preventing a glitch during signal generation

[bmihura]

I’ve got two high speed digital IO cards (NI PCI-6541) generating simultaneous waveforms, and I’m getting a 120 – 150 ns glitch from one during the generation process. The waveforms have ~ 200k samples at a sampling rate of about ~ 20 MHz.

For each board, I call this sequence of  functions:

niHSDIO Init Generation Session.vi
niHSDIO Assign Dynamic Channels.vi
niHSDIO Configure Sample Clock.vi
niHSDIO Export Signal.vi
niHSDIO Configure Generation Mode.vi
niHSDIO Export Signal.vi
niHSDIO Configure Generation Repeat.vi
niHSDIO Write Named Waveform (1D U32).vi
niHSDIO Write Script.vi
niHSDIO Get Session Reference.vi

It’s the “niHSDIO Write Named Waveform (1D U32).vi” function that causes the glitch. I’ve run into a similar thing before and it was the signals briefly going to a high impedance state during an initialization call; is there some way to prevent that?

[Matt_A]

Hello All,

Bruce has contacted National Instruments support regarding this subject. I am currently working with him to resolve this issue. I will post the resolution to this forum after we have determined a solution. Please feel free to comment to this thread or add any insight you may have.

[Matt_A]

Hello All,

This issue was resolved by changing the initial state in the niHSDIO Init Generation Session.vi. The "glitch" occurred because the niHSDIO Init Generation Session.vi was configured to reset the device at the begining of the program. This set the device into a high impedance state. By changing the input so that the device did not reset we were able to eliminate the problem.

[bmihura]

Nice job Matt.