1769-IT6, inaccurate readings, CJC removed

Anyone know a fix for this?
1769-IT6, 2 modules both have same problem.
Setup for K tc, 1x engineering units data, degrees F, 60Hz filter.
Problem---with CJC thermistors in place on TBs the readings are about 50 degrees above actual. With thermistors removed the readings are correct +_ a degree or so.
I would like to put the thermistors back on for better accuracy, but......Any ideas?

Check polarity of thermocouple leads.

Red is always negative.

https://www.google.com/search?q=the...9j69i57j0l6.6768j0j8&sourceid=chrome&ie=UTF-8

I'm stumped. The reported temperature should be significantly less than actual hot junction temperature., not more than, when the CJ is disabled or missing.

CJ temperature is the difference between the terminal block and the freezing point of water. If the terminal block is 72 Deg F, then the CJ temp is 40 Deg F (72°-32°=40°). The CJ value is added to the mV conversion value to get the reported temperature value.

Without any CJ measurement, the reported temperature would be 32°F if the thermouple were isothermal, - same temperature at both ends in a quiescent state.

Reversed polarity of the thermocouple will drive the reported temperature in the wrong direction, so reversed polarity with a slightly elevated hot junction temperature and no CJ would be less than 32°F.

The CJ error detection circuit for the 1769-IT6 uses 77°F as a default CJ,



To report a temperature value 60°(presumably F) high, I wonder whether the module is using the open-circuit-on-powerup-77° value, plus a measured value when the CJ thermistor is added after powerup.

Have you power cycled the board with CJ's installed to see if then reads correctly?

This assumes you've installed the exact model CJ thermistors needed for this module, not something you found in the junk bin.

Well danw you may be on to it. While the thermistors are definitely oem, we are not certain they are the thermistors that came with the cards in play. Do you know if AB produced different thermistors for the IT6?

I haven't a clue. I'm an instrument guy who's been involved with thermometry using thermocouples for decades and have seen lots of issues with thermocouple inputs because of CJ's; I just copied the statement from the A-B 1765-IT6 manual


I don't know what A-B has used for external CJ thermistors or whether the CJ's are common across different modules. Either someone else here might know of you'll have to contact A-B.


But if I were you, I'd definitely try a cold start to see if the installed CJ's work properly from an initialization form a cold start state.


Dan

Are you sure you have the correct compensating cable installed?

If the installed cable corresponds to a thermocouple that generates more mV than type K and the sensor head is hotter than the AB module, an error similar to the one you describe would occur.

Good call, lfe. J wire will do it.

I power cycled and re-enabled each channel manually with the thermistors connected and read +50 C. Power cycle and re-enabled each channel manually with therms disconnected get correct readings. Extension cables verified type K,yellow-red.

Anyone know the resistance of the thermistors at room temp?

Have you inadvertently at some time done a auto calibration with the CJC's disconnected ? Try an auto calibration with CJC's connected

1. Rockwell part number 1769-CJC is only for the universal inputs embedded on an L2 CompactLogix CPU. It is sepcifically mentioned that it doesn't work with the 1769-IT6

2. Part number 5300004-04 by Spectrum Controls is the part number for CJC on the 1769-IT6.


I can't find any info about either. Perhaps it isn't even a thermistor, but an IC?


Let us know what the solution is in the end.

Tried autocal all ways possible (as autocal occurs on enabling a channel, according to my interpretation of literature).

Thanks for the part number. Will follow up with fix, assuming I get there.

Not all CJC's are created equal. If the 1769-IT6 is calling for a specific CJC that's the one you must use. Removing it, in some cases, will "deploy" a default value which can get you pretty close to accurate depending on the setup but the accuracy can't be depended upon. Install the correct CJC and re-calibrate the inputs (if doable, I'm not familiar with the IT6) and that should get you back on track. If not you may be wanting to call Rockwell.

gofastnow said:

Extension cables verified type K,yellow-red.

Click to expand...

Yellow as + and Red as - ?

If the CJC is of the wrong type, then anything is possible. Just because the CJC is a 2wire device does not make it a thermistor. Even if it is, then it would have to be the correct type. I know so IO I have used over the years used for CJC:
PT100
PT1000
Thermistor (not published type)
IC

Just a daft thought,


Is the thermistor the correct type? i.e.
PTC (positive temperature coefficient) or a NTC (negative temperature coefficient)


PTC increases resistance with increased temperature, NTC decreases in resistance with increased temperature.


Steve