NI-9213 Low-Frequency Oscillation in High-Speed Mode

[Gubenator]

I am collecting thermocouple data with the NI-9213 in two thermocouple types, N and K. When I use the high-speed mode, I see a very low-frequency (0.02 Hz) oscillation with the thermocouple readings, at sampling frequencies from 50 Hz to 2 Hz. This oscillation does not appear when using the high-resolution mode, or in the PT100 sensors collected with the NI-9216 module. I've made the usual checks, the thermocouples are not grounded, grounding the shielding on the thermocouple wire, external electrical sources have been turned off, different slots have been tried etc. The oscillations are not seen when the probes are connected to a PXIe system. 

I am using a NI cDAQ-9178 chassis with NI-9205, NI-9213 and NI-9216 modules, with data collection with LabVIEW 2018. 

Is there information on what the difference between how data is collected in the high-frequency and high-resolution modes is, or on how this issue could be resolved? 

Screenshot shows high-speed mode with a 20 sample hamming filter applied, and high-resolution unfiltered data taken within a minute of each other. 

[Henrik_Volkers]

I don't know the SigmaDelta ADC that is used in this card. Usually in 'high speed' mode the internal filters are changed. I would asume some down mixing effects .

0.02 Hz ... some power line frequency (or higher hamonics) mixing?  

[Gubenator]

Thanks for this, do you know where details on the ADC could be found?