaneously (full duplex)
or if the can only do one at a time (half duplex).
FIGURE 28.
Physical Medium: Medium is twisted pair copper wire or fiber optic cable. For wire, Cat 5e
twisted pair copper wire would be used to support 100 Mbits/sec data rate. The 5e standard
specifies the electrical characteristics of the wire and its insulation such that the signal integrity
will be maintained.
Signal Characteristics: Different versions of Ethernet use different voltage levels, ranging from
-2.5V to 2.5V.
Bit Encoding: Ethernet uses a 4B5B bit encoding scheme. This maps every 4 bits into a group
of 5 bits for transmission. This 4-bit to 5-bit mapping is pre-determined so that there are always
transitions in every group of 5 bits.
Bit rates: Common Ethernet bit rates include 10 MB/second for (10 BaseT networks) and 100
MB/sec (for 100 BaseT networks). There are also 1 GB and 10 GB Ethernet standards.
A044E523 (Issue 1)
49
6. Protocols
4-2013
Distance Limit: For copper, 328 feet (100 meters) is the limit due to signal degradation over
that distance. The beauty of Ethernet though is that this distance can be extended almost
indefinitely with switches and routers. These are intelligent devices that clean up the signal and
forward the message on to its destination.
Duplex: Ethernet originally ran half duplex and used hubs to extend the network. The network
could only be extended so far before the collision domain limit would be exceeded. Todays
Ethernet is full duplex and uses switches to extend the network. There is no collision domain
issue since the switches manage the full duplex traffic flow, so network size/distance is
practically unlimited.
Other Ethernet Media Examples
Fiber Optic Cable: There are several Ethernet standards for fiber optic cable. One of the most
common is Ethernet 100 Base-FX which is capable of internode distances of over 1 mile.
Radio Frequency: IEEE 802.11g is the most common wireless (Wi-Fi) standard. IEEE 802.11g
networks can run at 54 Mbits/sec covering a radius of at least 65 feet (20 meters).
6.3.2
FIGURE 29.
50
A044E523 (Issue 1)
4-2013
6. Protocols
FIGURE 30.
A typical frame will consist of ID (identification) addresses of the sending and receiving devices,
control information, data, and data integrity bits. The control information is used to signal the
beginning of a new frame, describe how the message is segmented (how this frame fits into a
larger message), and how it is to be routed. Data integrity bits are used to verify that the
message has been received correctly. In this example the data integrity scheme is known as a
Cyclic Redundancy Check or CRC.
This does not necessarily cover the case of intentional (hacking) corruption,
only signal corruption due to electrical interference, noise, etc.
Application layers higher in the stack are concerned with making sure there has not been
intentional corruption. For example, Ethernet networks use a 32 bit data integrity scheme called
a Cyclic Redundancy Check (CRC). This is an algorithm that calculates a number based on all
the other bits in the frame. This number gets calculated by the sender and sent along with the
frame. The receiver then also calculates the CRC and compares its result with the CRC
received from the sender. If they match, the frame is good. If not, it must be thrown away and
resent.
A044E523 (Issue 1)
51
6. Protocols
4-2013
VLANs allows for creating separate networks that share a medium. The networks do not
know the others are there. It is a method to manage and isolate traffic. Note this is different
than VPN (Virtual Private Network).
FIGURE 31.
6.3.3
MAC ADDRESS
52
A044E523 (Issue 1)
4-2013
6. Protocols
In simple control networks using protocols such as Modbus, LonWorks, or PCCNet, the logical
address can be as simple as a one digit number. A device is assigned a logical address when it
is commissioned and that address typically does not change throughout the life of the network.
If a device fails and needs to be replaced, the new device will typically be assigned the same
logical address as the device it replaced. All the devices on the network will use logical
addresses to communicate with each other.
In more extensive networks, such as LANs (Local Area Networks) or WANs (Wide Area
Networks), devices will have more extensive logical addresses that will change in some
cases. For example, Internet Protocol (IP) defines a 32 bit logical address for devices known as
the IP address. In WANs that span multiple sub-networks and routers, the routers will use the
information encoded in the IP address to identify a path or route between the sending and
receiving devices. (See below for discussion of IP address and subnet mask.)
There are several protocols that routers use to direct messages to their destination. Two of the
most common are Open Shortest Path First (OSPF), which is typically used in large enterprise
networks, and Border Gateway Protocol (BGP), which is typically used on routers that run the
internet.
Since routers work with logical addresses and operate on the Network Layer of the OSI model,
they are often known as layer 3 devices.
FIGURE 32.
FIGURE 33.
A044E523 (Issue 1)
IP PARAMETERS
53
6. Protocols
4-2013
Under the current, most commonly used IP standard (IPv4), IP addresses consist of 32 bits, or
binary digits. The IP address is typically expressed as the decimal values of 4 bytes, each byte
representing 8 bits. For example, an IP address of 192.168.0.4 is actually the decimal
representation of the following:
11000000.10101000.00000000.00000100
(
192
168
An IP address is divided into two portions, network and host. The network portion of an IP
address represents the network to which the device belongs. Devices can only communicate
directly with devices that are on the same network. Devices communicate with other networks
through routers.
The host portion of the IP address represents the device, or interface of the device. A device
could have multiple interfaces, each with their own IP address. The host portion of the IP
address must be unique on the network, or IP subnet.
The Subnet Mask divides the network and host portion of an IP address. Like the IP address,
the subnet mask is a 32 bit number. For every bit of the subnet mask that is a 1, the
corresponding bit of the IP address is part of the network portion of the IP address. Bits of the
IP address corresponding to a 0 in the subnet mask are part of the host portion of the IP
address.
For example, consider a subnet mask of 255.255.255.0. This would be written in binary as
11111111.11111111.11111111.00000000. This means that the first 3 bytes (24 bits) of the IP
address would be the network portion of the IP address and the 4th byte (last 8 bits) would be
the host portion. See the example below.
Subnet Mask = 255.255.255.0
IP Address = 192.168.0.4
Network portion of IP address = 192.168.0.
Host portion of the IP address = 4.
Devices that have identical network portions are on the same subnet and can communicate with
each other directly. Devices that have different network portions are on different subnets and
need to communicate with each other through routers.
The third parameter that needs to be specified is known as the default gateway address. This
refers to the IP address of the router located on that network. The network portion of the default
gateway address will be the same as all of the other devices on that network. Devices on the
network will need to go through this router to communicate with devices on other
networks. Routers use the IP address, subnet mask, and default gateway address to define an
efficient route between devices.
The term gateway is used here to refer to the connection point between two networks. In this
case that connection point is a router. In general, it is not appropriate to use the terms gateway
and router interchangeably.
54
A044E523 (Issue 1)
4-2013
6. Protocols
6.3.4
A044E523 (Issue 1)
55
6. Protocols
4-2013
FIGURE 34.
56
A044E523 (Issue 1)
4-2013
6. Protocols
6.3.5
6.4
A044E523 (Issue 1)
57
6. Protocols
4-2013
For instance, consider sending a start command to a generator set from a remote HMI
communicating by Modbus/TCP. The Modbus command (write the start bit to the appropriate
Modbus register on the generator set control) would be formatted with the appropriate security
data session layer. The transport layer on the HMI will add error checking info and will divide the
message into the appropriate segments. The Network Layer will add the IP address of the
sending and receiving devices and routing information. The data link layer will add
synchronization and frame integrity information and start and stop bits, and the physical layer
will send the frame out over the communication media. At the receiving device, the process
works in reverse as the layers decode the message while passing it up the stack until the start
bit gets set in the Modbus register of the generator set control. Figure 35 is a sample of the
frame construction.
FIGURE 35.
6.5
FIGURE 36.
58
TCP/IP MODEL
A044E523 (Issue 1)
4-2013
6. Protocols
Figure 37 shows how the two models map to one another. Note that the mapping between the
models is not perfect. For example, there are elements of the OSI Session layer that are defined
in the Transport layer of the TCP/IP model. The differences between the two models are subtle
and are outside of the scope of this document.
FIGURE 37.
6.6
Common Protocols
This section consists of a list of protocols that are commonly used in control networks, with a
brief discussion of each. The section is divided into hardware protocols (defining the Data Link
and Physical Layer attributes) and software protocols (which define the higher layer attributes,
or in some cases defining all of the layers.)
6.7
6.7.1
Ethernet
Ethernet is the most commonly used data communications standard on Local Area Networks
(LAN). Ethernet is defined by the IEEE 802 family of standards which define Physical and Data
Link layer properties for LANs and Metropolitan Area Networks (MANs). IEEE 802 was originally
developed to define communication over twisted pair cable but has since been expanded to
include communication over fiber optic cable and over wireless media.
Some of the more broadly applied sections of IEEE 802 are listed below.
IEEE 802.3 - Defines the media access and physical layers for 10 MB, 100 MB and 1 GB
Ethernet connections over twisted pair cable, coaxial cable, and fiber optic cable.
IEEE 802.11 - is the standard for wireless LANs (WLAN). There are multiple subsections of
802.11 which cover networks of different speeds and different frequency bands. Wi-Fi
networks are designed to IEEE 802.11
IEEE 802.15 - Defines the media access and physical layers for Wireless Personal Area
Networks (WPANs). WPANs convey information over short distances among a private
group of devices. This is the Bluetooth standard. The technology was originally developed
by the Bluetooth Special Interest Group and the standard was written around the
technology. Unlike LANs, a WPAN has little or no connection to the world outside of the
private network.
A044E523 (Issue 1)
59
6. Protocols
6.7.2
4-2013
RS-232
RS-232 was developed by the Electronic Industries Association (EIA) to define the physical
interface between a PC and peripheral equipment, such as modems and printers. The RS-232
standard defines communication between two specific types of equipment known as Data
Terminal Equipment (DTE, typically a PC) and Data Communications Equipment (DCE, typically
a modem) over a distance of no more than 50 feet (15 meters). RS-232 also specified the pin
designations on the 9 and 25 pin connectors that were most common at the time. RS-232
defines separate transmit and receive wiring paths with the transmit pin on one device
physically connected to the receive pin on the other device. Although it was designed for a very
narrow scope of applications (communication between only two devices no more than 50 feet
(50 meters) apart), RS-232 is still commonly used in those applications today due to its
simplicity and relatively low cost to develop and deploy. Cummins Power Generations Mon
protocol, which is used by the InPower service tool to connect to controls, is built on the RS-232
physical layer.
6.7.3
RS-485
RS-485 was also developed by EIA as the standard for connecting more than two devices on a
network over longer distances. Unlike RS-232, which has separate circuits for transmitting and
receiving signals, RS-485 specifies a single circuit for both transmitting and receiving signals,
known as a transceiver. The RS-485 transceiver is designed in such a way that multiple
transceivers can be connected to the network without affecting the one transceiver that is
transmitting a signal. RS-485 was designed so that signals can travel long distances without
being attenuated by resistance and inductance in the wires and have a high level of noise
immunity.
RS-232 allowed two devices to send messages at the same time by having separate send and
receive paths between the two devices. Since RS-485 uses only a single transceiver for both
transmitting and receiving data, this is not possible. At any given time, a device may either
transmit or receive data but the device cannot perform both operations. This is known as halfduplex communication. There is also a 4-wire implementation available on the market which
supports full-duplex communication. These devices consist of 2 separate RS-485 transceivers,
so that one can transmit and the other can receive at the same time.
Cummins Power Generations PCCNet protocol, used to communicate between a generator set
and its accessories, such as human machine interfaces (HMIs) and annunciators, is built on the
RS-485 physical layer. RS-485 networks can communicate with up to 32 devices at half duplex
at distances up to 4000 feet and can be expanded by using repeater products available on the
market.
6.7.4
6.7.5
60
A044E523 (Issue 1)
4-2013
6.7.6
6. Protocols
6.7.7
6.7.8
Modbus
Modbus protocol was published by Modicon in 1979 for use with their PLCs and has become a
de facto standard industrial communication protocol. Modbus uses a master/slave format and
was originally developed to communicate over a serial link, typically RS-485. There were
originally two forms of Modbus; Modbus RTU and Modbus ASCII, with the difference being in
how the data bits were packed into a frame. Modbus RTU packs the bits more efficiently
(meaning the overall message length is shorter to convey the same amount of information so
the network runs faster) so Modbus ASCII is rarely used today.
Modbus TCP is becoming more popular as TCP/IP and Ethernet protocols are becoming more
common in control networks. At the application layer, Modbus TCP is the same as Modbus
RTU. The difference is in how the data is transported at the lower layers.
Another version of Modbus, Modbus+, is available on the market. It uses a peer-to-peer token
passing communication scheme (as opposed to master/slave) and is capable of running at 1
MBit/s. Modbus+ requires a dedicated co-processor and is proprietary to Schneider Electric so
its use is limited to applications where high speed communication is critical. Cummins Power
Generation uses Modbus+ for internal communication in some Digital Master Control (DMC)
applications.
6.7.9
LonWorks
The LonWorks platform was developed by Echelon Corporation and is used primarily in building
automation systems, particularly in HVAC and lighting systems All seven layers of the protocol
stack are implemented using Echelons Neuron chip. Although technically an open protocol, it is
very difficult to develop LonWorks products without buying the Neuron chip. Installing LonWorks
networks typically requires an integration tool. Echelon's LonMaker commissioning tool is most
commonly used for this purpose. Use of LonMaker requires paying a royalty per device
commissioned.
A044E523 (Issue 1)
61
6. Protocols
4-2013
LonWorks standard network variable types (SNVTs) and peer to peer communication gives
integrators a great deal of flexibility in developing a customized network, particularly when
working with devices from different manufacturers. Manufacturers are also able to create their
own User-defined Network Variable Types (UNVTs) giving them the ability to optimize their
application of LonWorks for their equipment.
Cummins Power Generation controls have used LonWorks communication since the mid-1990s,
however, in recent years the preference in Cummins Power Generations key critical protection
markets has been for Modbus communication at the application layer and RS-485 and Ethernet
at the lower layers. Cummins Power Generation controls are moving towards supporting
Modbus for 3rd party communication and a proprietary protocol for communicating between
Cummins products.
6.7.10 BACnet
BACnet is a data communication protocol used mainly in HVAC and lighting control networks.
BACnet was developed with the guidance of the American Society of Heating and Air
Conditioning Engineers (ASHRAE). BACnet is an application layer protocol and interfaces with
several data link and physical layer protocols, including RS-232, RS-485, and Ethernet. BACnet
applications have also run with LonWorks data link and physical layers.
It is not uncommon to have LonWorks and BACnet devices on the same network and there are
many Lon to BACnet converters commercially available. In most cases, individual devices on a
network do not need to communicate with each other, but need to communicate with a common
building management computer running a commercially standard SCADA package. Most
SCADA packages will have standard drivers for Modbus and many will have drivers for
LonWorks as well. Most often it is not necessary to convert LonWorks or Modbus devices to
BACnet for all devices to be monitored and controlled by a single Building Management System.
6.7.11 Profibus
Profibus was originally developed in Germany in the late 1980s and was initially used and
promoted by Siemens. It was developed for use in factory and process automation systems with
the intent of reducing the wiring required to communicate between a central controller and
distributed sensors and actuators. Profibus protocol defines all seven layers of the OSI
model. There are several different physical layer implementations including RS-485. Profibus is
a proprietary closed protocol and is not openly published.
62
A044E523 (Issue 1)
4-2013
6. Protocols
6.7.14 PCCNet
PCCNet is a Cummins Power Generations proprietary protocol used exclusively for
communicating between a generator set or switchgear controller and its peripheral devices,
such as HMIs, annunciators, and expanded I/O modules. It is not used to communicate with any
3rd party equipment or monitoring system. Limiting the types of devices that can communicate
over PCCNet makes it a simple network to use and to support. PCCNet is a plug and play
network, with no configuration required for devices to communicate. The plug and play
characteristic simplifies service as well because no special tools are required to replace
components
PCCNet communicates over an RS-485 physical layer.
6.8
Protocol Conversions
In network applications it is common to have a requirement for equipment from several different
manufacturers to communicate with each other on the same network. In many cases this will
drive a requirement to convert between protocols. Conversions may be required over all 7
layers of the OSI model, or over only the hardware layers (layers 1 and 2), or the software
layers.
A044E523 (Issue 1)
63
6. Protocols
6.8.1
4-2013
Ethernet Conversions
As the various Ethernet standards for twisted pair cable, fiber optic cable, and wireless
communication have become the dominant networking infrastructure standard in all kinds of
networks, conversions between the serial communication protocols commonly used in
equipment, such as Modbus RTU and LonWorks, are quite common. There are many
commercially available gateways that are easy to configure for this purpose, particularly with
devices that communicate over RS-232 or RS-485 serial links. These gateways, often known as
device servers, have been on the market as long as LANs have been common in office
environments, as there was a need for devices such as printers to be served from any PC on
the LAN rather than from a dedicated server PC.
Commercially available Modbus serial to Ethernet gateways can be used to communicate over
Ethernet with a CPG control that has Modbus communication. Consult the CPG factory for
assistance in configuring these networks.
A LonWorks to Ethernet converter is available from Echelon for communicating LonWorks
network variables over Ethernet media. Consult with the factory for support in configuring the
LonWorks to Ethernet router.
Networks communicating over TCP/IP require a time server somewhere on the network. A time
server is a program that synchronizes the clocks of all the devices on the network. Most TCP/IP
networks have a PC running a time server program. Echelons LonWorks to Ethernet router
ships with a time server program that can be installed on a PC.
6.8.2
64
A044E523 (Issue 1)
Monitoring Systems
7.1
Overview
This chapter describes monitoring systems, and defines purposes and basic functions. On-site
supervisory control and data acquisition (SCADA) systems and web-based monitoring systems
are explored, and typical applications for each type of monitoring system are described.
Considerations for communicating between Cummins Power Generations equipment and 3rd
party monitoring systems will be explained.
7.2
Purpose of Monitoring
There are many reasons cited for monitoring power generation systems. When implemented
correctly, the benefits of monitoring can be broken down into three areas: Improving system
reliability, reducing operating costs, and meeting agency requirements.
7.2.1
System Reliability
Monitoring can improve system reliability by notifying appropriate personnel that the system has
some problem or issue so that it can be rectified before a total failure occurs. For example, a
monitoring system can alert operators or maintenance people that a standby system may not
come on line in the event of a utility failure or that a generator set is in need of maintenance to
continue reliable operation.
There are four root causes that account for almost all of the failures of standby generator sets to
come on line in the event of a utility failure. They are low/dead battery, low fuel, not-in-auto
(service technician moves the Run/off/auto switch to Off to do some maintenance and forgets
to return the switch to the Auto position), and the engine is too cold because a block heater
has failed. A simple monitoring system that monitors and reports on just these parameters
(battery voltage, fuel level, Run/Off/Auto switch position and coolant temperature) will greatly
improve reliability.
Planned maintenance and prognostics are important factors in the reliability of a system,
particularly in prime power or load curtailment applications, and a monitoring system that alerts
service technicians that maintenance is required will contribute to reliability. Maintenance
notifications can be simple, such as the engine approaching some number of run hours, or can
be more complex such as load level, operating temperatures, or fuel consumption indicating that
the engine is not performing as efficiently as it should and further analysis is required. If the
system is operating due to a normal power failure, a monitoring system allows remote control of
the system to facilitate load management and other functions without actually being in the
generator room.
7.2.2
Operating Cost
A monitoring system can minimize operating cost of a generator set by making service and
maintenance more efficient and giving indication of a system that is not operating efficiently so
that corrective action can be taken.
A044E523 (Issue 1)
65
7. Monitoring Systems
4-2013
Some service organizations bundle a remote monitoring service with a planned maintenance
agreement because the monitoring system allows the service organization to operate more
efficiently. For example, a low fuel notification may be sent directly to a fuel delivery contractor
who can re-fill the fuel tank without any prior interaction with the service organization or end
user.
A monitoring system will allow service organizations to have better visibility of when
maintenance is required. When there has been an alarm, the monitoring system may allow the
service organization to do some diagnostics before dispatching a technician to the site so that
technicians will have the parts they will likely need and will more likely be able to repair any
problems with the generator set on the first trip to the site.
An end user may also be able to use a monitoring system to minimize the operating costs of the
system. Having access to fuel consumption and energy usage data can help make decisions on
questions, such as when to run the facility on generator power versus utility power and whether
there is value in operating equipment at times of day when energy costs are lower.
7.2.3
Agency Requirements
Regulatory agencies such as the Environmental Protection Agency (EPA) and the Joint
Commission on Accreditation of Healthcare Organizations (JCAHO) often require record
keeping and reporting on the standby power system. JCAHO, for example, requires a report that
documents key engine and generator parameters during all mandated test runs. Monitoring
programs are often used to record the necessary data and generate the reports automatically.
In the USA emergency power systems must have remote annunciation of alarm and status
conditions in a manned location. Conditions to be monitored include whether the system is
running on normal or emergency power, whether the system is in automatic mode, and whether
engine parameters such as coolant temperature or oil pressure have gone outside of a normal
operating range.
7.3
7.3.1
Status Display
The most immediately visible function of the monitoring system is the status display. This is
simply a snapshot of the current status of the equipment and can be as simple as a couple of
lights on an annunciator panel, or it can be a very sophisticated animated touchscreen Human
Machine Interface (HMI). Data displayed can be either discrete or analog. Discrete data is less
expensive to gather and simpler to communicate and in most cases is sufficient for
emergency/standby power systems. The most commonly used display for emergency/standby
power systems in the United States (US) is the annunciation required by the National Fire
Protection Agency Standard for Emergency and Standby Power Systems (NFPA 110). NFPA
110 requires indication of several discrete data points which describe the status of an
emergency power system. Figure 38 is an example of an annunciator with light-emitting diodes
(LED) dedicated to each of the NFPA 110 required events.
66
A044E523 (Issue 1)
4-2013
7. Monitoring Systems
FIGURE 38.
ANNUNCIATOR
In some critical protection or prime power applications, end users require real time display of
analog parameters so that they may be able to make adjustments in real time to optimize the
reliability and efficiency of the system. Figure 39 is an example of a more capable and complex
HMI displaying the status of a system of paralleled generator sets.
FIGURE 39.
In the example displayed in Figure 39 an operator can get more detailed analog data of
individual generator sets by clicking on one of the generator sets on the HMI. Figure 40
displays the screen that appears from clicking on one of the generator sets on the HMI
displayed in Figure 39.
A044E523 (Issue 1)
67
7. Monitoring Systems
4-2013
FIGURE 40.
In this example a replication of an annunciator is on the left side of the screen. (Note that this
can be displayed regardless of whether there actually is an annunciator in the system.) The
center section of the HMI displays key engine and generator parameters. The HMI in this
example includes the basic control functions of starting and stopping the generator set and
manually opening and closing the paralleling breaker. The operator can execute these functions
by clicking on the appropriate button on the right side of the HMI. Notice that the genset and
paralleling breaker are displayed at the top of the HMI. The breaker is shown as closed in the
single line diagram fragment and the generator and breaker are red indicating that the bus is
live. The color green indicates a dead bus.
Some operators may prefer to see a meter or bar graph display of generator and engine
parameters rather than a digital display as shown in Figure 40. This allows them to observe the
status of the analog parameters at a glance. Figure 41 shows another HMI screen which
mimics analog meters.
FIGURE 41.
68
A044E523 (Issue 1)
4-2013
7.3.2
7. Monitoring Systems
Alarm Notification
In standby power systems, arguably the most important function of a monitoring system is alarm
notification. As discussed earlier in this section, the ability of a monitoring system to notify a
service technician that the system has a fault, is at risk of having a fault, or is not ready to run
for some reason, will contribute substantially to system reliability.
Alarms can be generated either on the discrete status of data points or on the event of analog
parameters being outside of a given range. Service technicians are typically able to configure
which events are delivered as alarms.
Many monitoring systems will require some acknowledgement in their alarm delivery system to
better ensure that the alarm condition is addressed. With some systems, if the alarm is not
acknowledged (by the recipient responding to the alarm SMS/text message or email) within
some period of time, the system will repeat the notification or will deliver the alarm notification to
someone else and repeat this until the alarm is properly acknowledged.
Some monitoring systems will allow certain alarms to be delivered only to certain recipients, a
process known as selective messaging. The most common example of this is a monitoring
system that sends low fuel alarms to a fuel delivery subcontractor. The subcontractor has no
use for receiving any other alarm notification.
The mechanism for delivering alarms is typically via email, as web-based monitoring systems
usually have email capability and service technicians usually carry mobile phones capable of
receiving emails.
7.3.3
Data Logging
In many applications, end users and service organizations will have a need to log data on power
system events and performance over time. As stated previously, the purpose of this is to make
the system more reliable, more efficient, and/or meet legal reporting requirements.
A simple event log listing events, such as generator sets starting and stopping and various
breakers opening and closing, can help diagnose faults in a system. Figure 42 displays an
event log for a paralleling system, listing all of the events associated with generator sets
synchronizing and closing to a bus. If there is a fault in the system, an event log such as this will
be very useful in diagnosing the root cause of the fault.
A044E523 (Issue 1)
69
7. Monitoring Systems
4-2013
FIGURE 42.
An alarm log allows service staff to see the status of system alarms. Figure 43 is an example of
an alarm log. Notice the key in the lower left corner, referring to active versus inactive alarms
and acknowledged versus unacknowledged alarms. An active alarm is one in which the system
is still in the state which caused the alarm. An inactive is one in which the system has returned
to the normal state. For example, a voltage sag that creates an undervoltage alarm will be
considered active as long as the voltage is below the threshold. The alarm will be considered
inactive if the voltage returns to the normal range.
Many monitoring systems require alarms to be acknowledged and will log who acknowledged
the alarm and when as part of the service record.
FIGURE 43.
70
ALARM LOG
A044E523 (Issue 1)
4-2013
7. Monitoring Systems
Logging data over time allows organizations to mine data in an attempt to better understand
how the system works, to optimize system operation. Knowing how much energy is used, at
what time, and by which loads, can help facility owners operate their facilities more efficiently.
Figure 44 displays a plot of a log of real power used in a facility over a period of 10 days. This
particular plot does not differentiate between power supplied by the utility versus power supplied
by a generator set. It enables a facility owner to better understand the energy demands of the
facility and make decisions accordingly.
FIGURE 44.
Prime power systems will often use data logging systems to analyze trends in system
performance. Changes in various temperatures in an engine can signify wear in components
indicating that maintenance is due.
Data logging systems that log analog data can quickly consume all available memory resources.
There are a few ways that this can be managed.
Logging frequency can be adjusted. For example, a standby power system data may be logged
once per hour when the generator sets are not running and once every five minutes when the
generator sets are running. When an alarm becomes active, the system could also be
configured to take a complete snapshot of all the monitor points and add it to the log to assist
in diagnosis.
If logging data at 5 minute intervals is not sufficient for data mining purposes, a circular buffer
scheme can be used. With a circular buffer, there are a fixed number of data sets that can be
stored. When this number has been reached, the next data set logged simply overwrites the
oldest set. This process continues indefinitely with the newest data set always replacing the
oldest data set. This scheme is good for diagnosing system faults as data sets taken once per
second leading up to a fault will be helpful in analyzing what caused the fault; however, without
intervention, data that is more than a few hours old will have been deleted. Monitoring systems
may use both a circular buffer logging data once per second (for example) to be used for fault
analysis and a permanent log which consists of data sets taken at 5 minute intervals to analyze
trends in system performance.
A044E523 (Issue 1)
71
7. Monitoring Systems
4-2013
Even data taken at 5 minute intervals can eventually exhaust the available memory of an on-site
monitoring system, so provisions have to be made to move this data to a permanent archive.
This is typically accomplished using a store and forward scheme in which the system
periodically sends data to some offsite archive. Data archiving routines can be implemented
automatically using a pre-set schedule or can be initiated by an operator.
7.3.4
Reporting
Agencies, such as JCAHO and the EPA, require reports documenting system
parameters. Almost all monitoring systems with data logging capability have at least a basic
reporting capability.
Monitoring systems generate reports simply by exporting logged data into a format that can be
used by an end user or technicians personal computer (PC). Acrobat and Excel are the most
common formats used for reporting, although in some cases, to avoid compatibility issues,
monitoring systems will export data in a very simple Comma Separated Value (.csv) format. A
.csv file is simply a text file with fields separated by commas which can be imported into many
spreadsheet or word processor programs using many different operating systems.
7.4
72
A044E523 (Issue 1)
4-2013
7.4.1
7. Monitoring Systems
SCADA Systems
SCADA systems, also known as Building Management Systems (BMS), offer maximum
flexibility in a monitoring system. SCADA systems are typically custom designed and in many
cases monitor the entire facility rather than just the power generation and distribution
equipment. SCADA systems can also be expensive, with the main cost drivers being the
number of custom screens and the number of data points the system needs to monitor.
Figure 45 displays an example of a custom HMI screen developed for a SCADA system
installed at a major International Airport. The screen displays a floor plan of one of the airport
terminals. Each grey box represents a transfer switch which an operator can click on to view the
current status of that switch.
FIGURE 45.
There are three main building blocks at the core of a SCADA system: Data acquisition
software, communication drivers, and a process server.
A044E523 (Issue 1)
73
7. Monitoring Systems
4-2013
7.4.1.2 Drivers
A driver, in the context of a monitoring system, is a software program that allows the data
acquisition application to communicate with the monitored equipment. It converts data between
the format used by the data acquisition software and the protocol used by equipment. Drivers
are typically supplied by the data acquisition software supplier and several different drivers will
be available in support of the program. For example, a data acquisition software product may
have drivers available for Modbus, LonWorks, and BACnet, and it is permissible to use more
than one driver in a particular application. A monitoring system may simultaneously use BACnet
to communicate with heating and air conditioning equipment and Modbus to communicate with
power generation and distribution equipment. It is not necessary that all monitored components
use the same protocol as long as the monitoring system has drivers to support all of the
protocols used in the system.
All of Cummins Power Generations generator sets are capable of communicating over Modbus
to enable communication with SCADA systems. Modbus is the most common application layer
protocol used in SCADA systems and most SCADA packages have a Modbus driver as a
standard offering. When system specifications call out communication with a SCADA system
using some other protocol, the first course should always be to inquire whether a Modbus driver
is available for the SCADA system. In most cases, a Modbus driver will be available at a low
cost. Adding the Modbus driver will be a much simpler, less expensive and more reliable
solution than performing a protocol conversion, which will require a 3rd party gateway and
custom programming.
A044E523 (Issue 1)
4-2013
7. Monitoring Systems
Remote access is most commonly provided through the internet and the Pulse system can be
provided with a web server as an option. End users and service technicians will be able to view
system status and download reports and data logs remotely through the web as they would for a
web-based monitoring system. Although this scenario does provide the benefits of a web-based
system, for purposes of this discussion, the Pulse system is still considered an on-site rather
than web-based monitoring system because typically the primary day-to-day users of the
system are on-site and all of the data manipulation happens on-site.
7.5
FIGURE 46.
A web-based monitoring system (or an on-site monitoring system with web access) has some
distinct advantages over a traditional PC based (thick client) monitoring system. First, end users
do not need to load any special software on their PC other than a web browser. All functions are
executed by the server. Second, users can view system status any place that has internet
access. Finally, users can receive alarm notifications via any medium where they can receive an
email or SMS/text message.
A044E523 (Issue 1)
75
7. Monitoring Systems
4-2013
FIGURE 47.
When any piece of equipment is exposed on a public network there is a risk of hacking. By
taking appropriate IT security measures, as discussed previously, that risk can be minimized.
People have expressed concern that someone could hack into controls and cause permanent
damage to the generator set. That likelihood is remote, as generator controls are designed so
that only basic commands and adjustments (such as start and stop) are available over the
network. Although the risk is minimal it should be taken seriously. It is appropriate to have
protective devices on site to take a generator off line and isloate equipment in the event that
something does malfunction. It is prudent for facility managers to analyze all possible
contingencies using a tool such as a Failure Mode and Effects Analysis (FMEA).
76
A044E523 (Issue 1)
4-2013
7. Monitoring Systems
While there are advantages to using a corporate network for external communication, it is
important to isolate the automation network from the corporate network for many reasons. One
key reason is to make the amount of traffic on the automation network be unaffected by traffic
on the corporate network. This is important since the automation network depends on the
network for controlling devices in real time. See Figure 48.
FIGURE 48.
7.5.1
FIGURE 49.
A044E523 (Issue 1)
77
7. Monitoring Systems
4-2013
In applications in which there is no internet access, an off-site web server is the only choice for
web-based monitoring. In addition, an off-site web server has advantages when the monitoring
system is in use by a service organization to service multiple installations. Data from multiple
installations can be sent to the same web server and displayed on a single web page so that
service technicians only need to access a single web page to view status on all of the
installations for which they are responsible. Figure 50 represents a web page listing generator
sets from multiple installations.
FIGURE 50.
An off-site web server will incur ongoing costs of operation. An off-site web server will typically
be hosted by a 3rd party who is contracting space at a data center. This will result in an ongoing
cost for the service. There will also be an ongoing cost associated with the cellular or satellite
connection between the installation and the server as well as the need to establish and maintain
that connection. A service organization using this system may bundle the remote monitoring
service with a Planned Maintenance Agreement (PMA) and pass these costs to the customer as
part of the PMA.
7.6
78
A044E523 (Issue 1)
8.1
Overview
This section descibes several different applications using network technologies. In general,
these were developed to contribute to maximizing system reliability and minimizing operating
costs. Networks, when combined with appropriate service and maintenance procedures, provide
the infrastructure to achieve these goals through the key functions of status display and control,
alarm notification, and data logging, analysis, and reporting. In this chapter we will look at a few
examples of networks that serve the customer well in this respect. Some of these networks are
very simple and some seem quite complex, but they all increase reliability and reduce operating
costs by executing the same functions. Even the complicated networks can be understood by
breaking them down into the simple concepts discussed in this manual.
8.2
A044E523 (Issue 1)
79
4-2013
The grocery store networks are a very simple yet very effective web-based monitoring system
achieved by understanding exactly the end users needs and limitations, and designing a
system to fit the application.
8.3
FIGURE 51.
8.4
80
A044E523 (Issue 1)
4-2013
The SCADA system communicates over Modbus. Each of the generator sets has a LonWorks
card. A ModLon gateway converts the Lon data from the gensets to Modbus so that the SCADA
system can use it. The fourth generator set has a LonWorks to Ethernet router located at the
generator set so that the Lon data can be communicated over the internet. A second LonWorks
to Ethernet router puts the data back on to the LonWorks physical layer so that it can
communicate through the ModLon gateway to the SCADA system. The network also has an
iWatch web server so that alarms can be delivered and system status can be viewed
remotely. Figure 52 is a block diagram of the network.
FIGURE 52.
The utility uses the generators for peak shaving and the monitoring system allows them to track
usage and fuel consumption. The data is used to analyze when it makes financial sense to peak
shave.
8.5
A044E523 (Issue 1)
81
4-2013
Both networks use a similar architecture at a high level. The generator sets are all connected to
a LonWorks network. Data from the Lon network is converted to Modbus by the ModLon
gateway located within the DMC cabinet where it is read by a Programmable Logic Controller
(PLC) which is part of the DMC. The PLC has an Ethernet connection which is connected to an
Ethernet switch to which all of the transfer switches are connected through fiber optic cable. The
airport system also has a Pulse system connected to the Ethernet switch. Figure 53 is a
representation of how the gensets, the DMC, and the Ethernet switch are connected.
FIGURE 53.
Due to the size of the facilities and the great distances separating the transfer switches and the
centrally located DMC, fiber optic cable is a good medium to use for communicating with the
transfer switches. The two networks use slightly different means to connect the transfer
switches to the fiber optic network.
The operators of the airport defined a piece of equipment called the Remote Transfer Control
Panel (RTCP). The RTCP connects the ATS to an Ethernet switch using Cat 5 cable. MAC
treats the RTCP as a black box leaving it up to Cummins to determine how to get ATS data to
the Ethernet cable. Not all of the ATSs in the network were supplied by Cummins, however
Cummins was responsible for making all the ATSs look the same on the network. This was
achieved by using the LonWorks network card with the Cummins switches and using Cummins
Control Communication Module (CCM) with the non-Cummins ATSs. (The CCM is designed to
connect other generator sets and transfer switches to a Cummins Lon network. All of the
generator set or ATS control and sensing signals are connected to the CCM and communicated
over the LonWorks network using the same network variables that the Cummins controls use.)
Within the RTCP, the LonWorks data is converted to Modbus over RS-232 using the ModLon
gateway. The Modbus over RS-232 is converted to Modbus IP using a commercially available
serial to Ethernet converter (also known as a device server).
82
A044E523 (Issue 1)
4-2013
Several of the RTCPs that are located in the same general area are connected over Cat 5 cable
to an Ethernet switch which will connect to the rest of the facility using fiber optic cable. The
Ethernet switch is also connected to a PLC which manages some discrete I/O. The start
command from the ATS, for example, doesnt go through the RTCP but goes directly to the PLC
as a discrete input. Figure 54 is a representation of the RTCP and the Ethernet switch.
FIGURE 54.
The convention center network uses a slightly different method to connect the transfer switches
to the fiber optic network. This network uses a gateway that can convert LonWorks directly to
Modbus TCP. Each of these gateways can connect to up to 8 transfer switches over Lonworks.
For the convention center network, 5 "boxes" have been assembled, each of which consists of
up to three of the gateways and an Ethernet switch. The gateways connect to the Ethernet
switch over Cat 5 cable and the Ethernet switch connects the box to the DMC over fiber optic
cable. Figure 55 shows the connections between the transfer switches, the gateways, and the
Ethernet switch.
A044E523 (Issue 1)
83
FIGURE 55.
4-2013
These two examples demonstrate that networks can move both data and control commands
over large distances and among a large number of components. They are examples of how
different protocols, from the high level application layer down to the physical media layer, can all
exist within the same system. We can break these networks down into smaller, simple networks
joined to each other through various types of gateways and the complexity of the system seems
less daunting. Like all networks, when combined with appropriate service procedures, they
contribute to maximizing system reliability and minimizing operating costs by providing the
capability for system status display and control, alarm notification, and data logging and
reporting.
84
A044E523 (Issue 1)
Appendix
9.1
Glossary
Address
An identifier used to uniquely identify nodes on a control network.
American National Standards Institute (ANSI)
The United States government body responsible for approving U.S. standards in many areas,
including computers and communications.
American Standard Code for Information Interchange (ASCII)
A standard for encoding data.
Annunciator
An annunciator is used to give remote indication of the status of an operating component in a
system. Annunciators are typically used in applications where the equipment monitored is not
located in a portion of the facility that is normally attended. The National Fire Protection
Association (NFPA) has specific requirements for annunciators in some applications, such as
hospitals.
Backbone
Typically a fast cable or fiber connection used to connect wiring closets, hubs, and switches. In
a bus topology the backbone is the bus.
Balun
A balance resistor placed between two cable sections of a network to give the entire cable a
consistent characteristic impedance.
Bandwidth
The amount of data that can be transmitted in a fixed amount of time. For digital devices it is
usually expressed as bits per second or bytes per second. For analog devices it is usually
expressed as cycles per second or hertz.
Baud Rate
The speed of data transmission in serial data communications approximately equal to the
number of code elements (bits) per second (BPS). Bits per second are also termed BPS with
the prefix (k) denoting thousands.
Binding
The process of making the logical connections to the network (also called connecting). This
involves connecting network variable outputs to network variable inputs using LonMaker for
Windows software.
A044E523 (Issue 1)
85
9. Appendix
4-2013
Bit
Binary Digit (i.e., a 1 or 0).
Boolean
A logical system used to express one of two states, such as on or off (yes or no, 1 or 0, etc.)
Bus
The main physical data-carrying media of the network data wire. It must be terminated at both
the start and end of the network. The wire is daisy-chained from one node to the next. The
Lonworks network bus cannot exceed specified length (which varies by cable type used) without
the use of a router. Stubs off the main bus wire of a Lonworks network cannot exceed 10 feet (3
meters). Bus can also refer to the physical power-carrying connection between generator sets
and loads in a paralleling system.
Bus Topology
All devices are connected to a backbone cable or bus. The bus topology is relatively
inexpensive to install. Typically both ends of the bus must be terminated.
Byte
For the purposes of control networking, a byte can be defined as a set of eight contiguous bits.
Cellular
Refers to a communication system that divides geographic regions into sections called cells.
The purpose of this division is to make the most use of the limited number of transmission
frequencies.
Channel
A channel is the physical communications media that connects the devices. Most
PowerCommand network installations will have only one channel (UTP cable and 78 kBPS
transmission speed). In a large network, there may be multiple channels and each channel may
or may not be of the same media type. Typically, channels are linked together using routers.
Characteristic Impedance
A parameter of communication cable. The impedance that an infinitely long section of cable
would have.
Client
A client is a device that accesses a service or shared data on another device or server.
Typically, clients would require specific permissions and software in order to utilize the services
or data on the server.
Comma Separated Value (CSV)
A record layout that separates data fields with a comma and usually surrounds character data
with quotes. PowerCommand for Windows uses the CSV record format.
86
A044E523 (Issue 1)
4-2013
9. Appendix
Configured/Unconfigured
The term configured, as used in Lonworks networks, refers to a module that has been logically
installed with LonMaker software. A network image (address and binding information) is stored
in the node. A module that has not been logically installed with LonMaker is referred to as being
unconfigured. The service LED will blink on an unconfigured module.
Connecting Devices
Connecting refers to the process of assigning connections; linking an output variable of one
device to an input variable of another device. This process is also called binding.
Control Communication Module (CCM)
Cummins device that gathers analog and discrete data from power generation equipment, such
as ATS or generator sets, and translates it into LonWorks data so that it can be used on a
Lonworks network. There are two versions generator set and ATS. The design allows for use
with other manufacturer's generator sets and transfer switches. The Module has 3-phase
voltage, current, PF sensing, 32 discrete inputs, 7 analog inputs, and 8 relay outputs and
operates on 5-36 VDC power.
Daisy Chain
A wiring method where each device on a network is wired in series.
Default Gateway Address
A device on an IP subnet that enables communication with other subnets. The device referred
to by the default gateway address is typically a router.
Deterministic
A network characteristic meaning that the time it takes a message to be transmitted between
two nodes is predetermined and consistent. Networks that are used for time critical control
functions need to be deterministic.
Digital Input/Output Module (DIM)
This device is a Cummins manufactured relay board with some dry contact discrete inputs. The
base DIM module has 4 discrete customer inputs and 8 Form-C dry contact outputs. An
Expansion DIM module can be added to the base for an additional 4 discrete customer inputs
and 8 Form-C dry contact outputs. The DIM requires 10-36 VDC power for operation. This
device allows discrete input and output data to be communicated over Lon networks.
Digital Master Control (DMC)
The Digital Master Control is used in paralleling applications to allow supervision of the
operation of multiple generator sets operating in parallel and to provide load management and
capacity control for the system. See T-016 for more information.
A044E523 (Issue 1)
87
9. Appendix
4-2013
88
A044E523 (Issue 1)
4-2013
9. Appendix
Frame
A sequence of bits that is part of a message communicated between two devices. A message to
be communicated is broken down into frames so that it will fit within the parameters of the
protocol. A typical frame will consist of identification of the sending and receiving devices,
control information, data and data integrity bits. The control information is used to signal the
beginning of a new frame, describe how the message is segmented (how this frame fits into a
larger message), and how it is to be routed. Data integrity bits are used to verify that the
message has been received correctly.
Gateway
A device that acts as an interface between two different communication protocols.
Hub
A hub is a network device that serves as the center point of a star configured network, enabling
other network devices to be connected. A hub broadcasts any message it receives to all other
devices connected to its ports.
Internet Protocol Address (IP Address)
Numerical identifier for a network device (192.168.1.1).
Latency
The amount of time it takes a packet of information to travel from its source to its destination.
Local Area Network (LAN)
A computer network that spans a relatively small geographic area. Most LANs are confined to a
single building or group of buildings.
LonWorks Network Services (LNS)
This is the client/server architecture developed by Echelon for communication with a LonWorks
network.
Master/Slave Network
A type of network operating system whereby one device controls all communication on a
channel. This controlling device is known as the master. The slaves are all other devices on a
network. This is different than the peer-to-peer type of network operating system.
Media
The physical path used for transmission of messages between network nodes. Media can be
twisted pair wiring, through the air (wireless), or fiber optic cable, among others.
Modem
Modulator Demodulator. A device that converts a computers digital pulses into audio
frequencies (analog) over a telephone line and converts them back into digital pulses at the
receiving site. Typically, one modem will transmit the data and another will receive the data over
a telephone line at a controlled transmission speed.
A044E523 (Issue 1)
89
9. Appendix
4-2013
Modbus
Modicons Modbus protocol is an old protocol, by technology standards, that has become a de
facto standard for generic network devices and in some cases for devices using different
technologies that need to communicate. Most manufacturers of control devices and protocols
have included the ability to communicate with Modbus protocol. Some examples of different
Modbus implementations:
Modbus ASCII ASCII characters are used to represent the data in Modbus ASCII
messages. This is a serial protocol typically communicating over an RS-232 connection.
Modbus ASCII is considered an Open Protocol.
Modbus RTU One RTU character is used to represent the data in Modbus RTU
messages. This is a serial protocol typically communicating over an RS-232 or RS-485
connection. Modbus RTU is considered an open protocol.
Modbus TCP Modbus TCP is a proprietary TCP/IP based protocol that communicates
over an Ethernet network. The data structure of Modbus TCP is similar to Modbus RTU.
Modbus Plus
An industrial networking system that uses token-passing peer-to-peer communications at data
transfer rates of one megabit per second (MBPS). The network media is shielded twisted-pair
cable.
ModLon Gateway Kit
A device that allows for conversion to Modbus RTU from Cummins Power Generation's
LonWorks protocol for devices.
National Fire Protection Agency Section 110 (NFPA 110)
This section deals with the regulations concerning emergency power system (EPS) regulations
on installation, operation, and monitoring of EPS. CPG equipment must meet the standards set
for EPS.
Network
A collection of nodes that communicate with one another over a common medium (i.e., wire,
fiber, wireless) where information is exchanged via sets of rules (protocols).
Node
The term "node" is a general term used to refer to any device communicating on a network.
Operating System
The operating system is the system software that is responsible for the direct control and
management of the system's hardware and internal control operations.
Parity
In error detecting schemes, parity is a bit (even or odd) that represents the binary sum of the
data transmitted. Primarily used when transmitting data over a long distance.
Peer-to-Peer
A network operating environment where any device on the main network bus can initiate
communication.
90
A044E523 (Issue 1)
4-2013
9. Appendix
Port
The external connector on a device at which the network cable or medium is attached.
Private Network
Network that uses private network address (LAN) (192.168.x.x, 172.16.x.x, 10.x.x.x)
Protocol
Set of rules used mutually by two or more devices or software applications to communicate.
Different protocols often describe different aspects of a single communication; taken together,
these form a protocol stack. Different protocols function at various levels of the protocol stack.
For example, the TCP/IP protocol is a transport protocol that functions at the lower levels of a
protocol stack, where Modbus is an application protocol that functions at the upper levels of the
protocol stack.
Protocol Suite
A collection of protocols that are designed to work together. Protocols such as HTTP, HTML,
SMTP, SNMP, TCP, and IP are often described as being part of the internet protocol suite.
Public Network (Internet)
System of globally connected networks.
Ring Topology
All devices are connected to each other in a closed loop, or ring. Each device is connected
directly to two other devices. Ring Topology networks are very reliable.
Router
A router is a device that directs network traffic between two networks. The term router is used in
relation to control networks to apply to devices that encapsulate and receive data over different
networks or media. Routers receive packets of data, filter them, and forward them to a final
destination using the best route. Most LAN and WAN routers direct packets of data based on
TCP/IP addresses. Routers do not send broadcast packets or corrupted packets. If the routing
table does not indicate the proper address of a packet, the packet is discarded.
Serial Port
A communication port at which data is transferred one bit at a time.
Server
A server is a device that provides services or resources, such as printers and files, for the use
of other devices on the network.
Simple Network Management Protocol (SNMP)
A standard protocol used for managing Ethernet networks.
Slave
A networked device that is controlled by another device. Slave devices do not initiate data
transmission. They respond to commands or requests initiated by a master device.
A044E523 (Issue 1)
91
9. Appendix
4-2013
92
A044E523 (Issue 1)
4-2013
9. Appendix
9.2
Acronyms
Acronym
Description
A/D
Analog to Digital
AC
Alternating Current
ACB
ANSI
AS, or AUS
Australia
ASME
ASTM
ATS
AVR
AWG
BACnet
BIL
CAN
CB
Circuit Breaker
CCC
CCM
CE
Conformite Europeenne
CGT
CSA
CSV
CT
Current Transformer
dB
Decibel
DC
Direct Current
DHCP
DIM
DMC
DNS
A044E523 (Issue 1)
93
9. Appendix
4-2013
Acronym
Description
E-Stop
Emergency Stop
ECM
ECS
EMI
Electromagnetic Interference
EIA/TIA
EN
European Standard
EPS
FAE
FMEA
FMI
FSO
GB
Gigabyte
GCS
GEN
Alternator/Generator
Genset
Generator Set
GFCI
GND
Ground (Earth)
GOOSE
GOV
Governor
HMI
I/E
Import/Export Control
IC
Integrated Circuit
IEEE
ILS
IP
ISO
kVA
kVAR
kW
LAN
LCD
LCL
LED
LLC
LNS
LSIG
MAC
MB
Megabyte
94
A044E523 (Issue 1)
4-2013
9. Appendix
Acronym
Description
Mil Std
NC
NEC
NEMA
NFPA
NO
Normally Open
NWF
Network Failure
NZ
New Zealand
OEM
OOR
Out of Range
OORH, or ORH
OORL, or ORL
OSHA
OSI
PB
Push button
PC
Personal Computer
PCC
PF
Power Factor
PGI
PowerGen Interface
PGN
PI
Proportional/Integral
PID
Proportional/Integral/Derivative
PLC
PLL
PMG
PT
Potential Transformer
PTC
PWM
Pulse-Width Modulation
RFI
RH
Relative Humidity
RMS
RTU
SAE
SCADA
SNMP
SPN
SW B+
SYNC
Synchronizer
A044E523 (Issue 1)
95
9. Appendix
4-2013
Acronym
Description
T-011
T-016
T-030
TCP
THD
UL
Underwriters Laboratories
UPS
UTP
VT
WAN
9.3
9.3.1
96
A044E523 (Issue 1)
4-2013
9.3.2
9. Appendix
Modification of Products
Generator sets and other related products are sometimes Certified, Listed, or otherwise assured
to be compliant to specific standards or codes. This generally applies to the product as
manufactured and shipped from the original manufacturer. Often these products are labeled or
in some way marked as such. Subsequent modifications to the product could alter or violate the
specific code compliance or listing. Product modifications should be, and in some cases must
be submitted to the local authority having jurisdiction for approval.
A044E523 (Issue 1)
97
9. Appendix
4-2013
98
A044E523 (Issue 1)
www.cumminspower.com
Cummins and the "C" logo are trademarks of Cummins Inc.
Copyright 2013 Cummins Power Generation, Inc. All rights reserved.