04-19-2015 11:59 AM
04-24-2015 11:54 AM - edited 04-24-2015 11:55 AM
25 MHz is a very large sample rate for a counter... can you elaborate on your application? There really shouldn't be any reason to sample the counter this quickly. If you're trying to timestamp your external pulses with the highest possible precision, I would instead use the external pulse as a sample clock and count the internal timebase (assuming the external pulse isn't also 25 MHz...). Doing this would actually get you 10 ns resolution (instead of 40 ns that you are hoping for with a 25 MHz sample clock) and the required throughput would be based on the rate of your external signal (which is presumably less than 25 MHz, which is the highest frequency that you can connect through the PFI lines anyway...)
To answer your question though...
The on-board FIFO is 127 samples per counter. If your task is measuring more samples than this, you would need to be able to read data off of the device at least as fast as it is coming in. On USB that is going to be highly system dependant (one benchmark here reported 8 MHz). If you really need the higher throughput you are asking for you will probably want to switch to the PCIe version of the 6341 (20 MHz on a single channel mentioned here, I'm not sure if higher rates were tested but I wouldn't be surprised if 25 MHz is sustainable as well).
Best Regards,