[HSDIO/PXI-6552] Using generated waveform as external clock?

[dkpham2007]

Hi,

I am using PXI-6552 with HSDIO driver to do some digital waveform generation and acquisition.

May I use generated waveform (that is output to...let's say...DIO 0) as external sample clock (i.e I connect DIO 0 back to STROBE and configure STROBE as the sample clock) as I would like to control the acquisition timing bit-by-bit? I've tried to do so but it failed.

[Mark_SAI]

Hello,

Have you got a VI with your attempts?

Thanks

[dkpham2007]

Thanks for your quick response, Mark! I'll post my vi when i'm back to work in 2 days time

[Ryan_M]

dkpham,

There is a restriction that the clock must be "freerunning".  What that means is that your "clock" (and it needs to be an actual toggle signal) needs to be toggling before you commit your acquisition session and as you close it.  If the clock stops between them, you run the risk of a hardware clocking error. 

[dkpham2007]

Thanks Ryan for your reply. I've indeed generated a clock signal that runned "forever" on DIO 0 before the acquisition session.

[dkpham2007]

Hi,

Sorry for my late response. Please see attached vi.

External pin connections are as follows:

DIO 0 (emulate clock signal) is connected to STROBE/PFI 5

DIO 1 (acquired waveform) is connect to DIO 9

DIO 2 (start trigger) is connect to PFI 1

In the vi, Generation session is runned first to generate all needed waveforms. Acquisition session is then runned with sample clock being STROBE. It fails!

[Ryan_M]

dkpham,

You're attached code does not have generation running first.  Since LabVIEW is a data flow programmaing language, you need to make sure that you have some sort of dependency from generation that gates the acquisition code.  For example, you can wire the error cluster from the generation INITIATE up to the acquisition INIT SESSION.  In this way, you'll be guaranteed that the generation will start producing the clock before the acquisition sessions needs to have a clock.

[dkpham2007]

Ryan,

Thanks for your reply. Actually I am aware of this issue so in this case I deliberately runned the generation first and checked all the waveforms by oscilloscope beofore running the acquisition part. However i still don't know why it doesn't work.

[Ryan_M]

What error are you receiving?  Does it work (though with bad data) if you use OnBoard clock for your acquisition? 

[dkpham2007]

I receive a time-out error on the read-waveform block. And YES, it works with OnBoard clock. I also works with an external clock generated by a function generator

Message Edited by dkpham2007 on 07-31-2009 04:13 PM