Connecting a PT100 via DIN or jackplug ?

[AndyBWood]

Evening All,

I'm in the middle of assembling my PID build and would like to make the PT100 sensor 'detachable’ so that its not hard-wired into the PID control unit. The obvious way to do this would be to have some kind of jackplug or DIN socket in the wall of the project box holding all the bits and bobs – plug in the sensor and away you go.

I’m fairly certain I’ve read somewhere that interfering with the cable can induce errors in the temp reading but…… does this just apply to thermocouples ?

Would it be ok to wire a PT100 (RTD) with some type of plug connector without any issues ??? :wha:

Cheers

Andy

 

[ScottM]

Evening All,

I'm in the middle of assembling my PID build and would like to make the PT100 sensor 'detachable’ so that its not hard-wired into the PID control unit. The obvious way to do this would be to have some kind of jackplug or DIN socket in the wall of the project box holding all the bits and bobs – plug in the sensor and away you go.

I’m fairly certain I’ve read somewhere that interfering with the cable can induce errors in the temp reading but…… does this just apply to thermocouples ?

Would it be ok to wire a PT100 (RTD) with some type of plug connector without any issues ??? :wha:

Cheers

Andy

No issues with an RTD lead. As long as whatever you use has a low impedance then you will be fine. The reason for the hassle with thermocouples is that the wire itself is specific to the type of thermocouple, thus splicing something in that isn't made for the purpose can throw the reading off (thermocouple wire auto corrects based on temperature).

The funny thing is that in this case the thermocouple would be a lot easier to deal with as you get some very nice mini thermocouple plugs with flush mounts etc. Loads of options

 

[MacKiwi]

I used an XLR connector for mine. Seems to have worked out ok so far...

http://uk.rs-online.com/web/p/xlr-connectors/6949415/

 

[jamesb]

If it's a two wire RTD then you'll introduce some errors. If it's three or four wire then as long as the connections are fairly similar between all wires then any error will cancel themselves out.

 

[Fil]

+1 for xlr fittings, they are robust and quite cheap on ebay..

 

[ScottM]

If it's a two wire RTD then you'll introduce some errors. If it's three or four wire then as long as the connections are fairly similar between all wires then any error will cancel themselves out.

What errors? It's only a measure of resistance. If the connection put in place makes no difference to the impedance then there will be no error in the reading.

 

[jamesb]

If it's a two wire RTD then you'll introduce some errors. If it's three or four wire then as long as the connections are fairly similar between all wires then any error will cancel themselves out.

What errors? It's only a measure of resistance. If the connection put in place makes no difference to the impedance then there will be no error in the reading.

Every bit of wire and connection gives you resistance. Remember with a pt100 it's only a few mOhm per degree.

With a two wire (uncommon) the resistance of the wires and connections are added to this.

With a three or four wire the circuitry the wires and connections almost cancel each other out.

 

[ScottM]

If it's a two wire RTD then you'll introduce some errors. If it's three or four wire then as long as the connections are fairly similar between all wires then any error will cancel themselves out.

What errors? It's only a measure of resistance. If the connection put in place makes no difference to the impedance then there will be no error in the reading.

Every bit of wire and connection gives you resistance. Remember with a pt100 it's only a few mOhm per degree.

With a two wire (uncommon) the resistance of the wires and connections are added to this.

With a three or four wire the circuitry the wires and connections almost cancel each other out.

I have to disagree. You would be really hard pushed, even for a PT100, to put an error in place IMO. You say it's mOhms but it's actually more like half an ohm. If you introduced a connection with a half ohm impedance then you've picked the wrong one. I would need to wind out a good few metres of cable to get even a 1 degree error, never mind putting in a small connector.

 

[jamesb]

With a three or four wire the circuitry the wires and connections almost cancel each other out.

I have to disagree. You would be really hard pushed, even for a PT100, to put an error in place IMO. You say it's mOhms but it's actually more like half an ohm. If you introduced a connection with a half ohm impedance then you've picked the wrong one. I would need to wind out a good few metres of cable to get even a 1 degree error, never mind putting in a small connector.

.385ohm/degC. Not necessarily linear over the whole range but for general use in our ranges it may as well be.

Depends what accuracy you're after. I wouldn't care less with tens of meters of cable for 3degC. 1degC still requires calibration really when it's installed, 0.5/0.1% definitely requires calibration in situ. Even a class A pt100 is only really reproducably accurate to 1degC.

 

[ScottM]

With a three or four wire the circuitry the wires and connections almost cancel each other out.

I have to disagree. You would be really hard pushed, even for a PT100, to put an error in place IMO. You say it's mOhms but it's actually more like half an ohm. If you introduced a connection with a half ohm impedance then you've picked the wrong one. I would need to wind out a good few metres of cable to get even a 1 degree error, never mind putting in a small connector.

.385ohm/degC. Not necessarily linear over the whole range but for general use in our ranges it may as well be.

Depends what accuracy you're after. I wouldn't care less with tens of meters of cable for 3degC. 1degC still requires calibration really when it's installed, 0.5/0.1% definitely requires calibration in situ. Even a class A pt100 is only really reproducably accurate to 1degC.

My point is that a well calibrated readout with a PT100 would see no noticeable change by wiring in a pin connector. I agree that there may be a slight change, but the temperature of the room would effect it more than putting in a piddly connector. Even trying to monitor if the connector made a change would be pointless unless in a test environment. Holding a brew to a certain temp though?

Saying that it would require re-calibrating and that it would introduce an error, given the discussion here, is wrong IMO. But hey... I'm just a calibrator

 

[jamesb]

My point is that a well calibrated readout with a PT100 would see no noticeable change by wiring in a pin connector. I agree that there may be a slight change, but the temperature of the room would effect it more than putting in a piddly connector. Even trying to monitor if the connector made a change would be pointless unless in a test environment. Holding a brew to a certain temp though?

Saying that it would require re-calibrating and that it would introduce an error, given the discussion here, is wrong IMO. But hey... I'm just a calibrator

My point was the number of wires. Most of them are at least three wire, in which case don't worry too much. Two wire ( like a lot of the cheap ones from China with the instruction to link the other terminal...) will be terrible

But for our homebrewing purposes there's much more to worry about. I know most of the stuff people use is inaccurate, but the important thing is that it is consistant per brewery.

Ex-Calibrator too. And designer for the last few years.

 

[evanvine]

Surely if you are considering the resistance of a plug and socket, should you not also consider the resistance of the screw terminals on the PID its self?

 

[AndyBWood]

Thanks for the disscussion chaps, appreciated.

In short :

With my 3-wire UK bought, hopefully well-made PT100 I’m fine.

After wiring through a decent connector there shouldn’t be a great deal of difference worth worrying about but…….

I can then re-calibrate the PT100 / PID reading againts my (hopefully very accurate ) reference thermometer I picked up on ebay and Bob’s your mother’s brother.

Andy

 

[evanvine]

That's exactly what I've done with both of mine Andy, they work fine!

 

[Fil]

afaik a pt100 probe reads 100ohms resistance at 0C so you could test any plug and leads and the accuracy of your reading device with a 100ohm resistor.

 

[EskiBrew]

James is absolutely right - there will be a measurable effect if using a 2 wire system; whether the introduced error is important is entirely up to you depending on the accuracy you desire.

Personally I would not use a 2 wire PRT; 3 wire minimum with all connections as similar as possible - MiniDIN connectors are fine :thumb:

 

[dennisdk2000]

Can't say I know much about the finer points of electronics (wish I did!), but the guys who designed and built The Electronic Brewery (http://theelectricbrewery.com/) use a three-pin XLR connector for their temperature probes, and here's how they build them: http://theelectricbrewery.com/temperature-probes?page=1

Now, if I only had the space, I'd be making a start on building my own shiny set-up... Ah well, a boy can but dream..!

Dennis

PS: Remember to share pics, so me all can drool and envy you!

 

[EskiBrew]

Yep, should have said earlier, XLR connector should be great too :thumb:

 

[dennisdk2000]

I believe the advantage of an XLR connector is that it locks in, where the DIN doesn't (you'll notice that they re-enforce the cable of the probe and joins at either end with thin wire rope). However, there is a difference in cost.

Dennis