Post by js27 on Nov 4, 2023 16:01:57 GMT -5
Hello,
We bought two uncalibrated RX-102B-RS thermometers and are having a difficult time calibrating them. We are calibrating them by attaching them to the base plate of our Bluefors dilution refrigerator and using the calibrated fridge thermometer as a temperature standard. This is of course not the perfect standard but we believe it has a precision/accuracy level we are satisfied with (10 +/- 1 mK).
We have installed the thermometers according to the instructions here: www.lakeshore.com/docs/default-source/default-document-library/rx-rs-installation-instructions.pdf?sfvrsn=1f61ed72_4 . In particular, we were careful to heat sink the ruox leads with GE varnish on a (normal, not OFHC) copper bobbin and we applied a little apiezon N to the thermometer and bobbin surfaces in contact with the dilution refrigerator. The thermometers were clamped with a torque of roughly 0.3 Nm. We also use a room temperature ferrite EMI filter (https://www.l-com.com/d-sub-ferrite-filter-emi-adapter-db25-male-female) since we do not have a cryogenic EMI filter currently. The temperature sensors are read out with 4 wire measurement using a Model 372 through the PhBr twisted pairs in our fridge which are anchored at each temperature stage.
I have attached our calibration curves below. Blue and red are the curves for our RX-102B-RS thermometers and in green is the R vs T curve for the Bluefors RuOx, which I was told they acquired from Lakeshore also. We are unable to explain why the resistance values diverge so significantly below ~0.1K.
We do not believe EMI filtering is the sole limiting factor at these high temperatures (we tried this with and without our RT filter and it didn't make much of a difference). Consequently, something might be wrong with our thermalisation. In a previous run, we had clamped the thermometers to 1Nm and gotten much higher resistances (~8 kOhm) on our ruoxes when the fridge reached base.
One other phenomenon that we are observing is that when we heat up our mixing chamber plate and let it cool back down and replot this curve, there seems to be some hysteresis effect as shown in the picture below:
We are unsure why this is happening, and in particular the directionality where the warmup curve is lower than the cooldown curve. Any input on what we may be doing wrong with our installation
Finally, I was wondering how valid is the "typical" curve on the RX-102B-RS product page for an uncalibrated thermometer?
Thank you very much for your help,
Joe
We bought two uncalibrated RX-102B-RS thermometers and are having a difficult time calibrating them. We are calibrating them by attaching them to the base plate of our Bluefors dilution refrigerator and using the calibrated fridge thermometer as a temperature standard. This is of course not the perfect standard but we believe it has a precision/accuracy level we are satisfied with (10 +/- 1 mK).
We have installed the thermometers according to the instructions here: www.lakeshore.com/docs/default-source/default-document-library/rx-rs-installation-instructions.pdf?sfvrsn=1f61ed72_4 . In particular, we were careful to heat sink the ruox leads with GE varnish on a (normal, not OFHC) copper bobbin and we applied a little apiezon N to the thermometer and bobbin surfaces in contact with the dilution refrigerator. The thermometers were clamped with a torque of roughly 0.3 Nm. We also use a room temperature ferrite EMI filter (https://www.l-com.com/d-sub-ferrite-filter-emi-adapter-db25-male-female) since we do not have a cryogenic EMI filter currently. The temperature sensors are read out with 4 wire measurement using a Model 372 through the PhBr twisted pairs in our fridge which are anchored at each temperature stage.
I have attached our calibration curves below. Blue and red are the curves for our RX-102B-RS thermometers and in green is the R vs T curve for the Bluefors RuOx, which I was told they acquired from Lakeshore also. We are unable to explain why the resistance values diverge so significantly below ~0.1K.
We do not believe EMI filtering is the sole limiting factor at these high temperatures (we tried this with and without our RT filter and it didn't make much of a difference). Consequently, something might be wrong with our thermalisation. In a previous run, we had clamped the thermometers to 1Nm and gotten much higher resistances (~8 kOhm) on our ruoxes when the fridge reached base.
One other phenomenon that we are observing is that when we heat up our mixing chamber plate and let it cool back down and replot this curve, there seems to be some hysteresis effect as shown in the picture below:
We are unsure why this is happening, and in particular the directionality where the warmup curve is lower than the cooldown curve. Any input on what we may be doing wrong with our installation
Finally, I was wondering how valid is the "typical" curve on the RX-102B-RS product page for an uncalibrated thermometer?
Thank you very much for your help,
Joe