Salvage a type K TC probe reading from a type J meter

[abl33]

I've run into an issue where a part was produced with a TC that was specified to be type J, but was delivered as Type K. It is impossible to replace this TC without replacing the entire parts which is something that we really want to avoid. My problem is that this part will go into a vessel that only has a type-J passthrough plug and type-J wiring back to the PLC that is reading the Thermocouple and I do not have the ability to reprogram the PLC.

I do not need a precision of greater than +/- 2C. In trying to work around this problem, is it possible for me to infer the voltage of the actual TC using a Type J and Type K thermocouple reference chart. Take the temperature reading output shown and find what that mV value is and then take that mV value and find what that value actually translates into on the Type K chart to know the correct actual temperature (or reasonably close to it).

I assume the two bimetallic junctions at the TC plug when in connects from type K wire to type J wire could induce an error, but I'm hoping that it will cancel out on both poles assuming they are both close to the same temperature. If they wouldn't be balanced, then is there any other way to salvage this or is it a complete loss?

 

[IRstuff]

What's doing the cold junction compensation and how?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[IRstuff]

What temperature range are you trying to measure?

It may be possible to perform a two-point calibration of the thermocouple, just to avoid having to deal with all the dissimilar junctions between the thermocouple junction and the data acquisition module.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[abl33]

Not a bad idea. I'm concerned about a couple key temperature points. 125C and 40C being the two most critical.

I might be able to take a meter setup for type J, connect a type J extension to it and plug in a type K thermocouple to that all within a temperature controlled space and observe those points.

The more I was thinking about the junction issue, the less I thought that would actually work given there are actually four different metals between the thermocouple and the extension plug.

We are on a critical time crunch and saving time with a simple conversion chart was very attractive. I think we are most likely to just scrap the part and re-produce and just eat the loss to have a chance at our deadline.

edit:
cold junction compensation is within the PLC IO card and I do not have access to modify that.

 

[IRstuff]

We've calibrated our own type Ts using an ice bath, but we had everything else programmed correctly. Adding a boiling point would allow you to calculate the error and compensate in post-processing

see:

https://www.foodrepublic.com/2011/04/04/how-to-calibrate-your-food-thermometer/

One tip w.r.t. freezing point; we used a thermos bottle and regular drinking water/ice; the error is relatively minor.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[hacksaw]

Get the supplier to send you a replacement, but use the type K temporarily.

 

[abl33]

It is not physically possible to replace the TC in that part, so a replacement TC in this situation will not do us any good. The part is used as temperature monitoring sample for processing other similar parts and it is not possible to change any of the physical items I mentioned above. The only way to salvage it would be with logic based on the "bad" reading we currently get and a conversion of that to an accurate one. That being said, I don't think its possible because we will effective have a chromel–alumel junction at the temperature point I care about and a chromel–iron junction, and an alumel-constantan junction at a temperature I don't care about but that I can know what it is with another closely located thermocouple to that point.

At this point, we scrapped that part and built a new one with the correct TC inside. The delay in delivery of the actual product is not acceptable and so we just had to eat the cost of making a 2nd part. In the meantime, if there is a conversion method, I'll have this part to play around with it to verify if we can do it in the future if this ever happens again.

 

[hacksaw]

If you used a smart transmitter it is a user selection or had access to the PLC coding, you designate the T/C type.

If you use an independent measurement you can directly compare the type K with your actual service temp, good luck

 

[danw2]

> given there are actually four different metals between the thermocouple and the extension plug.

Yup. It is a custom, multiple thermocouple source and an operating range characterization of the hot junction would be valid only under the assumption that the intermediate junctions are held at a constant temperature at whatever temperature the characterization occurs. If the intermediate junction(s) (singular or plural) temperature shift(s) then the characterization curve is out the window.

 

[abl33]

>Yup. It is a custom, multiple thermocouple source and an operating range characterization of the hot junction would be valid only under the assumption that the intermediate junctions are held at a constant temperature at whatever temperature the characterization occurs. If the intermediate junction(s) (singular or plural) temperature shift(s) then the characterization curve is out the window.

yes, unfortunately it would be. This is a TC that is embedded within a part that is run within an autoclave, so we are very restricted as to what we can physically do and for sure no "smart transmitter" will survive the autoclave process without some extreme protections of its own. At this point, I assume I could in theory take a similar setup and run it in a normal oven and map the specific points I would care about but for now I don't think its worth figuring it out given I also think it likely that the mV value on these three junctions will end up outside of the valid TC sensing range and the PLC will not return any value at all and the likelihood of this happening again is small.