choice of filter

[devchander]

Told you so

Pleasing the boss is a super human feat

still benco, did you check out the resolution of your card and its accuracy??

your readings may still be correct and adhering to the accuracy level promised by your DAQ card

[benco]

Thank you maximus

it's very nice

 i send you data and a picture

[benco]

thanks devchandler

It may be the problem

but how can I check the DAQ-card?

Then I try to make my own filter ( with shift register )

it quiet hard but I will make it

[Martin Fierz]

Hi benco,

I think you are solving the problem from the wrong end. Of course you can use filtering in LabView to reduce the noise. But this kind of noise shouldn't be there in the first place! When I see this kind of behavior in a system where I'm trying to make a measurement, I usually take a closer look at the signal with an oscilloscope, and then I call in my electronics technician to help me solve it if I can't solve it myself.

Filtering bad signals is better than doing nothing, but getting good signals is much better!

cheers

  Martin

[devchander]

---

but how can I check the DAQ-card?

Which DAQ are you using??

Are you using differential or single ended connection??

First, check your AI acquisition with a constant and accurate DC source and see what accuracy you are able to achieve

Comparing these readings with those on an Oscilloscope as Martin suggests is a good approach

[winterishere8]

Benco

Please take a look at the attached picture. If this looks ok, I will mail you the source code or the DLL. Used my smoothing which is basically very similar to moving average with 3 window and 3 passes.

For your noise, I would check the cables, I've faced similar problems with cabling.
Also, I would make sure your test setup is grounded.

Hope this helps.

[winterishere8]

I've included a picture with median filter (rank 5). I think this might work better for you.

[benco]

Thanks maximus

The smoothing seems good but I think not so "smooth" for me. but can you send me the VI, I can show him the result.

I think I just ask a electronics technician as Martin said (I don't know any electronics technician but I ask my boss).

I would keep you informed (But not before Wednesday or Thursday because I don't go at the institute before)

Thanks all of you

Ben