PID Controller - MEW 269 - wrong connector

That's exactly why my Rotary Table Indexer uses a 5 pin DIN plug & socket to power the stepping motor. It's an awful bodge, intended to be temporary and never fixed. It's one of my shameful secrets.

A good reason not mentioned for using proper connectors is that some sensors come pre-fitted with them. Not having to rewire to match an XLR would save a few minutes work.

I think Dave's corrections are useful, but don't think the article was spoilt by the omission. Sometimes authors can't win. If he says he used XLR connectors in his build, some might take that as an essential instruction. I reckon Nige is right: XLR was used because it was handy.

Dave

Sorry Muzzer, but not so. That's why, even in the same form factor, you have a range of connectors which are appropriate for E, J. K etc. - they use appropriate connection pins or plating. e.g.: http://www.precision-measurements.com/pdf/connector-systems-for-thermocouples.pdf

Dave
(Edit: corrected link)

Edited By Dave Martin on 16/06/2018 18:27:19

Another useful datasheet here: **LINK**

https://docs-emea.rs-online.com/webdocs/15bb/0900766b815bb539.pdf

Edited By Michael Gilligan on 16/06/2018 18:33:14

and here are two RS data sheets that list the contact materials used for different types of thermocouple
**LINK**
**LINK**

from which:

[quote]

  All THERMOCOUPLE connectors use true thermocouple alloys for optimum accuracy, Except for Type R, S & B which use compensating alloys.

  Thermocouple Pins

  Type /  ‘+’ Positive Pin  /  ‘-‘ Negative Pin

  J  /  Iron  /  Copper Nickel

  K  /  Nickel Chromium  /  Nickel Alloy

  T  /  Copper /  Copper Nickel

  E  /  Nickel Chromium  /  Copper Nickel

  N  /  Nickel - Chromium - Silicon  /  Nickel Silicon

  R  /  Copper  /  Copper Alloy

  S  /  Copper  /  Copper Alloy

[/quote]
(Edit: - table format disappeared so separate columns with /  - crossposted with MichaelG at the same time!)

Edited By Dave Martin on 16/06/2018 18:42:37

If one is not to fussy about accuracy then using any old connector for thermocouples is not going to be a problem although using a bulkhead connector is definitely not a good idea as the internal and external parts of the connection will probably be at different temperatures.

Mention was made earlier in this thread about controlling a heater at 100 degrees C, the PID controller will have no problem but if accuracy is required then a thermocouple (whatever type it is) is a poor choice of sensor. A PRT (Platinum Resistance Thermometer) is a much more suitable and accurate device and is not fussy about connectors.

I suppose this whole thread revolves around what is 'fit for purpose'. If the authors project did what he wanted then that is what matters. If Neil wants to use XLR's because they are robust and latching thats fine, personally I put electrical integrity first.

Ian P

Not me! I don't do much electronics at home now; having been a professional electronics engineer for 40 odd years I've got a bit bored with it. So I've binned or flogged most of my components. As and when I design a circuit I just order what I need from Farnell and it's here the next day. There's no point in wasting time hunting through boxes of components for something you may, or may not, have. And for SM components it's darn near impossible to tell what the value is anyway, without measuring it. Much simpler to buy new.

Mind you I have kept the better quality components such as Jackson Bros variable capacitors and Denco radio coils.

Andrew

If one really wants robust and latching connectors one could try Lemo. We used them extensively on high end racing cars, mostly Indy cars and a bit of F1. Mind you, if you have to ask the price you can't afford them!

Andrew

A bath of boiling water would suffice for maintaining 100 degrees C (373K)!

My 100 degrees C was a typo, I meant to put 100.5 degrees C for which boiling water is unsuitable!

Many of the controller I installed were for cooling/refrigeration systems for which thermocouples just do not cut it.

Ian P

.

^^^ Assuming that just the right barometric pressure prevails.

MichaelG.

After reading this thread, it seems less improbable that people died in the middle ages during arguments over whether it was 6 or 7 angels that could dance on a pinhead.

This really is like that argument, so abstruse as to be arguing for the sake of it. Just as pointless as the argument over some extra zeros appearing in the same article in MEW 269. Life happens. Just live with it. Or don't. Your choice, but is it really important enough to go round and round in circles about this sort of thing? There's a limit to how much pedantry is amusing and fun to read. I read most posts on this forum for the gems people come up with - and there are quite a few, but dilution of the gems still has to be waded through to find the gems., and, I dunno, this weekend has been a bad one for dilution.

Regards

Richard.

Edited By richardandtracy on 17/06/2018 21:19:57

Add the appropriate amount of common salt, or any other salt for that matter? As long as the bath were maintained at constant volume.... 58g of salt per litre of water would do fine.

.

Richard,

If you really find this discussion so tedious, you might like to spend a few minutes reading this note by Prof. Peter Harrison instead: **LINK**

http://nq.oxfordjournals.org/content/63/1/45.full.pdf+html

It's one of several items linked from the Wikipedia page: **LINK**

https://en.wikipedia.org/wiki/How_many_angels_can_dance_on_the_head_of_a_pin%3F

MichaelG.

Your figures are seriously adrift Richard! Modern quantum analysis suggests an upper bound of 8.6766×10⁴⁹ angels per pinhead.

Well maybe, but professional engineering is all about attention to detail, and an extra zero can have unexpected consequences! While MEW isn't for professionals there's no reason why it shouldn't strive to get things right.

Andrew

Can I ask a question, you will have to be gentle with me as I only did GCSE Electronics at school.

If it is so important to have the right connector materials so like materials come into contact at a joint then what happens when you come to connect the connector? Assuming it is soldered to a circuit board or uses screw terminals then they would not match and again if type of wire is critical what happens when the signal has to flow around the circuit board which would not be the same as the thermocouple wire?

That's an excellent question, Jason

The full answer really depends on how [and how well] the measuring instrument is designed, but; as I mentioned on p1:

"... the connector block will become the cold reference junction ..."

If the instrument is designed accordingly, that's fine.

MichaelG.

.

Edited to put the second 'n' in connector 

Edited By Michael Gilligan on 18/06/2018 13:38:16

That's more than I did at school.

Andrew

To answer Jasons's question,

Assuming the thermocouple is connected to some measuring/indicating/controlling device, it is calibrated using a millivolt source. Simply inject the corresponding mV that equate to a given temperature and adjust the electronics to ensure the device indicates that....Usual adjustments are Zero and Span......I have encountered devices that have a linearity adjustment too. The Zero adjustment doesnt necessarily correspond to the literal value 0, it may be the measuring/indicating device could measure between a negative and a positive temperature....

Of course, the mV ipnut signal injected is type dependant, this is where reference tables are used for alternative types of tc...and of course taking into account cold junction compensation.....in fact some of the more sophisticated test equipment has cjc built in.....

Edited By John Rudd on 18/06/2018 15:29:50

So kind of comes back to what I said that the controller could be programmed to compensate for any error caused by the connector and that would also be what Michael said about correctly designing the instrument to allow for how it was connected.

Andrew, I seem to remember that there was no O Level Electronics offered by any of the boards so that is no doubt why us comprehensive kids got to take the GCSE