Thermocouple Interface

I've been reading up on thermocouples (TC) interfaces and the concept looks simple in theory, but possibly not so easy in practice. It looks like it is important to use an amp with a very low Vos because of the small signal.

I am thinking that the cold junction compensation can be done in software by using a thermometer on the board near the connector. Otherwise it doesn't look too hard. I'll be using low pass filters to limit high freq noise and most likely use a synchronous sampling method to integrate 50/60 Hz noise away. If the TC is connected to the board with an extension cable, where is the cold junction that needs to be compensated, on the board still or at the TC to cable connection or both? I expect it will be at the extension cable junction since the extension cable to connector will be identical metal types. I guess we'll avoid extension cables...

I am not looking to use any of the TC interface chips available because they tend to be pricey and not so small.

Are there any aspects I am missing that will make this not so easy?

Rick

Reply to
rickman
Loading thread data ...

That's about it. One thing that you're missing that will make it easier is that anybody who's worked with thermocouples for any amount of time at all know that thermocouple extension cables are made of the same stuff as the thermocouple, and are matched to it, precisely so that the cold junction is in a controllable spot.

If for some reason I needed to make a system that terminated the thermocouples remotely, I'd make a postage-stamp-sized board with processor, op-amp, ADC and thermistor, that sends a digital message home to mama.

--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
 Click to see the full signature
Reply to
Tim Wescott

would these be helpful?

formatting link

Cheers Don...

=====================

--
Don McKenzie

Dontronics: http://www.dontronics-shop.com/
 Click to see the full signature
Reply to
Don McKenzie

Thermocouple extension cable is thermoelectrically essentially the same as the thermocouple wire, except that it usually operates at a lower temperature, is often less expensive, and may be flexible, sheathed, screened, armoured etc. You treat an extended thermocouple just as if it were one long thermocouple.

Extension wire may be made of the same metal pairs as the thermocouple, or metals which have the same thermoelectric characteristics, albeit at a lower temperature.

This means that you have to use the right type, connect it using the correct polarity, and that the cold junction is where the extension cable joins your circuit and connects to copper, often in some sort of 'isothermal block' with a CJC thermistor.

I'm sure you knew this, but it wasn't clear from the above.

And there are plenty of dedicated chips, of course.

Cheers

--
Syd
Reply to
Syd Rumpo

I think the idea is that a cheap microprocessor and a good op-amp may be less expensive (and possibly more accurate) than a dedicated chip.

Certainly that's why I'd be thinking along the lines that Rick is.

--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
 Click to see the full signature
Reply to
Tim Wescott

Analog Devices have quite a good introductory video series called Thermocouple 101 here

formatting link

and Agilent(once HP) have a great app note for several thermal interfaces including t/cs called AN290 Practical Temperature Measurements

formatting link

-Aubrey

--------------------------------------- Posted through

formatting link

Reply to
antedeluvian

o

Yes, this is pretty much what I had figured out as I wrote my post. I actually studied this sort of thing as a chemist and I was making it too hard. If you join the TC wires to two wires of the same composition, that is your "cold" junction where ever it is. So as you say, it is necessary to use extension wires of the same material as the TC.

Thanks to everyone for your replies. It is good to know I am on the right track.

As to why not use a standard TC chip... I want the circuit to serve as a general ADC input for flexibility. I need to weight the trade offs. Sometimes flexibility is more work than just using one of each. TC signals are very low level and need a lot of gain. That is very different from typical ADC inputs.

Rick

Reply to
rickman

And often in practice have horrible common mode noise, let's not forget.

--
Rob Gaddi, Highland Technology -- www.highlandtechnology.com
Email address domain is currently out of order.  See above to fix.
Reply to
Rob Gaddi

I have not built this, but it may be a good example to follow.

formatting link

Reply to
hamilton

Computer-MicroController

formatting link

Holy crap! I just took a look at your board and it sells for $12 using a $10 chip! How do you do that? I guess your volumes are a bit higher than I expect mine to be. This part is also not recommended for new designs. How long do you expect to be able to ship this?

I don't see a manual on any of the product pages. The manual is listed, but there is no link.

Rick

Reply to
rickman

Computer-MicroController

formatting link

Hi Rick, this was made by Olimex, and released only 64 days ago.

As such, I am still waiting for the PDF manual to be released from Olimex.

As they do build in thousands, I don't see them using a product that is obsolete. But I haven't done any research on this to verify one way or the other. You may need to ask them, but I know they do very careful research before committing to a new design.

Cheers Don...

==========

--
Don McKenzie

Dontronics: http://www.dontronics-shop.com/
 Click to see the full signature
Reply to
Don McKenzie

ml

bit

.

bsolete. But I haven't done any research on

they do very careful research before

A *new* dsign eh? Olimex says the design uses the MAX6675 and the Maxim web site says the MAX6675 is not recommended for new designs. That usually means the part is not long for this world. Maxim has a similar part that is likely pin compatible so I guess it's not a big deal.

Rick

Reply to
rickman

obsolete. But I haven't done any research on

they do very careful research before

I'll let Tsvetan of Olimex know your findings Rick, and see if he has a comment about it.

Cheers Don...

==========================

--
Don McKenzie

Dontronics: http://www.dontronics-shop.com/
 Click to see the full signature
Reply to
Don McKenzie

There you go Rick, I just got this (below) response back from Tsvetan of Olimex.

Cheers Don...

============================== Hi Don this guy is right, Maxim released new IC with better ADC accuracy 14 bit which is pin to pin compatible, so once we sell these 1000 MAX6675 I got in stock we will switch to the new MAX31855 the difference is that the new chip can work with minus temperatures and over

1023.75 C which may be restriction for some applications, but if you do not need to measure minus temperatures and over 1000C MAX6675 is still good! Best regards Tsvetan ==============================
--
Don McKenzie

Dontronics: http://www.dontronics-shop.com/
 Click to see the full signature
Reply to
Don McKenzie

Olimex.

=3D=3D=3D=3D=3D=3D

Hey, I gotta like Tsvetan!

Rick

Reply to
rickman

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.