Noise spikes in waveform generated by DAQ

[viswa]

Hello all,

                 I am trying to generate few millivolt amplitude square waveform say 5mV at 12.5Hz, i see some spikes with high voltages suddenly appears in the waveform. This is actually causing a lot of problem in uniformity of the waveform. I precisely need to generate very low amplitudes, so i thought of using an RC circuit external to the board. 

I am using USB 6229 board with LV8.2. How can i generate very low voltage waveform with minimal noise??? Any help would be greatly appreciated.

An oscilloscope image of the waveform is attached.

Maximum frequency needed would be around 6kHz.

[Wayne.C]

viswa,

I don't think that spike is noise.  I think you have a problem with your vi.  Post it so someone can look it over.

[Philip.k]

Hello Viswa,

Could you tell me which version of the DAQmx driver it is that you are using. Could you try installing the latest version available on are website and let me know if you still see the issue.

Here is a Link to the latest version

http://joule.ni.com/nidu/cds/view/p/id/1614/lang/en

Regards, 

[ChrisR.]

This looks like DAC glitches to me.  http://digital.ni.com/public.nsf/allkb/9B13E25B0B0EC197862562B4007D2E33

Probably the best thing to do is to generate a much bigger signal with the board's analog output and attenuate it with a resistive divider to the level that you need.  For example, instead of generating a 5mV signal, generate a 5V signal and pass it through a 10k/10Ω divider.  You'll have much less noise this way, and your glitches will be 1000x smaller. 

[cirrusio]

I would agree with Chris R. that this looks a lot like glitching.  As is stated in the article Chris provided a link to, glitching is most problematic when the MSB switches at the midrange of the code (i.e. around 0 - where yours is occurring).  Have you considered DC-coupling with an offset that would remove the signal from this midrange?  Also, given that the glitch energy for this device is 100 mV at 2.6 us, you might reconsider your choice of DAQ and use one with a more reasonable output range.

Cheers, Matt 

[viswa]

Hi Matt,

                     Thanks for the reply. I considered generating 5V waveform and attenuating it with a potentiometer. When i see potentiometer output on oscilloscope seems too noisy and i felt that it would be better off with direct output of DAQ board. I feed the waveform to laser control which changes wavelength very precisely in few pico meters. I could see some instability in laser power and wavelength in an interferometer but was not sure of it.

Well my project is nearly finishing, so i do not have money to replace USB 6229 with some other board. I have to look for alternatives. 

Thanks everybody for your replies!!!

Message Edited by viswa on 05-14-2010 05:18 AM