OT 2-wire, 3-wire, 4-wire thermocouple

Hello guys, I know this is off topic but can anyone explain me difference between 2-wire 3-wire and 4-wire thermocouple. I saw in catalog same thermocouple but with three different version. Is 3.wire is bettern than 2-wire and why?
Maybe answers is valid not only for thermocouples but other sensors...
Thanks

A 2-wire is the least accurate and 4-wire the most. The extra connections are to allow the measuring of the typical losses within the cabling between device and controller.....


Hope this helps

Can you be more specific please, I saw on data sheet that for example 1 and 2 are sort circuited and 3-4 are short circuited. I notice that same thermocouple is used, but I don't understand how more wires will improve accuracy

This may help re-affirm....

http://zone.ni.com/devzone/conceptd.nsf/2d17d611efb58b22862567a9006ffe76/2068782e403b155a862567ed00508f3a?OpenDocument#1

I think you are thinking of RTDs, not thermocouples.

An RTD is more precise than a thermocouple. RTDs measure temperature based on the resistance change of the probe material, usually a very fine platinum wire. Because this resistance change can be very small a means of cancelling out the resistance of the lead wires is needed. This is done by adding a third wire and even a fourth wire.

RTDs are not interchangable with thermocouples, though some instruments have input terminals for both types of devices.

Thermocouple probes can be purchased which have two, three, or even more thermocouple junctions inside a single probe sheath. These will have two wires per thermocouple. However, because you mentioned three-wire what you are looking at is almost certianly an RTD and not a thermocouple.

Read the following PDF file for more information on how the three wire and four wire RTD models compare.
http://www.omega.com/temperature/Z/TheRTD.html

Whoops , never noticed that!! I just saw 2/3/4 wire connections and thought we were discussing RTD's.


Rob.

With a 3 wire system you have one for the supply that is shorted at the sensor position to another one.

This short circuit / close loop is used to measure the resistance of the wires from your instrument to the sensor.

You take out the connecting wires resistance so you only measure the sensor resistance which is the thrid wire. The current flow from this third wire is matching the variation in temperature.

Imagine in a hot day, the wires would elongate hence becoming of a smaller diameter, the junctions could also have a variation in resistance and whatever could affect the TOTAL resistance. So this wire remove this incertitude (is this a word?) from you measuriong circuit.

In such system, Wire 1 and2 and 3 SHOULD be the same resistance but are they?

One way to prove it is to repeat the same pathern with yet another wire (No.4) to measure and remove the resistance of wire # 3.

Hope this helps.

Normally you would NEVER need a 4 wire RTD. Are you in a LAB?

This is or should be used in high precision measuring system OR when the cables are in so deep doodoo that you want to make certain nothing is giving you a false measurement.

Yes it is about RTDs I stand corrected.
Thanks for help

Basically all RTD's only have 2 poin junction the 3rd & 4th wires are connected to the "resistor" a 2 wire device.

you can use 2,3,4 wire RTD's on any RTD input no matter how many wires are required. i.e a two wire to four wire simply connect a link from each RTD connection to the two others on the card, if using 3 wire input & 2 wire rtd connect one side of 2 wire rtd to the 3rd connection & iff using a 4 wire rtd to 3 wire input link 2 of the reds (usually come with 2 reds & 2 whites) together on one input & the two whites to the other connections

the only thing is noise & wire resistance the idea of the extra wires are for noise & wire resistance reduction.

parky said:

if using 3 wire input & 2 wire rtd connect one side of 2 wire rtd to the 3rd connection

Click to expand...

You're right, and I've used this trick many times. You should always check how much error is going to be induced, though, by jumpering at the input card.

For example, on a recent job in Idaho I had a 3-wire input and a 2-wire RTD on an alarm unit. I jumpered as you described. There was approximately 500 ft. of #18 AWG between the input and the RTD location (250 ft. each direction). That added about 3.25 Ohms to the resistance. (#18 AWG is 6.5 Ohms/thousand feet at 77 °F.)

I was using a Pt 100 Ohm RTD, 0.00385 Ohm/Ohm/°C alpha, and my alarm trip point was 240°F or 121 °C. This corresponds to 146.5 Ohms RTD resistance. The 3.25 Ohm error meant that when my input measured 121 °C the actual resistance at the RTD was 143.2 Ohms, and the actual temperature monitored was only 115 °C or 239°F.

In my case the error was not significant, and was on the safe side, and I could tweak my trip point if I wanted to. So, the jumpering worked just fine. It had the added benefit of confusing the heck out of the guy from GE, who wanted the contractor to rip out all of the RTD wiring and was fairly obnoxious about it!

On the other hand, in some applications a 10°F error is going to be a problem.

The moral of this long story is that it is OK to work around differences between the preferred and the actual equipment. Just make sure you know the impact of the work-a-round and account for it in the control.

For that length of a run (500 feet), you could also make your own 3 or 4 wire out of a 2 wire just by running 3 or 4 conductor to the sensor.
Jumper at the sensor connector instead of the Input module.

You are right, Keith, but in this case the contractor had already pulled two conductor cable, his conduit was full, and his time was up.

Always the deciding factor! To get the numbers to match, you could put precision resistors to match wire resistance, but an offset in software would make more sense. Is it just a straight offset, or do you need a correction factor?

It's essentially a constant offset to temperature across the measurement range. For example, 100 Ohm = 0°C and 103 Ohm = 8°C for a change of 8 °C. At higher temperatures, 150 Ohm = 131°C and 153 Ohm = 139°C for a change of 8 °C.

The simpified formula for resistance at a given temperature is:
R = 100 x (1 + A x T).

For a given temperature difference this solves to:
R2 = R1 + (100 x A x (T1 - T2))

For a given resistance difference this solves to:
T2 = T1 + ((R1 - R2)/100 x A)

This is for a 100 Ohm Pt RTD, and A = 0.00385

Thank you all for your replies, it is very interesting and valuable. I learned a lot here...