USB DAQ, Laptops, thermocouples and ground loops...

[psmorris]

Hi all, I hope someone out there can help me...

I'm trying to use a USB-6211 multifunction DAQ device to measure outputs of several devices, including 4 thermocouples. However, I have a severe noise problem. My hardware set up is as follows:

Thermocouple wire 1 --> USB-DAQ AI0
Thermocouple wire 2 --> USB-DAQ AI8

USB-DAQ connected via usb to laptop which is running of a standard (non-grounded) power supply.

In this configuration, if I use labview (DAQmx) to read the output of the thermocouple, I get nothing but ridiculously high levels of noise (Amplitude in excess of 10^7!) - I've set up the DAQmx for a J type thermocouple and have the cjc set to the approximate room temperature where I'm making the measurement. Strangely, the noise improves significantly if I connect AIGND on the DAQ device to AI8 (the second wire of the thermocouple). But the amplitude of the noise is still about 20-30 degrees. However, if I then disconnect the power supply from my laptop (and run it off battery), the noise decreases further and the signal is useable (noise amplitude ~+/- 1 degree).

I need this system to run for several hours, possibly days, so there is no way I can run the laptop from the battery... can anyone suggest anything I may be doing wrong, or which may help to solve the problem? Might it help if my laptop PSU had a three pin input? (I'm in the UK, but the mains connector which goes into the power brick is a simple 2-pin one, not 3-pin...)

Any suggestions would be greatfully recieved!

Thanks

Paul Morris

[pallen]

Hello Paul,

I'm no expert when it comes to thermocouples and ground loops.  But just from reading your post it makes me wonder a few things.

How close are your thermocouple wires running to laptop's power supply?  

Those supplies are switchers (usually) and I know that mine has gotten so noisy that I can actually hear it when it's under load.  It would be easy enough to extend the DC side of brick's cable and relocate the brick as far away as possible.  

If that doesn't do it for you, you could try an isolation transformer.  But typically laptops don't have a ground prong.  So I'm not sure if you'd be seeing an actual "ground loop".  But it could be possible.

Good luck.

[psmorris]

Thanks for the reply,

I'm wondering if its not something to do with a "floating" differential measurement (please excuse my lack of clarity, electronics is not something I have a huge deal of experience with!).

If I connect the AI ground pin from the DAQ card to ground, the problem appears to disappear... (although I need to test it more thoroughly yet!)

So, hopefully "problem solved" although I'm not sure if I understand why the laptop power supply influences the issue...

thanks again!

[pallen]

Sounds exactly like your signal was floating.

Hopefully you've got it solved now and can get to the fun part.