Unwanted pre-samples using a start trigger using NI-9239

[Karen_Isenor]

Hello,

I am using an NI-9239 inside of an NI-9188XT Chassis. I am providing an external digital TTL pulse to trigger data acquisition via the BNC PFI0 line. To test my triggering, I have set up a simple program which generates an Analogue Voltage task, sets the timing to Finite Samples with an appropriate sampling rate, provides the trigger source and rising edge, reads/displays the data, and finally stops and clears the tasks. I provide an analogue sine wave via a signal generator to one of my channels, and the trigger line to a second channel.

After I start my program, I manually raise the external trigger from low to high. The data acquisition is triggered, and both the sine wave and trigger data is displayed - however, in every test case I can see ~40 samples *before* the trigger (regardless of sample rate, always 40 samples).

See the attached images.

I can work around this by always sampling an extra 40 samples, and clipping the first 40 samples, however I feel this is a) bad practice and b) may be susceptible to subtle changes. I'd like to know why these samples are recorded, and how to disable them OR verify that it is a constant, and any potential sources that could change it.

[Kevin_Price · Accepted solution]

This is due to the delay characteristics of the delta-sigma converter used by the NI-9239.  There will be info in the manual that describes and characterizes it, but you cannot get rid of it.  Here's a link to some more info, much more can be found with more searching.

Based on a little prior experience with other delta-sigma devices, the delay will be dominated by a fixed # samples but there will be a small additional delay that may vary with sample rate.  Again, the manual will provide details now that you know to seek them out.

-Kevin P