LabVIEW
Does anyone know how to prevent data from jumping from low to high?
@GRCK5000 wrote:
How do you prevent this jumping from happening. I thought about using "In range and coerce" block. I just don't think it will work or maybe I am not using it correctly (See VI).
I have no idea why you ever showing us such a silly VI. The code makes absolutely no sense (independent of the problem you are trying to solve).
Yes, your problem could be hardware or software. How do you generate the ramp of values and how does it relate to the instrument reading? It could be that you are formatting the message sent to the instrument incorrectly or that you parsing the response incorrectly. Still, I am with the others that you might have faulty hardware or things connected, shielded, terminated, etc. incorrectly. What kind of relay temperature are you measuring? (classic coil, SSR?)? Is the relay idle or operating during measurement? An operating mechanical relay will create transient fields that could induce extra voltages in the thermocouple? Thermocouple voltages aver very low, so it would not take much!
A good thing to confirm would be if your "spikes" are coming from your signal, or somewhere else.
Make the "smallest" code you can and record some data. If you can, use MAX and log your channel data directly.
Log Measured Data with Task Created in MAX
If you still see the same "noise" in your signal, then at least you know it isn't being created by some other means within your code.
Unexpected Noise or Fluctuations from Thermocouple Measurement
Ground loops can cause "noise" to appear in temperature measurements. If the noise is only there during the heating or "on" cycles, then you could very likely have a ground loop.
I've had success with USB Ground Loop Isolators. Basically a small device that sits between your PC and your USB DAQ device.
A quick but not ideal solution is to increase your number of samples and then implement averaging or filtering.
I've had a USB-TC MCC device many many years ago that had threading issues- it was NOT thread-safe. If you ran a LabVIEW program that was using the provided VI's at the same time you ran the MCC-provided test panels you'd start getting bogus data. I can't recall the specific nature of the bad data, but I got confirmation from their engineers that it was indeed NOT thread-safe. Once I closed the provided test panel the issues went away.
