DAQ Timing Issues Using USB-6341

[ianscandela]

I need help synchronizing my DAQ IO signals properly. What I am trying to do is move two galvo motor with mirrors attached incrementally (4-100 points) in sync with a laser diode signal. Picture pointing a laser pointer at a mirror and moving the mirror around while holding the laser pointer stationary; that is the broad view of my setup.

I want to send a laser pulse signal once for each incremental movement of the galvo mirror, no more no less. I have breadboarded a PWM trigger signal that I am using to both turn on and off the laser AND move the galvo. The signal is 200-300 Hz, where the HIGH time is always 350us. Only the low time changes when switching frequencies.

My DAQ IOs are:

• 2 analog outputs (ao0 and ao1), each wired to the galvo mirror to select position
• 1 digital trigger input (PFI0) that is wired to the aforementioned PWM trigger signal
• 1 digital output (p0.0) to turn on/off the trigger board

I am able to successfully get everything working in sync, but am always getting one extra pulse due to a timing delay on the trigger on/off signal. The DO I am sending to turn the trigger off takes anywhere from 3-7ms extra, causing for extra unwanted pulses coming from the laser after the mirror has stopped moving.

Here is a picture of the main loop of my code as well as a scope photo of my signals.

Signals above are:

yellow = trigger signal (PFI0)

blue = trigger on/off signal (p0.0)

green/purple = two galvo mirror signals (ao0 and ao1)

In the scope photo I am trying to run a total of 4 pulses, but the trigger does not stop until 5.44ms after the fourth trigger pulse, and I end up getting an unwanted pulse from the diode. This 5.44ms varies +/- 2ms each operation.

If anyone has any suggestions as to how I can turn the trigger signal off more quickly I would be so grateful!

[Kevin_Price · Accepted solution]

Well, you could just generate that "PWM trigger" signal with one of your 6341's onboard counters.  Then you can set it up with the high/low time combo you want, and configure it for Finite Sampling and exactly 4 pulses.  And you'd no longer need that software-timed DO task that can't perfectly sync with the hardware-timed AO task anyway.

-Kevin P

[ianscandela]

Thanks, this worked. To anyone interested, here is what I ended up with: