Intermittent noise in pressure transducer voltage signal

[MikeStebbins]

Hello,

I am currently using LabVIEW 9.0.1 along with a USB-6212 DAQ to control an experiment that includes measuring the pressure of a water-based hydraulic system using an Omega PX209 pressure transducer (0-200 psi, 0-5V output).  The included graphs show pressure vs. time for what should be a pressure ramp-up and then constant pressure hold test. My question is specifically about the noise in the signal that builds after some time of data collection (or rarely, from the beginning of collection, as shown in graph 3), not the loss of pressure over time.

The wiring set-up and pressure transducer data sheet are included.  During the tests, nothing changes in the environment that I've noticed that would explain the noise (i.e., no additional electrical activity near the transducer/wiring/DAQ, nothing moved, etc.)  The transducer shielding is soldered to a wire connected to ground.  The DAQ is housed inside a grounded metal project box, which is several feet away from another metal project box that houses the power supply.  The wiring from transducer to DAQ is approximately 3 feet long.  The included .png of the code (A, B, and C represent different states of the state machines) shows the differential measurement technique I'm using.

Questions:

Does the wiring into the DAQ look correct?

Can you think of any reason why the signal would get noisy after time or sometimes start and continue noisy?

If it's a shielding problem, what other shielding should I do?

Does this look like a possible ground loop issue?

I'm still fairly new to LabVIEW, so I would gladly take any suggestions on my code, as well.  Thank you very much for your help.

Mike

Omega PX209, 0-200 psi, 0-5V output pressure transducer data sheet

[Henrik_Volkers]

In such a situation I try to get more information about the noise.

Can you make a zoom of the noise? Is it hum, is it spiky ?

Try to make a measurement with a higher (and highest possible) samplerate. Fetch 100ms , 

(You can use the DAQmx Testpanel, your software or (my favorite) a fast analog scope)

A 'nice' source of noise are inverter driven pumps or switched power supplies.

[MikeStebbins]

Henrik, thank you for your response.  I was able to zoom in on the first of the 3 pressure graphs in the OP, see included picture.  I zoomed in on two locations; 1 in the lower-noise region and another in the higher-noise region (A and B, respectively).  Unfortunately, I can not collect any more data with the transducer at this point, as an Omega application engineer asked me to send it in for inspection so it is gone.  As I mentioned in the OP, I wasn't able to hold a constant pressure with the transducer and the engineer suspects that I might have a leak in my transducer.  He said that this might also explain the increased noise that occurs later on in the experiment.

I do have a relatively inexpensive switching power supply a few feet away from both the pressure transducer and the DAQ, inside of a separate metal project box.  When you say 'nice', do you mean a 'strong supplier of' or an 'easy to deal with' source of electrical noise?  Or both?  I'm a mechanical engineer, so while I tried to study and shield the different components appropriately, I certainly wouldn't claim to know everything about the subject. 

Thanks for your help, Mike

[Henrik_Volkers]

As far as I can see you sampled the sensor with 1kS/s. Enough to see that this is not line hum . The noise could be of a much higher frequency. We will see if you get back your transducer.

Switching power supplies are can , if badly constructed / designed , not (enough) filtered, be a really bad source of noise/interference.  I said 'nice' because they are mostly easy to identify, but sometimes hard to avoid. (Just had a eletronic ballast of a light source , that went mad and radiated in the MHz range).

[MikeStebbins]

Thanks for the explanation, Henrik. I'll update this post when I know what the situation with my pressure transducer is. 

Mike

[VSolorzano]

Surely by now Mike has solved the pressure transducer problem. Just for the record I had the same problem with an Omega pressure transducer and as Henrik suggest, it was caused by a switched power supply (TDK Lambda).