Why would temperature reading drift on an SCXI-1122

[RicK.]

I have an RTD probe connected thru an SCXI-1122. Over the course of about 30 seconds the reading drifts over a range of about 5 dC. What would cause such a drift and how can it be corrected? The LabVIEW program accesses the data at a .5 hz rate and averages the last 5 readings.

[Filipe A.]

Hello;

What might be happening there is that when you do SCXI measurements the CJC channel is read only once, at the beggining of the code execution. If the temperature changes on the terminal block, since the CJC is not constantly compensating that temperature drift, you might eventually get some erroneous readings.

You might try to restart the acquisition periodically to have the CJC channel recompensating the teperature drift.

Hope this helps.
Filipe A.
Applications Engineer
National Instruments

[AVR]

What is the nature of this drift? Is it sinusoidal? Try acquisition over a long period (5 minutes) and look at the nature of the drift. RTDs and thermocouple are notorious for their ability to pick up 60 Hz and 120 Hz noise. Is the RTD attached to a device that is "plugged in" or is the RTD in free air? The difference is significant as the RTD mounted on a grounded object/surface would actually be grounded and the free air sensor will be a floating source. The most common source of electrical noise comes from the improper referencing/mixing of floating and gounded inputs.

For example, in our test facilites we will have RTDs and TCs mounted on test components to measure surface temperatures (these are grounded) and other TCs to measure airflow temperature (these are
floating). If these TCs are brought to the same SCXI module, regardless of the grounded/floating selection for each channel and the use of resistor packs, ALL of the measurements are affected by 60 Hz noise and other nasty ground-loop effects (errors of 10-50 deg F). If we bring the grounded signals to one SCXI-1100 and the floating to a separate SCXI-1100, the noise level is below the quoted acuracy of the TCs.