Two RX-102A-CD sensors deviating from standard curve

maclean
New Member

Posts: 2

Two RX-102A-CD sensors deviating from standard curve Nov 14, 2022 17:58:31 GMT -5

Quote

Post by maclean on Nov 14, 2022 17:58:31 GMT -5

Hi,

We have two Interchangeable Rox sensors, U09632 and U09633, that we've been using to measure sample temperatures in a dilution refrigerator (DR). The two sensors are classed as "uncalibrated", which I think I understand to mean "roughly follows a standard resistance-temperature curve to within +-10mK at 50mK". (this is confusing wording on the lakeshore website and documentation!) The curve we've been using is available for download from www.lakeshore.com/resources/sensors, under "Standard curves --> (Rox) RX-102A (.zip)"

However, we've been confused by the readings we're getting from these sensors. In our current configuration, they are directly adjacent to a built-in thermometer in the DR, so that we can compare their reported temperature to a sensor that we take as trustworthy. We have two separate types of cryogenic thermometry in our lab: a custom readout that we typically use for all cryogenic temperature measurements, and a Stanford Research Systems (SRS) AC Resistance Bridge. We've been measuring the sensors with both readouts, and getting similarly confusing results.

For example:

- The DR sensor reports 50mK

- with our lab readout, the two rox sensors report 25kohm = 85mK. The bias is constant current of 5nA with an AC modulation at 17Hz.

- with the SRS bridge, the rox sensors report 29 kohm = 75mK, both when using a constant voltage bias of 30uV or when using a constant current bias of 30uV/100kohm=0.3nA.

The sensor is mounted on the cryostat stage with a stainless steel bolt torqued to 6inch-lbs, with belleville washers, on a clean gold-plated surface. The thermometry wires are clamped to the cryostat cold stage, broken out and heatsunk on a 800mK stage, and exit the cryostat via a 4K wire feedthru to warmer stages.

Any insights? Is this not the correct calibration curve to be using?

Lake Shore Jeff M
Lake Shore

Posts: 327

Two RX-102A-CD sensors deviating from standard curve Nov 15, 2022 8:11:12 GMT -5

Quote

Post by Lake Shore Jeff M on Nov 15, 2022 8:11:12 GMT -5

Hello Maclean, I have asked two of my sensor experts to look at what you are reporting and let me know their thoughts. As soon as I hear from them, I will enter another response.

In regards to the correct curve to use, you are using the one for this application. In addition, your understanding of the +/- 10mK accuracy at 50mK is also correct.

One thing that we have seen in the past is that the sensor in the Bluefors DL has a filter installed that blocks induction of a RF signal that can cause self-heating. Our sensor does not include this filter so if you are in an environment that has potential RF induction, that can be what is causing the higher readings.

Check back later to see if my sensor experts have any other suggestions.

maclean
New Member

Posts: 2

Two RX-102A-CD sensors deviating from standard curve Nov 16, 2022 16:05:22 GMT -5

Quote

Post by maclean on Nov 16, 2022 16:05:22 GMT -5

Hmm, that's very intriguing but I'm not sure I understand the implications. Could you say a bit more about the mechanism, or is there a resource somewhere I could refer to?

Would having a filter present anywhere in the thermometry circuit resolve this, or would it need to be near to the thermometer itself? I'm imagining that the induced currents would be present in the whole wire loop, which runs out to the warm readout systems. We do currently have some filtering present at the in/output to the cryostat wire feedthrough. Generally the work we do in the cryostat is with RF signals, so it's very possible that they're present in the environment.