Strange perodic fluctuations in voltage measurement

[ninvoice]

Hi Bill,

This is an interesting idea. Although it has been there since it came up, the magnitude of the fluctuations did change from time to time, sometimes ~1 psi, sometimes ~30 psi. This could be checked also if I take the pressure sensors to my labmate's computer, which is located in a different room.

yanxiang

[Michael_78]

Of your spike is very quick, the magnitude you see is likely to do with the exact time the sample is taken. I.e. you are aliasing the interference. The lab noise idea is good. You have a lot of ideas to go at now, hope you find a solution.

[billko]

Oh,sometimes if wires are running close together and aren't shielded and one has pulses, you may see those pulses on the other.

[Tirmizi]

try to make proper grounding of your measurement system (sensor plus your system chassis etc ) . there should be no potential diff. presence of pot diff cause creation of ground loops which corrupt the measurment data and canot be removed via filtering unless proper grounding is done.

[Tirmizi]

check this link also

http://forums.ni.com/t5/LabVIEW/Voltage-spikes-using-USB-6009/td-p/1522550

[ninvoice]

I have done some other tests. This time I tried to isolate the components. So obviously, with three 1.5v batteries in series as the power supply, the output singal of a single presuure sensor is fairly stable.However, the power supply box has no intrinsic fluctuations, or at least not showing up as big ones on the graph. But once the pressure sensor is hooked up, the fluctuations showed up, and the more sensors are connected to the same power supply, the higher the fluctuations are. Now I noticed each contributes to about 5 psi max fluctuations. That is why I could get ~30 psi of fluctuation when more are connected. The frequency of the spikes is fairly stable and is about 1/1.6s. Not sure if this number is associated with the sample rate. The wires may potentially be problematic as well.

yanxiang

[billko]

Good luck!  Problems like this are sometimes very difficult to track down.

[Michael_78]

More transducers = more cable length. Look at proper shiedling and grounding techniques, use differential inputs. Can you post a picture of your connnections with cable details? Try 4 wire connection right to the transducer maybe.