Westermo U100 User Manual
OnTime Industrial Networks
100 Series Installati on Guide
OnTime Industrial Ethernet
100 Series
V 1.0
- 1 -
1. Introd uc tion To OnTime Ne tworks
1.1. History Of Ethernet
1.2. Industrial Ethernet Market
1.3. Switches vs. Hubs
1.3.1. Switch Operation. MAC Address Learning
1.4. Twisted Pair P ort S pecification
1.4.1. MDX/MDIX Technol ogy , Crossed / S traight Cables
1.4.2. Auto-Negotiation Protocol & Manual Configuration
1.5. Fibre Optic Port Specification
1.5.1. Fibre Optic Care
2. Switch Specifications
2.1. Power Supply Specif ic ation
2.2. Power Supply Connections
3. Techni cal Specification
3.1. Returns Procedure
3.2. Physical Specification
Index
OnTime Industrial Ethernet
100 Series
V 1.0
- 2 -
Section 1
Introd uc tion to OnTim e
Company History
OnTime is dedicated to the implementation of industrial and deterministic Ethernet
infrastruc ture. OnTime Networks is a priv ately hel d company based in Norway and S weden.
We work cl osely with a number of large aut omation compani es; enhancing older proprietary
networks and working in partnership developing new network technology.
Mission Statement
OnTime's mission is to provide an extension of Ethernet to the factory floor by offering
Ethernet products that fulfill industrial and real time requirements.
Core Tech nology
OnTime's Ether net switches are based on a robust and r eliable i ndustr ial design for maximum
life cycle and minimum life time costs. Real time properties are implemented in order to
achieve determinism for real time criti cal applications.
OnTime Industrial Ethernet
100 Series
V 1.0
- 3 -
Section 1.1
History of Ethern et
In late 1972, Metcalfe and his Xerox PARC colleagues developed the first experimental
Ethernet system to interconnect the Xerox Alto, a per sonal workstation with a graphi cal user
interface. The experimental Ethernet network was used to link Altos to each other, an d to
servers and laser pri nters.
The signal cl ock for the experiment al Ethernet interf ace was deriv ed from the Alto's sy stem
clock, which resul ted in a data transmission rate on the experimental Ether net of 2.94 Mbps.
Robert Metcalfe's first experiment al network was called the Alto Aloha Network.
In 1973, Robert Metcalf e changed the name t o "Ethernet," to mak e it clear that the system
could support any t ype of computer; not just the Xerox Altos and to poi nt out that his ne w
network mechanisms had ev olv ed well beyond the Aloha system. He chose to base t he name
on the word "et her" as a way of describing an essent ial feature of the system: the physi cal
medium (i. e., a cable) carries bits to all stations, much the same way that the old "l umi niferous
ether" was once t hought t o propagate el ectrom agnetic wav es through spac e. Thus, Et hernet
was born.”
``The diagram ... was dr awn by Dr. Robert M. Metcalfe in 1976 to present Ethernet ... to the
National Computer Confer ence in June of that year. On t he draw ing are the or iginal t erm s for
describing Ethernet. Since then other terms have come into usage among Ethernet
enthusiasts.''
OnTime Industrial Ethernet
100 Series
V 1.0
- 4 -
Section 1.2
Industrial Ethernet – What Are The Differences?
Within the UK Market, and possibly the majority of Europe, Ethernet is moving into the
Automation Industry. Manufacturers are exporting their legacy protocols onto Ethernet,
designing new IP based communication protocols and providing embedded Web-Pages
within PLCs to provide real-time information using simple tools like Internet Explorer and
Netscape.
However, the domain of Ethernet has always been controlled by the IT department who
configured office networks normally with an iron fist and dictated to the company how the
network would be de signed with c om plex recover y prot ocol s li ke spanning t ree a nd SNM P to
help with f ault fi nding and system analysis. If a network failure occurred t he IT department
would casuall y look at repairi ng the equipment - there was no real r ush as it was an office
network. Howev er, with Industrial Ethernet you need very fast repai r time, and, with an IT
department not present on the factory floor the maintenance personnel need to be made
aware of the fault, find the error and repair it - quickly.
Industrial rated Switches are intended to be installed in harsh conditions and electrical
environments with the added benefit of fast recovery of a network failure. The On-Time
switches are an excellent example of how such Switches should be designed – very high
operating temperatures, fast r epair of redundant ring, layer 2 and layer 3 priority switching etc.
Without doubt, On-Time switches ar e technically superior to many similar m odels available on
the market.
OnTime Industrial Ethernet
100 Series
V 1.0
- 5 -
Section 1.3
Switches vs. Hub s
A hub consists of a number of ports normall y with either RJ-45 ( copper) socket s and / or fi bre
optic ports that have a number of different st yl es of fibre optic sockets. Usuall y a ‘patch cable’
is connected to the hub; the other end is normally connected to a device (PC, Printer etc).
Note: It should be noted that when a hub requires an ‘up-link’ connection to a further
hub a cross-over styl e cabl e is req ui red.
A hub has no intelligence and therefore is unable to identify addresses or any information
contained wit hin the Header fr ame of an Ether net packet. This m eans that it i s not capabl e of
determining whic h por t to send the frame to. Therefor e, every frame is sent to every port.
Note: Industrial hubs can only connect to equipment that operat es at the same speed .
A network of repeaters and hub s is called a ‘Shared Ether net’ or ‘Collision Domai n’. Various
systems will all compete with each other using ‘Carrier Sense Multiple Access / Collision
Detect’ (CSMA/CD) pr otocol. This means that onl y one system is allowed to proceed wi th
a transmission of a frame within a Collision Domain at any one time. This is a major
disadvantage when usi ng Hubs and Repeaters within a network.
If a hub sees a collision on a cable segment, it is detected and a ‘jam’ signal is generated.
The ‘jam’ signal is sent to all connected dev ices. This ensures t hat every devic e is aware of
the collision and t hey do not attempt to transmit duri ng the collision.
All Ports Receiv e the Sam e Ether net Frame
To summarise, hubs operate with the following limitations:
• Only a single speed of oper ation – no abilit y to automati cally c hange between 10M or
100M.
• Only one system is allowed to proceed with a transmission of a frame within a
Collision Dom ain at any one time.
• Hubs require special ‘cr ossed’ cabl es to enabl es l i nks f rom Hub t o Hub. (I f no up-l i nk
port is present)
OnTime Industrial Ethernet
100 Series
V 1.0
- 6 -
Section 1.3.1
Switch Ope r a tion
Introduction
A switch has to forward and receive packets from one LAN or device to another. The switch
could forward all pac k ets, but if this was the case it would have similar behav ior to a hub.
It would be more int elligent if the switch only forwarded packets which need to travel from one
LAN or device to another. To do t his, the switch must learn which devices or LANs are
connected to each port. In simplistic terms; it needs to learn the destination and source ports
of each and every packet r ec eiv ed on eac h individual Switch port. O nc e lear nt, any identically
addressed packet will be automatically be f or warded.
Error Detection
The switch stores ev ery incoming packet and scans thi s for error s, usually by checki ng the
frame CRC (cyclic redundancy check sum). If any errors are found or det ected the packet i s
discarded. In addition each frame is checked for size. Undersized packets (less than 64
Bytes) and over si z ed pac k ets (more than 1518 bytes)* are also discarded.
Once these basic checks have been carried out the switch can then start learning packet
source and destinat ion informati on.
Note: When implementin g Packet P rio rity this increases to 1522 or 1536 Bytes.
Flooding
The switch needs to make a decisi on r egar ding which port(s) t he pac ket is to be forwarded to.
This decisi on is based upon the MAC tables that are mai ntained and updated automatic ally
by the Switch. The process is kno wn as Layer 2 Swit c hing.
When first powered on the MAC tables within the Switch are empty. When a packet is
received on a port the S witch doe s not kno w where th e desti nation MA C address is l ocated.
The Switch learns the address by ‘flooding’ the packet out to all ports. Eventually, the
destination node r esponds, the address i s located and the S witch remembers the desti nation
port. In simpli stic terms; when a Switch receiv es a packet on a port it stores the source M AC
address in the MAC table that corresponds to that Port. The flooding technique is always
used with Broadcast and Multicast packets. If the switch is equipped with multicast
management then multicast packets will not be fl ooded.
OnTime Industrial Ethernet
100 Series
V 1.0
- 7 -
Loading...