Frequency operation of PXI 6723 Card

[chrisj247]

Hi,

I have recently purchased some NIPXI cards for the PXI1033 chassis.

My issue is in relation to the PXI6723 Card that we purchased for analog output generation, in particular, the generation of frequencies by the card. From the test panel of the card in MAX, I can see that the generation of the signals is heavily reliant on the Update rate for the card and the buffer size used.

These somewhat decide on the frequency generated so i was wondering if you could point me to a tutorial or documentation on analog generation for that card. Obviously, if i have a standard Waveform generator i turn the frequency knob and thats it. But for these programmable cards there is a lot more control over the frequency, such as the sample clock, update rate, buffer size, cycles per buffer etc etc. Its all new to me in Labview so i am just looking for some guidance as to how to produce these signals.

Secondly, the PXI6723 is said to produce 42kS/s on all 32 channels but this didnt specify how many samples were make up the waveform and restrictions on using acquistion cards with synchronisation necessities. Can the card produce 32 channels of quasi-pure sine wave signals up to 42kS/s and be synchronised to the aquisition card (PXI4462) clock? (In a seperate issue we figured out that the 4462 sample clock could be exported to the 6723 to synchronise the devices - however one would have thought the PXI_Clk10 would have been the choice here but we couldnt get it working - this may still be up for change also)



Many thanks,
Chris

[LeeM]

Hello Chris,

  
     I would recommend having a look at the examples in the LabVIEW example finder to start with. In particular i would look at 'Cont Gen Wfm-Ext Clk.vi' This is a basic example that allows you to output a waveform through the card and source an external clock source to do this. I would recommend exporting the sample clock that you wish to use from the 4462 on one of the RTSI lines and then setting this as the sample clock source for the Analog output card.

Another good example to look at is the 'Cont Gen Voltage Wfm-Int Clk non regeneration.vi' This does not regenerate the same waveform in the buffer but instead creates a new waveform to output after the previous one has been written to the buffer.

It should be noted that there may be some gliltching if you are updating the waveform. This is because the buffer will already have a waveform in there to output before you input the new one so a new frequency may not have an exact number periods that fit into the pre-defined buffer. If you are looking to generate dynamic waveforms then the best option for the best results would be one of our arbitrary waveform generators here.

Good Luck,

LeeM
NIUK

[chrisj247]

Thanks for the post Lee. I have had a look at these examples and the method you mention with the sample clock and the RTSI line from the 4462 to the 6723, is what i am using and it appears to work, insofar as it keeps a fixed phase throughout the run with AO and AI.


However my main concern now is with the capabilities of the analog generation card. I think i now understand that the card can produce 32 channels at 45kHz on a PENIII machine say (as per spec sheet). However, for each sine waveform we would need for comfort about 100 samples. If i wish to produce a maximum of 20kHz sine wave this would require (20k*100points = 2Msamples)? I dont think the card can achieve this. It is a bit frustrating since we explained the simple application to a LabView sales/technician who came on site here and he assured me that the card could achieve what we needed.

(I need an 32 channel analog output card that can produce 20kHz sine wave comfortably. I thought this would have been simple enough since it is just function generation.) Perhaps i am missing something in my VI  but i think it may be the card i was sold. Any suggestions?

Thanks for your time,
Chris

[LeeM]

Hi Chris,

   Looking at the specification on the card you are using:

http://sine.ni.com/nips/cds/view/p/lang/en/nid/12551

It appears that the card should be capable of doing 45KS/s on each of the 32 channels. Is it possible for you to post your code so that i can see how it is architected?

Thanks,

LeeM
NIUK