How to Program Unique AI Scanning Sequence

[Jason S.]

I an using an encoder as an external trigger to collect analog data. I'd like to do the following sequence, using the index pulse on the encoder as a trigger, and the A channel as the start trigger for each sample:

180 samples from channel 1
180 samples from channel 2
180 samples from channel 3
180 samples from channel 4
wait for next index pulse and repeat

I cannot program this all into the AI Config buffer because the buffer only holds 512 points. And, I don't want to sample them in scans (1-2-3-4, 1-2-3-4, etc.). I can't sample them in scans and throw away the data I don't need, because the system would not be able to keep up (the sampling rate would be 4x what is required). The time between channels is insuffic
ient to reprogram the configuration for the next channel.

How can I set this up (NI-DAQ or similar, please)?

thanks-
Jason

[Ben]

To properly answer your question we need to know;

1) What development environment are you using?
2) What hardware are you using?
3) How fast are you trying to sample.

Ben

[Jason S.]

Ben,

1. I'm actually using the E-series register level programming manual to write this myself for xPC Target (I know this isn't supported, but the programming isn't a problem). So, if you can tell me how to do it in NI-DAQ on say, Windows 98, I could duplicate that here.

2. 6025E
3. up to 144kS/s, most likely around 72kS/s

I'm thinking about using the external multiplexer setting. Any comments?

thanks-
Jason

[aasenjo]

Jason,

Based on the restrictions imposed on the application:

1) Limited buffer size (memory limitations?)
2) Can't scan all channels and discard data (system limitations)
3) Can't reconfigure (Time constraints)

I can't think of any option other than an external mutiplexer.

We have one that will give you the 1 to 4 ratio, the AMUX-64 ( http://www.ni.com/pdf/products/us/2sc478-481e.pdf). It is controlled by the digital lines on the E Series board. Controlling the AMUX manually through the digital lines is not something that we directly support but if you are doing RLP you can probably figure it out really easily. Here's a KB that might help:

http://zone.ni.com/devzone/conceptd.nsf/webmain/38C2B8A17C7768AA86256802007B8C16?OpenDocument

KB Online - 2
07G055U might also be useful.

The other option could be one of our SCXI multiplexing modules, but since you are doing RLP you would also have to learn how to program the switches through registry access.

I hope this helps,

Alejandro Asenjo
Applications Engineer
National Instruments

[Ben]

HI Jason,

I may not be able to fully answer your question but I may be able to help get another question together that you can post to both the MIO and other lists.

As you know, there a number of steps that have to be caried out to do a a single triggerd I/O. It may be possible to open multiple references to the I/O device and go throughthe steps of configuring all of the indiviuals triggered I/O you want. Do everything as normal but stop short of actually starting the I/O.

The idea would be to start the first I/O and as soon as it comples, grab the data from it and fire off the second I/O. Continue till done. He idea is that most of the time invloved with starting an I/O is messing withthe config. If you do it all ahead of time, the latency of
you system may let you do what you are trying.

Like I said, not a complete answer, but maybe helpfull.

Ben