Thermocouple readings fluctuating JType

[sonotk]

When acquiring temperature using NI 9213 of J type thermocouplethe temperature readings are fluctuating. Attach is the screen shot of the temperature showing the fluctation. The actual temperature is between 123 to 125 deg C and it drops after an interval to around 63 deg C. The width of the drop is approximately the same.

Some options tried were

1. Connecting the negative terminal of thermocouple to chassis ground. This did not help and the temperature dropped to almost half.

2. Connecting the COM to chassis ground did not help either.

3. All the wires and connections were tested and when testing using handheld thremometer directly the readings does not fluctuate.

Any suggestions as what might be causing this?

TIA.

[Henrik_Volkers]

Using my uncalibrated cristal ball:

I see some EMC problems...

Is the drop in sync with the heater? Or in sync to something else?

The drop with the grounded terminal is an indicator...

If you make some readings with 1.2kSPS .. how does the signal look like?

What are you monitoring, how long is the TC and style? (cable length, ultra fast thin TC? or shielded?)

HOW does the overall grounding looks like?

[sonotk]

Hello Henrik,

The thermocouples are monitoring temperature in different ovens. There were three channels tested and the test results were the same.The drop happens at the same interval across the channels.

The two channels have J Type with signal range set from 140 to 190 deg C in Max, the actual reading for these two is around 125 deg C. The thrid channel has S type where the signal range is set to 25 deg to 1275 deg C. This also reads around 1000 deg C and then drops to 750 - 800 deg C for the same short duration.All three channels drop at the same time and recover.

The DAQ chassis 9184 itself is grounded and that is the only ground I know of.

Thermocouples are pretty long and run across the room before connecting to DAQ. The wires come out to a board and there is another wire from that board going to DAQ module. Shielded wires are not used.

Thanks.

[Henrik_Volkers]

I still vote for (a grounding loop /isolation fault at) one of the oven heaters as the source for the drop. (check the current consumption of the heaters ,  look for correlations, ovens often have packet driven heating shemes)

If you disconnect all channels and shortcut the inputs, all channels should read CJ temp.

Check each channel individual, then add one after another. and look for the drop down.

If you connect the TC to a scope input , the noise can be a hint too.