Equipment noise

[Thibault_Prevost]

Hi,

I have a force sensor connected to an SC-2350 board via a feedtrough module (SCC-FT01). The board is connected to a laptop through a DAQCard-6036E and receives its power directly from the laptop.
I am faced with very strange signal fluctuations - up to 10-15V spikes - that seem to occur in a random fashion. These signal fluctuations do not seem to come from the force sensor since: 1) everything looks fine whenever the sensor is connected to an oscilloscope 2) the sensor itself cannot physically output a signal larger than 5V. It therefore seems to me that the spikes themselves are artefacts coming from the board, the DAQ card, the laptop or Labview... This is quite frustrating. Any suggestion on where the problem could come from?
Attached are the typical normal (top figure) and abnormal (bottom figure) signals I get:

Normal noise level



Spike trains


Thibault.

[AnalogKid2DigitalMan]

The frequency of the spikes is 60Hz- AC power line frequency noise possibly?

Is the laptop plugged in when the spikes occur? Do spikes occur when the laptop is running just off of battery power?

[Thibault_Prevost]

Yes, whenever the laptop gets plugged, the spikes take over everything. But the slight problem is it also occurs when the laptop is on its battery, although more randomly but frequently enough to make any measurement almost impossible.

Thibault.

[AnalogKid2DigitalMan]

Are there any electrically noisy devices (pumps, motors, etc) devices in the vicinity of your setup? Is your sensor wiring near any AC power wires, how long is the sensor wiring, is it shielded?

Is differential input the correct setup for your application?

Many questions, since intermittant noise issues can be a bugger to troubleshoot!

-AK2DM

[Thibault_Prevost]

There is no other noisy device in the vicinity but I moved the system to another room - quiet office with no AC power line in the immediate vicinity - to see if there would be any difference: the noise was still there (!).
I also tried to switch to RSE and NRSE inputs but I got either no signal or something that seemed completely decorrelated with the expected the sensor output.
Also, all the wires are shielded and the multi-wire cable connected to the sensor is about 3 feet long. I moved it around while acquiring the signal with Labview but I couldn't see any correlation between the noise occurences and the wire movements... I don't know if it helps but the kind of noise I get looks also occasionally like this:

This looks really weird to me. I'm under the impression that the problem might come from the SC-board, the DAQ card or the laptop because whenever I measure the force signal with an oscilloscope using the same wiring and under the same conditions, I get no such spikes.

Thibault.

[Chris_D]

Thibault,

Here is a excellent article on the NI Developer Zone about noise considerations and noise troubleshooting.

Field Wiring and Noise Considerations for Analog Signals
http://zone.ni.com/devzone/cda/tut/p/id/3344

[AnalogKid2DigitalMan]

Thibault,

The last graph you posted almost looks like a transient electrical connection occuring and allowing power line noise to be induced. Just a hunch on my part.

I would check every electrical connection- wires, connection points, plug in modules (remove and reinsert), etc. with the power off.

-AK2DM

[Thibault_Prevost]

I double-checked the wiring, regrounded everything and switched the terminal configuration for my sensor to RSE: it seems to work fine now.
Thanks again!

Thibault.