Synchronizing AO, AI 4461

[dksys2000@yahoo.com]

Hi All, Well I am using a Pci 4461 I am Synchronizing ai0, ao0 and ao1. I am outputting two sine waves and capturing them. I am Rotating the phase, then I average them so ao0 and ao1 should cancel. It does this most of the time, but every 3^rd to 100^th cycle it gets off but about 3 data points. This will ruins the test

Any ideas.

Thans

Dan

[dsb@NI]

Are you summing the two signals in hardware and acquiring with ai0, or are you using both input channels and averaging them in software? Can you simulate the desired behavior in LabVIEW? If you can provide more details (sample rates, test frequency, rate that the phase is being rotated, etc.) and even attach your VI, the community can probably give more guidance.

[dksys2000@yahoo.com]

I have 2 outputs (ao0 and ao1) and 1 input ai0 sync using Multi-Function-Synch AI-AO-Ext Dig Trigger.vi that was changed a little. I have varied all the sample rates, test frequency, rate that the phase is being rotated. It still happens every now and then.

Sample rates from 44.1k to 200k

Test frequencies from 500 Hz to 40 kHz

Phase rotation 180 2 times

45 8 times

You can see the rotation when you zoom in at the start and when it glitches you see that too. It shifts off just a little.

I read the waveform from a spreadsheet.

Output them and read them back average them send then to the fft and usually it works but not all the time.

[dsb@NI]

Thanks for providing some extra details, but I still think a picture is worth a thousand words... Screenshots of success and of failure would be nice. Also, please attach your VI. I can modify my example, but you can help us help you by attaching your modified VI.

Why are you reading the values from a spreadsheet? How many values are in the spreadsheet, and are you interpolating when the phase falls between listed values? Does your algorithm work in software (taking the hardware out of the picture for the moment)? When you say that every so many cycles, you see a shift, I am thinking that is a rounding/interpolation error when reading values from the spreadsheet.

[dksys2000@yahoo.com]

We need to use a spreadsheet to control the setup and values.

The amount of values can varies

I have tried taking the hardware out of the loop and the algorithm works software with no problems.

I have also tried using the same values multiple times and still get the glitch with no pattern. It works most of the time but still I get the glitch every so many times.

Dan

[dksys2000@yahoo.com]

Here is a screen shot in sync then one 1 point out of sync

[dksys2000@yahoo.com]

This one 1 point out of sync if you look at “raw waveform after delay and 1 pervious waveform should be 180 deg off” the green waveform is one point off. This throws off the FFT AVE