Generating digital pulses from the pci 6722

[ag14]

I am trying to generate signals that will drive a piezoelectric motor. I need an analog signal that I will send to an amplifier and 6 digital pulses with varying offsets. I am using a pci 6722 and was hoping to get an offsets between the digital pulses to be around 2 microseconds. I am struggling  to find the best  way to get such exact time delays between multiple tasks. Any help  would  be greatly  appreciated.

[Eric-E]

Hello, if you are stuck using this PCI module, you will most likely encounter a problem of jitter within the windows operating system. The jitter of Windows will prevent exact timing which is the purpose of Real-Time operating system. You can always try this first hand to find how precise this is, but the jitter will become apparent right about at this resolution. If you do not need the timing to be precise, you can probably get reasonably close, but I would highly recommend some new hardware.

[ag14]

The actual offset times is not as important as the digital pulses firing sequentially and all within the envelope of the analog signal for multiple steps.  I guess my question is less about hardware limitations, and more about how execute my desired process in Labview. My experience in Labview is pretty rudimentary and I am struggling to see how to synchronize events in labview.  We are using a BNC 2110 to output our signals, and it is my understanding that the BNC 2110 has a 1 Mhz clock.  We are trying to import that clock from our device use that to clock timed loops and we are having some difficulties with the details.  I am honestly not sure if I am on the right track.  In hopes that it might add some clarity I attached a picture of the signals we are hoping to produce.  The orange signal is our analog output.  Between times t3 and t4 we need the 6 digital triggers to switch to "on" and we need them to all switch off simultaneously at t6.  Again we are definitely willing to compromise on speed if we can use the card we already have.

[Eric-E]

Hello ag14,

I'm not very familiar with this type of application, but I think the stat you would look at for this problem is the 800 kS/s per channel for one channel. This would give you one digital signal each microseccond (ish) so that's why I said that this hardware might not be the best for this particular application. You may not have enough digital points to create the desired waveform, but it might be just enough. 

You might want to take a look at this page regarding your implementation question. You would get rid of the trigger and property node, but this is the general structure of a finite output.

https://decibel.ni.com/content/docs/DOC-12104

[Dennis_Knutson]

Note that you cannot use the digital I/O for anything close to the timing you need. The digital I/O is software timed as you can see from the spec so the 800kS/s rate mentioned above does not apply at all. With software timing, you might be lucky to get 1ms resolution and that would have a large amount of jitter.

The BNC 2110 is a dumb terminal clock. Where did you see that it has any clock?