olga
New Member
Posts: 3
|
Post by olga on Nov 25, 2019 10:10:38 GMT -5
Hello,
I am using Cernox 1030 calibrated sensors in two different packages: CD and SD. there are 2 SD sensors and 1 CD sensor placed into the same copper bloc very close to each other (less then 1 cm distance). Special attention was paid to minimize thermal conduction through the wires.
I observe the noticeable difference in the temperature given by the SD and CD sensors. Two SD sensors give the same temperature within the calibration uncertainty, but the CD sensor gives significantly different value. At 295 K the difference (CD-SD) is - 333mK, at 4K the difference (CD-SD) is -20mK. I joint here an image with values.
What could be the reason of this?
- using 84 calibration points and linear extrapolation between, instead of calibration coefficients and analytical formulae?
- calibration not longer valid (calibration of two SD sensors made in dec 2018, calibration of CD sensor made in nov 2016 (pas sensor was never used))? - connections (2 pins with 4 wires for SD, and direct 4 wire for CD)? - different mechanical strains (SD sensors fixed with a drop of varnish1073, CD sensor is mechanically fixed between two copper plates)? - for 4K the measured voltage was quite high (around 40mV instead of 2mV as for calibration), but for 295K the voltage was OK, but the difference CD-SD was also ![]() ![]() measured.
Thanks in advance
  |
|
|
|
Post by Lake Shore Jeff M on Nov 26, 2019 13:41:44 GMT -5
It is not possible to say why you are seeing the discrepancy between the SD packaged sensors and the CD package sensor without having the CD Package sensor returned for a calibration verification.
There are several things that can cause the difference in the readings as follows:
1. The CD Package is not getting proper thermal transfer
a. Did you apply thermal grease to the package and use the proper torque specification?
2. The instrumentation is not in calibration
a. What instrumentation did you use for your measurement?
b. If using a multiple channel instrument, try exchanging one of the SD channels with the CD sensor
and see if the issue follows the sensor or stays on the channel
3. The CD Package sensor calibration is not correct
a. This would require that the sensor be returned to Lake Shore for analysis and recalibration if
necessary.
|
|
olga
New Member
Posts: 3
|
Post by olga on Feb 21, 2020 9:37:28 GMT -5
Dear Jeff,
thank you for your detailed answer.
Here is the answers on your questions:
1. a. I am sure of the proper thermal transfer (grease was applied, sensors properly fixed). 2 a.b: Different Keithley 2002 multimeter were used, different channels of the Keithley scanner were tested, but I always observe the same difference: CD-SD = -330mK at 295K. 3. I have done some extra measurements with another calibrated CD sensor,
and I observed the same difference SD and CD sensors (-330mK at 295K).
In the calibration certificate it is mentioned that the excitation of 2mV was applied.
Does it mean that during the calibration you apply a 2mV voltage to the sensor, and you read the current?
I my experiment I use the current source and I read the voltage. (The value of the current is adjusted to keep the voltage around 2mV). Is it a correct way to do ?
Thank you in advance.
|
|
|
|
Post by Lake Shore Jeff M on Feb 24, 2020 14:48:33 GMT -5
Yes, when we calibrate the sensors we try to maintain a 2mV or less signal across the sensor, however, when they are used we recommend a 10mV excitation voltage be maintained for temperatures greater than 1K. The reason we use 2mV is due to the fact that we have a large number of sensors being calibrated at the same time so we try to eliminate any possibility of heat build up due to the large number of sensors.
I do have one more question on your arrangement. Please describe what you mean when you say you have a 2 pin with 4 wires for the SD sensors.
|
|
