6120 external sample clock

[istarnes]

We have been successfully using the 6052e for some time now to acquire our radar�s analog output. This output consists of a data signal synced with a 20 Hz analog trigger that also comes from the radar. This has been a nice easy solution but now we have increased our frame rate from 20Hz to 60Hz. Quantitatively, the old system was taking 20Hz X 9,984 samples = 200 KS/s, while the new system takes 60Hz X 9,984 samples = 600 KS/s. The 6052e no longer could handle the increase so we purchased the 6120.

The problem we have now is that 6120 does not seem to support the same acquisition technique that we used with the 6052e.

Here is what we are doing with the 6052e:
We use the 6052e to generate a pulse train with frequency correspondin
g to our sample rate, which is gated by a signal that turns on when the trigger pulse occurs and stays on for a fixed number of samples. We then route that pulse train back into the board as our external sample clock. The problem seems to be that the 6120 does not support the external sample clock.

The reason we use this technique with the 6052e is to guarantee that we get the same number of samples between triggers.

Ian Starnes
Sr. Software Engineer
NIITEK, Inc.

[Bill B]

Ian,

I am going to guess that you are experiencing this problem because our S-Series boards such as your 6120 use a pipelined FIFO ADC process. Specifically, our S-Series boards do not send the last 3 samples of an acquisition when using an external clock. The last three samples remain in the FIFO on the card. Therefore, if you begin your acquisition, receive a trigger pulse, then receive 9,984 clock edges on the sample clock, and then read all the samples in the buffer, you will only receive 9,981 samples. Consequently, if you set the AI Read.vi to read 9,984 samples it may timeout because it is still waiting for three more samples to be acquired. This would create the allusion that the external clock was not working with the 6120.

The good news is that i
t is easy to solve this issue. You simple need to read 3 less samples the first time AI Read is called. Therefore, the first time you call AI Read you need to read 9,981 samples and then the next time you call AI Read you will continue to read 9,984 samples. This will ensure that you will get the same number of samples between triggers except for the first time which will be short three samples. However, with this method, the first three sample of each trigger will be the last three samples from the previous trigger signal.

Please see this KB for additional information on how to ensure that you get the proper amount of data for each trigger signal that you receive:
http://digital.ni.com/public.nsf/websearch/D64CD277A6B739A186256A73007E7BCC?OpenDocument

Please let me know if you are still having difficulty or if the information I provided did not help resolve the issue.

Regards,
Bill B
Applications Engineer
National Instruments