Analogue measurement get strange interference

[DevAutoit]

Hi

I'm using a cDAQ-9184 with at analogue input card "9201" which has 8 channels for signal and 1 pin for common grnd signal.

I'm trying to measure two pressure sensors which give a value of 0.5V when there is no pressure, and one channel which measures a torque-signal,  1V = 100Nm.

When measuring the signals without moving the test piece with a servo-motor I will have what to expect. (see picture below).

I've added all 8 channels so you can see my problem.

So channels 0-2 are connected, 3-7 are not connected.

The strange thing happens when the servo-motor is on, it doesn't matter what RPM the servo is running at. 

In my voltage-measurements I will get strong electrical disturbances. Which I cannot understand 😃

If the servo is causing this via the GND, shouldn't the disturbance show up at the same time on all channels.

See below screenshot

As you can see above, the disturbance show up on each channel about once every 8.5 seconds or so.

When I only look at one channel. The disturbance will show every 8.5 seconds.

Isn't this very strange?

Anyone has any ideas?

I've tried to run the NI-card on a totally separate power-supply. But if the error is in the GND signal, this wouldn't help...  So, still the above disturbances.

Have you had any similar problems?

/Mats Eriksson (programmer, not measurement expert 🙂 )

[JonasM]

Hi Mats,

As you say, if this would be a grounding issue you would probably see the disturbance at the same time for each signal read. Althouhg it is a multiplexed module so it would be a small gap between them. But not 8.5 seconds.

Anyhow the module you have is not appropriate for measurement of grounded signals. Below you can read about ground loops and returns.

http://www.ni.com/white-paper/3394/en

Is it possible that you have a module with possibility to measure differential or one with NRSE measurement? This way we can eliminate the ground loops to see if this is related to ground.

[fredericb]

Hi there,

That's interesting. It might come from the power supply itself.

Try grounding all channels you don't use and ground one of those to to a power supply ground.

Also it wouldn't hurt to replace the DAQ with an oscilloscope and see if the same happens if so you can say whether the noise comes from your instrument or from your DAQ.

Also out of curiousity have you tried to measure the frequency of the noise? if that frequency band is of not of interest for your measurement you can easily filter it.

Another silly mistake easily done (i did it last week) is if the cables from your instrument are badly connected to the DAQ i suggest take them out and put them back in.

That's all i can think of

Frederic.

[DevAutoit]

So,

we changed to a 9205 module and got rid of most of the interference from the servo-motors.

A small 5V transfomer connected near the cDAQ device caused some other interferences 🙂 .... So we changed to a more expensive transformer and voilá, almost totally clean signals... ripples of +- 10mV instead of 1-2V 🙂