USB-TC01

[Tishmo]

I'm using multiple USB-TC01 devices with K-type thermocouples.  I'm seeing occassional periods of oscilation of about 0.5 degree C with about 3-4 cycles per hour.  It comes and goes at different times of day and for varying periods, but seems to be synchronized across all of the channels.  When one channel does it, they all do it.  The attached plot shows four of the channels, but I have 14 total connected.  Has anyone seen this behavior and know how to get rid of it?

[M.J]

Hi Tishmo,

Have you tried using a different computer to see if this issue is persistent? If possible can you try on a computer running Windows XP? Since this device utilizes software timing it is controlled by the clock of the operating system which can cause measurment discrepencies.

[Tishmo]

I haven't tried another computer, but the one I'm using is running XP.  I was wondering if there was a common reference since it's across all of the channels.  I should have also mentioned that the lead lengths are significant, around 120 to 150 feet.  The thermocouple leads are run between two buildings through a conduit, so I thought I might be seeing localized heating effects, but there are no junctions, it's a single continuous run from the sensor to the junction.  There is one more odd thing about the data which is that the high termperatures clearly show up during the hot time of day but the low temperatures seem to clip instead of showing a typical diurnal cycle.

[cade]

Hello,

Looking at the specs of the USB-TC01, it seems that he accuracy of the the device is +/- 1 Deg C which is greater than the oscillation you are seeing.

http://www.ni.com/pdf/manuals/372997a.pdf

Therefore, oscillations might not hold much value based on the way you are measuring them.  Aside from that, the noise could be coming from a variety of sources, especially considering that the cables are that long and exposed to outside conditions without being twisted pair.  Becasue of this, there may not be a way to get rid of this noise you are seeing.

You can try to use a more accurate measurement device but that probably won't fix the issue as there is surely noise introduced into your signal.