Cold Junction Reference Temperature Control

[57eele]

Hi
I am trying to source a cold reference oven/device that will maintain the CJ temperature at 45 DegC.
This is to replace existing devices that are decades old.
The number of thermocouple junctions is limited to around 50.
This is for the power industry in the United Kingdom.
Thank you in advance.

 

[hacksaw]

the current method is to correct for the cold junction using software, of course you know that, but it is going to make it very difficult to track down a turn-key manufacturer of such an oven today.

you'll have a choice of repair or rebuild the existing units, build your own temperature block and oven, or upgrade the temperature measurement system.

 

[danw2]

Assuming you are looking for a temperature regulated 'chamber' that holds the 'cold end' thermocouple terminal block connections at 45°C because that's the fixed value CJ used by whatever device, then I suspect, along with Hacksaw, that you're in the DIY (do it yourself) realm.

If I were looking, I'd go to the lab and grab one of those thick heavy 2000 page lab catalogs and look up ovens or incubators. I'd look for one that advertised flat line control, as opposed to on-off control to minimize CJ error, since any CJ error is a hit on the T/C reading. I'd check the oven temp reading against a certified lab grade calibration thermometer to zero in on the target temperature. It would be handy to be able to see the lab grade thermometer.

But going through all that to monitor 50 temp points? Why not place and avoid the hassle?

 

[IRstuff]

Seems to me, the lack of such systems indicates that thermocouples are not been used to the accuracy that you seem to want. This is, of course, borne out by the fact that PRTs and the like are used in calibration systems.

While you could stuff the cold junction into an oven, that seems a bit brute force, but the basic concept would still be valid. If you are going to roll your own solution, you might consider a closed-loop thermoelectric temperature controller and glue your cold junction to that. Add in a bunch of thermal mass, allow it to stabilize, then start measuring.

Alternately, you might just consider "really" upgrading, and use a PRT or thermistor as a replacement.

You've not even indicated what accuracy you're looking for, or whether that's really necessary for your application. After all, your system presumably went "decades" without a stabilizated cold junction, so why is it suddenly necessary now?

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize