I have been dealing with PT100 calibration for a while and normally I have used this sensor at max 40 degree but 60 and 70 degrees are adjusted for calibration and these deviation is very high about 2-3 degree. Does this issue affect my normal measurements.
Hello Elgin.
I would rather suggest that you adopt a two point calibration method - first point at ZERO degrees and the other at HUNDRED degrees. Repeat this procedure for 4-5 times and then check the stability of your system against other calibrated or reference temp. Indicator. If the problem persists dig into your front-end electronics. Best luck.
Eery
Hello Dear Sir,
Thank you very much for precious values. I checked my electronical system and other sensor, Everything is ok. Normally, My refrigerant outlet sensor is calibrated at -20, 0, 10, 30, 70 degree by independent company and I checked it my fluke calibrator and then adjusted -10,20,40,50,60 degree with fluke calibrator. in spite of calibrating all values ( for 50, 60, 70 deviation is about 2,5 degree) my deviation is 0,23 degree for 10 degree. I think that this deviation is based on limit value. I have been using this sensor at max 40 degree. Maybe, when I remove the deviation of 50,60,70 , my measured value diverges.
Ergin,
If you have a well-calibrated sensor and you are also cross-checking it with your Fluke calibrator, then it is certainly the problem with your signal-conditioning electronics.
Also, I as have mentioned the TWO-POINT calibration method (example of 0-100degrees), since your lower end is well below ZERO, you can also calibrate it to say -20 degrees (or below if possible) and then at the higher value of 100. This should minimize any errors in your calibration slope.
Thanks and best luck!
The deviations you describe seems higher than one would expect from this type of RTD. Without knowing more details of your system and procedures it is difficult to guess what can be improved. It would be useful to see your calibration data (independent temperature vs measured PT100 resistance). Does your calibration correction fit a curve to the data or does it interpolate from tabulated data? Is you Fluke temperature measurements of the sensor generally higher or lower than the displayed readings? Do you have a 2, 3 or 4 wire sensor (not that this should make a big difference, but still)? How do you determine the deviation between your sensor displayed reading and the actual system temperature?
Dear Crause,
The sensor calibration was performed by independent caompany for each sensor at 5 temperature values. Then, I interpolated the intermediate values. For instance, they measured -20, 0, 10, 30 ,70 degree and I have adjusted -10, 20, 40, 50 and 60 degree for each sensor even I have checked it with Fluke calibrator at each value and my reading at calibrator was very good but I have been observing 2 degree deviation at a sensor. I don't exactly know that the problem is based on whether about electronic or sensor. Thanks
Ergin,
I have a doubt about your sensor (PT100)excitation and probably the front-end electronics. Sometimes, the circuits are in good agreement with the theory but in practice unforeseen parameters creep in. One major culprit might be the very small offset that might be getting amplified with your full-scale or higher temperature readings causing your slope to shift away from the standard calibration curve.
This is just one suggestion. However, you know better what your approach is!
Thanks!
- Badges
- Science topic
- Similar topics
- Methodology
- Sensors