I'd like to set up a DAQ scheme with a retriggered pulse train sample clock that does not continuously read samples, but instead synchronously reads the next N bursts of samples at arbitrary times. More specifically, my setup is as follows. My analog input is an integrated photodiode signal. The end of the integration is signaled by a falling TTL edge. The repetition rate is 500 Hz. A few times a second I command and then wait for a translation stage to move. Following each move I'd like to tell the DAQ to read M (~ 10) fast samples after the next N (~ 3) falling TTL edges.
I know how to set things up to do the reading continuously, and I think that I can implement the arbitrary reading by allowing the AI task to overwrite unread samples. However, when I perform the read I want to know for sure that it does not begin in the middle of a sample clock pulse train. I've considered using the falling TTL edge as an AI start trigger in addition to it's use as the CO start trigger, but I'm concerned that various delays might cause the AI to start late and miss samples.