Reminder of the thermocouple measuring principle
At the terminals of a circuit formed by two conductors (couples) of different type (e.g. ironconstantan) an electromotive force (emf) can be measured which is directly proportional to the temperature variation recorded in the hot zone.
The materials used depend on the temperature range to be measured. The couples are symbolised by the letters: “J-K-S “.
To be able to move the measurement zone (cold zone) away from the hot zone we use
compensation cables: they are less expensive than thermocouple cables (markings JC –KC – SC).
The extension cables provide the same function with a greater precision (lower tolerance).
They are more expensive than compensation cables (markings JX – KX – SX). A colour code is used to identify the different cables according to each country’s standards.
Thermocouple compensation cables
Extension cables are used to extend the thermocouple circuits, they take the form of an electric cable whose conductors are made of the same materials as those of the thermocouple.
Compensation cables are made of other materials (cheaper) whose thermoelectric characteristics are identical up to 100°C. Compensation cables are made of other materials (cheaper) whose thermoelectric characteristics are identical up to 100°C.
Practical information for Special Instrumentation Cables
The positive conductor always has the same sheath colour. The negative conductor is always WHITE.
Important: any temperature inversion of a compensation cable generates spurious thermoelectric junctions which affect the measurement precision and stability (fluctuations linked with the variation in ambient temperature).
Compensation cable codings
Colour codings
Network cables
Profibus cable
The Profibus-DP protocol uses a high speed RS485 serial link and imposes an impedance
of 150 ohms.
Electrical characteristics:
Service voltage: 100 V
Test voltage 2,000 V
Impedance: 150 Ω+/- 10%
Capacitance: 30 pF/m
Electrical resistance: 50 Ω/km
Loop resistance: 100 Ω/km
Ethernet cable
Ethernet cable is now very often used because of the developments in PLCs and PLC control systems which are based around industrial computer networks. It uses two pairs of twisted wires, one pair is used to receive data signals and the other is used to transmit data signals.
The two wires in each pair must be twisted together along the whole length of the segment, this technique is often used to improve signal quality.
RJ45 Connector
This connector is derived from that used for the telephone (RJ11) but it is physically incompatible with it (it is wider) and can contain more wires (8 compared to 6, whereas the conventional RJ11 uses only 4 wires).
The standard pin configuration is shown here (following figure) with the standard colours. It must be noted that the odd pins are always those with striped colours.
Straight RJ45 cable
It is the most widely used model, it is always used when connecting an Ethernet interface to a hub or to a switch. The pin assignments are identical at each end of the cable:
RJ45 crossover cable
Used to directly connect two Ethernet interfaces together. The pin configuration at one end is different to allow communication to take place: the transmission and reception wires are reversed.
As you can see in this example, using a crossover cable we can directly connect two computers in an Ethernet network.
Determining the type of RJ45 cable
Take the two ends as shown in the following diagram and look closely at the order of the wires:
IMPORTANT: The colour codes indicated in the previous tables correspond to a standard but it is quite possible to find cables which use wires with completely different colours.
- if the colour sequence is identical at each end: straight cable
- If pins 1-2 and 3-6 are reversed: crossover cable
- Neither one or the other: “unknown” cable, or at least a cable which does not respect an Ethernet configuration.
RJ45 cable categories
Each RJ45 cable is attributed a category which defines the maximum amount of data (data
speed) speed which it can transfer without errors.
Normally, each RJ45 cable you find on the market is marked along its length with a set of data and the cable category. The following table gives the different categories with their current usage:
Making your RJ45 cable
The first thing you need is some category 5 cable. This can be obtained from computer or
electronics retailers, etc. You can also choose to shorten the cable supplied with the
machine.
The crimping tool for RJ45 connectors is more expensive. This one is relatively cheap
(around 15 euros). Near its hinge it has two blades which strip the cable, as well as a
crimping part and a cutting part.
I prefer to use this one, it only crimps but it does a very good job!
You also need the RJ45 connectors and any coloured caps which may be necessary. The
caps are used to identify the different cables and protect the Ethernet connector’s clip. You
obviously have to insert the cap before crimping the connector.
I would not use one on this cable because it is designed to remain in place and, because
of its length, we can immediately see what it is connected to.
The connectors can be different, with guides, without guides, screened. In this case these
are simple connectors.
On this side we can see the metal pins which will later fit onto the cable’s wires.
Personally, I don’t use the stripping part
of the various tools, I prefer to use a
utility knife, and make a cut between 2
and 3 cm from the end of the cable
In this way, I know that the conductors will
not be damaged and I have enough length to
untwist the pairs and lay them flat.
And the conductors are laid out flat in the correct
order:
- white/orange
- orange
- white/green
- blue
- white/blue
- green
- white/brown
- brown
For a straight cable, the two ends must be placed in this manner.
For a crossover cable, one end must be as indicated above and the other must be as
specified below:
- white/green
- green
- white/orange
- blue
- white/blue
- orange
- white/brown
- brown
The two pairs currently used (base 10 and base 100) are those which will be crossed over
(or left straight). After cutting the conductors straight, we slide them fully into the connector, as shown here.
Then push the sheath as far as possible.
We can see that the cables are fully
inserted.
connector
The yellow arrow indicates the contacts which will be inserted onto the conductors, the blue arrow indicates the plastic locking device which will secure them and their sheath in the connector. We insert the connector carefully but firmly into the crimping tool.
When the connector has been clipped into the tool, check that the cable and sheath are fully inserted before crimping. When it has been crimped, repeat the operation for the other end of the cable.
And our cable is now ready to use. In this example we made a 10 cm long cable but you can make the cable any length you wish.
Reminder on HUBS and SWITCHES
General description:
All the data flowing in the network transits via these units. They have from 4 to 48 RJ45 ports and can thus interconnect as many Ethernet interfaces. They can also be interconnected together via an uplink port.
This uplink port is always shared with one of the conventional RJ45 ports, which means that you lose one port when you connect two hubs/switches together. Thus, connecting two 8-port hubs together will result in a total of 14 ports available instead of 16. There are two types of hubs/switches:
conventional hubs/switches (desktop/palmtop): the most commonly used models, they are also the cheapest, and they are designed to be placed on a desk and therefore have an attractive design. They have from 4 to 16 ports and, depending on the model, the power supply may be internal or external.
rackable hubs/switches: they are the largest and most expensive models, they are designed to be installed in rack units (more commonly known as cabinets). They have from 8 to 48 ports and have an internal power supply. The recent models are beginning to integrate gigabit Ethernet ports (1000baseTX or 1000baseFX). The topof-the-range models very often have a remote monitoring software suite. It must also be noted that there are stackable models available. Their specific feature is that they can be interconnected together (up to 8 devices depending on the models) via special boards thus allowing very high speed transfers between all the switches without the risk of creating bottlenecks.
Hub Operating Principle
A hub retrieves the signals from a port and sends them to all the other ports. This means that each data packet from an Ethernet interface connected to the hub is sent to all the other interfaces present on this hub. Thus we are sure that the intended receiver of the packet actually receives it.
The problem is that the packet is also received by all the interfaces which it is not destined for. This generates a lot of unnecessary traffic on the network and the network becomes more and more saturated as more and more Ethernet interfaces are added to it. Since a hub has no means of managing the traffic it receives, the packets very often bump into each other (collision principle).
These collisions fragment the packets and so they have to be sent again, increasing the transfer times and therefore greatly reducing the effective speed of the network.
Switch Operating Principle
Whereas hubs only transfer packets over the network, switches are capable of managing the packets they receive in different ways. Their main feature is that they can consult the MAC address of the sender and of the receiver in each packet.
The MAC address is the unique ID number of each Ethernet interface. By keeping a trace of these MAC addresses, a switch knows which port each Ethernet interface is located on.
Practical Example
A packet arrives on port 2 with X as destination address and Y as source address. The switch immediately knows that address Y corresponds to port 2 since the packet arrived via this port. At the same time, a packet arrives by port 5 with Z as destination address and X as source address. The switch now knows that address X is on port 5 and thus knows the destination of the first packet from port 2 (with MAC address Y).
In theory this series of events happens only once for each MAC address because each switch has an address table containing this data for future reference. In addition to reducing the unnecessary traffic on each port, recent switches are capable of reducing the number of collisions even further by using CSMA/CD (Carrier Sensing Multiple Access/Collision Detection).
This feature is used among other things to check the state of the line before sending data. If it detects that there is traffic on the line, it waits till the line is free before making the transfer.
CSMA/CD also allows the switch to query each packet it receives and to reject those which are fragmented or damaged, thus reducing the unnecessary traffic even more. Finally, a last technical point: most switches are of the “store-and-forward” type. This means that a switch retrieves a complete packet before sending it to its destination.
The switch can therefore analyse the packet (e.g. to find out if it is a fragment resulting from a collision) and decide if it must send it or reject it. Store-and-forward switches must be differentiated from the cross-point models: cross-point
models start to send the packet before they have completely received it.
This gives a shorter latency time but these models are much more expensive and the store-and-forward technologies have reached such a level of efficiency that cross-point switches are extremely rare. All the switches available on the market are store-andforward switches.