18kHz limit

[Jason Lawrence]

I am using DIAdem to process a waveform signal that I need to sample at 40kHz (in order to generate a FFT frequency domain to 20kHz).  I can adjust acquisition and save rate in the configuration file we're using to configure the .vi channels (written to .tdms in groups).  Even when I set 30-41.5kHz (acq. and save rate), the time step associated with my waveform channel (when disassociated through ANALYSIS-->CHANNEL FUNCTIONS-->"NUMERIC CHANNELS<->WAVEFORM CHANNELS") I get 18kHz and no more.  For comparison, when my save rate is 1500/s, I get a frequency domain of 750 on my FFT and a time step of 6.666E^-4seconds (5.555E^-5 in the 18kHz case).  The signal in question goes through a 9205 card (slot 6 of a cDAQ-9188 chassis) that is then wirelessly transmitted to our ethernet hub.  The 9205 supports 250,000S/s, so that should not be limiting us to 18kHz.  9188 clock routing (80MHz/20MHz/100kHz)?  What gives?  Any ideas?

[Brad_Turpin]

Hi Jason,

What happens if you run the FFT in LabVIEW before you save the data to disk?  What's the highest frequency the LabVIEW FFT returns an amplitude for?

Brad Turpin

DIAdem Product Support Engineer

National Instruments

[Jason Lawrence]

Hmm.  Possible.  I can look into that.  Is there a 9kHz limit on DIAdem then?

[Perry_S]

Hi Jason,

If you are sampling at 18kHz, the maximum bin that you could have is 9 kHz. It is a limit of the dataset that you have acquired.

Regards,

Perry S. 

[Jason Lawrence]

Perry,

I'm aware of that, thank you.  That's why I'm trying to boost the sample rate to 40k+, but I keep getting a maximum frequency domain on my FFT of 9k (and when I disassociate the time channel I get 18kHz).  I could have explained the situation better.

[Jason Lawrence]

I'm thinking more and more this must be a 9188 clock issue.  How do I work around that?  Move modules?  It claims GHz bandwidth, but perhaps it's defaulting to the slowest clock speed (100kHz base clock?).  Is this appropriate for a different forum?  Can I move this issue there?

[Perry_S]

Hi Jason,

I seem to be misunderstanding what you are trying to do. Could you post one of your data samples for me to take a look at so that I could get a clearer picture of what you are doing? Thanks.

Regards,

Perry S.