pci 6229 ai/ao clock

[ultracold]

Hi,
Does any one know how to use the digital output from the same card as the clock of the analog channels internally?
I am actually not sure whether it's possible or not.
The card is pci-6229.

thanks

[Brian F.]

Hello ultracold. 

Thank you for posting to the NI discussion forums. 

There isn't a way to route a digital output line internally to the clock of the analog channels.  However, if you wish to use an external clock as your sample clock, I would recommend instead using a counter which can be easily routed to the analog input clock.  The other option is to connect the digital output to a PFI line externally and designate that line to act as the clock. 

Thus, the basic question that should be cleared up to choose the best solution for your needs is:  Why do you want to use a digital output to act as the clock of your analog channels?  Is there a reason why using a counter would be undesirable?

Let us know and we will be happy to help you further!

Brian F
Applications Engineer
National Instruments

[ultracold]

Hi Brian,
Thank you and I am sorry I didn't explain this clearly. I am actually trying to generate some long time (>100 second) analog arbitrary waveforms with 2 micro second resolution.  The amount of  analog data is thus very large and my computer really have trouble to handle it. But actually the waveforms don't have many edges. What in my mind is to use the digital channel to generate an arbitrary waveform with all the timing information of the analog edges. If I use this signal as the clock, the analog data will be cut to almost nothing. Since the digital output accepts compressed data, it's not too much data either.  Do you have better ideas to handle large amount of analog data?

I don't think the counter/timer can generate an arbitrary waveform, can it?
I also tried  using a cable to connect the digital output to a PFI line as the analog cable. It actually worked, but with some delay. This makes it hard to synchronize with other digital events from the same card.  But if there are no other ways, I will have to live with it.

Regards

[Kevin_Price]

Quickly chiming in...

No, the counter timers CANNOT generate an arbitrary pulse train waveform.  Buffered counter output has been HIGH on my wishlist for many years.

NI makes arbitrary waveform generators that should be capable of such things.  Haven't used them myself.

Another nice option comes from one of NI's Alliance partners, Viewpoint Systems.  They make a board called the PCI-DIO-64 which lets you define pairs of timestamps and digital port bits.  In output mode, it updates the port only on the timestamps you buffer.  If changes are rare, your DAQ output buffer can be very small.  (It works analogously for input, timestamping digital state changes with high timing precision but without redundant samples). 

I used their legacy version of the board that ran in an ISA slot to generate custom stepper motor motion profiles.  It's good stuff.

-Kevin P.

[ultracold]

Thanks for the reply.
I guess the question I am asking is how to handle large amount of analog data. The waveform I need is
so long, if data are organized point by point, my PC will run out of memory soon.

Since the analog waveform I need only has several changes but with long delays between changes, I think there should be
a better way to handle them.

Best

[Brian F.]

Hello ultracold. 

As I read your post, I think you may need some type of analog triggering to aid in your acquisition.  As I understand your application, you are gathering data at a fairly fast rate for a long period of time, which can be fairly memory intensive.  Most of this data is not important as it isn't around one of the 'changes' that you mention.  Is this accurate?

If so, you can use analog triggering that only triggers when the analog input that you are acquiring rises above a certain threshold value.  The specific VI that can accomplish this is the DAQmx Start Trigger vi, with the selection of Start>>Analog Edge on the pull down menu directly below the VI. 

If my first statements are correct, I would recommend giving this analog trigger a try.  If it is inaccurate, please post more specifics so we can decide how to tackle this issue in the most effective way.

Have a great day!

Brian F

Applications Engineer

National Instruments