are there any possibilities that Labview can cause low frequency spikes on captured data?

[cstudy]

Hello All,

I used Labview 9.0 to capture data from an ultrasound displacement sensors. The sampling rate of the sensor is about 12 Hz. The results data randomly have very low spikes frequencies of about 0.3Hz to 0.7Hz. Where are the sources of the spikes? Is it from the sensor? From power supply? From LabView?

Very happy if you can give me any clues.

Thanks

[aeastet]

When using ultra sonic devices I find that distance, angle and surface (how hard, soft, etc...) are the biggest contributor to noise. If you get a sensor and use it on the outside of it range it will be nosier than if you use the low end. The closer the better for these devices.

Do you have data so we can see it?

[cstudy]

Thanks Aeastet. Here is the data captured. You can see some of the spikes at the end of the waveform. Labview is sampling at 20 scan/second.

Thanks

[aeastet]

I have seen this type of noise before. Are you working in a noisey environment? Also can you sample faster and filter the data or average the data to help with this? I would try to move the sensor closer to the measurement surface to help with noise as well.

[Hummer1]

Rather than averaging, you might pick the median value...(the one closest to the average...) noise will produce some outlier values that influence the average some, but by picking the median you are choosing a value close to the average that may not have been influenced by your noise environment....

Good Luck.

[cstudy]

Hello,

The purpose of this test is to find down where is the possible source of the noise spikes. Software massage is another issues. Attached is another waveform captured from the same device but the LabView scan rate is 1000 scans/second. The lab is not that noisy.

Thanks

[Hummer1]

If you "zoom in" on the data, does it look like the bad data point occurs at the same "distance" every time...It looked like in the first data set that the spike occured a little before reaching zero and a little after crossing zero...could you have a bad A/D bit....?

[cstudy]

Hi Hummer1,

To address the DSP system, I did change the sensor by a signal generator. The waveform captured is OK in that case. The distance between the spikes (that you were talking about) is about 0.75sec .

Thanks

[NIquist]

If a signal generator signal is clean then it must be your sensor.  Either it is glitching (can you replace it?), or it's detecting random noise even though you say the environment is quiet.  Are there any devices in the vicinity that might be cycling on and off (compressors for instance) creating noise spikes?

[johnsold]

Two things to try:

1. Look at the signal with an oscilloscope to see if you see the same interference.

2. Speed up the sampling rate.  If the interfering signal is higher in frequency than half the sampling rate, aliasing will occur.  Aliasing is a very effective way to confuse the observer about what is really going on.

What kind of data acquisition device are you using?  Are the inputs properly configured (differential, single ended, grounded, ...)?

Lynn