Platform:
LabVIEW 7.1 for Windows
cFP 2020
cFP TC 120
cFP AIO 600
other cFP modules
I am getting some strange behavior out of our cFP TC-120 and cFP AIO-600 modules. We are using LabVIEW to control a rapid thermal anneal machine here at the university. The main method of gathering a temperature is to use a pyrometer that is sighted on the wafer within the chamber. When we removed the outdated control system on the machine, with the purpose of updating it with LabVIEW, we ran into a problem with the old pyrometer. Long story short, we had to replace it.
When we acquired a new pyrometer, we knew that we had to find some way to independently calibrate it. We decided, since we had an extra input on our thermocouple module, to use a thermocouple wafer. We are getting some very strange behavior from the TC module. The program I wrote to calibrate the pyrometer is saving readings from the TC through the TC module, and from the pyrometer through a current input on the AIO module. While we expected some lag between the two, we have found that the TC temperature output is stair-stepping very badly. There seems to be as much as a second between samples from the TC module. While the overall track of the TC seems to be accurate, this is a very puzzling result, and makes calibration more difficult. Is this the normal function of the module? I would expect temperature samples many more times a second. Also, the overall temperature track of the TC seems to be leading that of the pyrometer, which is opposite of what we expected. Although, there is a propreitary display device that the pyrometer hooks up to between it and the cFP, so that could account for the delay.
Anyway, I have attached a spreadsheet showing the behavior of the TC wafer and module vs that of the pyrometer. The program samples every 50 msec. Looking at the TC column, you can see a new temperature appears only every second or so. I suspect that the problem might be because the temperature output resolution is low. That or the Express VI outputs in LabVIEW can't handle precise enough data. Is this kind of behavior normal for the TC 120?