I try to calibrate the temperature of FP-TC-120. I found that at low temperature of 30 oC, the difference between FP-TC-120 reading and actual temp( I got from thermometer) is about 3-4 C. However, with the increase of the temperature, the differece of the FP-TC-120 and actual temp increase. And the difference can goes up to over 10 C when the actual temperature is 70 C. I feel a little strange. Do I need to make a calibration curve of actual temperature versus FP-TC-120 reading. I worry about how I can get the precise temperature reading based on FP-TC-120 reading. Furthermore, when I connect different channel of FP-TC-120 with different K type thermocouples(of different dimeters), the reading show big differences up to 10 C. Thanks a lot.
The errors that you are seeing are a bit beyond what would be expected and should be able to be lessened but are not as large as you may be thinking they are. First of all, let's try lessening the errors you are seeing. There are several things that you should try. Make sure the FP-TC-120 is properly configured for the thermocouple type that you are using. This is done by going into FieldPoint Explorer and right-clicking on the FP-TC-120 @X and choosing Edit This Device, then by clicking on the Channel Configuration button. In the channel configuration window, you can specify what thermocouple type is attached to each channel. Make sure that you have specified K Type. Then, in the upper left corner, activate the drop box that says Type 1: Analog Input and change it to Type II: Analog Input. This allows you to modify the CJC sensor. Make sure that the CJ Source parameter is set to Internal (Internal tells the FP-TC-120 module to use the current measured temperature from the CJC sensor to cold-junction compensate the readings, alternatively you can set it to assume the temperature is always 0 or 25 C.
Next, let's make sure you are hooking things up as ideally as possible. If you have multiple FieldPoint modules on the bank, located the FP-TC-120 module between the modules that generate the least heat (this can be determined by looking at the power consumption of each module and choosing modules with the least power consumption).
Another thing to consider is thermocouple length and thermocouple diameter. Thermocouples will transmit heat along their wires, this heat (or cold) will change the CJ temperature for the particular terminal. Since the FP-TC-120 compensates for the temperature of all terminals with a single sensor, not 8 sensors, a single thermocouple that has a lot of heat transfer will have an increase in the error on the channel. To mitigate this, use thinner diameter thermocouple wire.
Finally, remember that K type thermocouples themselves are only accurate to +/- 2.2 C. The FP-TC-120, mounted in ambient temperature conditions between 15 and 35 C, measuring temperatures between 0 and 750 C, has a worst case accuracy spec of 0.8 C, so between the thermocouple and module, you can have 3 C worth of error.
We have a system where we use several thermocouples and we have individual calibration tables for each sensor...and I am in teh process of replacing all teh hardware with cFP and cFP-TC-125 modules, and I was trying to locate some ideas on howto include this calibration feture aswell...I'm thinking a table read from a file(a few known calibration points) for each TC-125 module (8 sensors) and then maybe fitting a 3rd or higher order polynom and then compensating the values with that...is anyone could provide some help on doing this in LV it would be great!