How can I generate a prescaled pulse stream using A/B encoder as input source?

[scwoods]

I have an optical encoder that generates an A/B pulse stream and I want to generate a continuous pulse stream to drive other hardware. I want to use one counter on a PCI-6601 board to receive the A and B pulse streams, decode in X4 mode, count down N ticks (prescale the tick stream), generate an output pulse, then reload and repeat continuously.

I tried using DAQmxCreateCILinEncoderChan() with X4 decoding, wiring A to AUX and B to SRC (reverse wiring for counting down), wiring OUT to GATE to force a reset on TC, enabling the Z index reset, and setting the Z index reset value and initial value to the prescale tick count. Unfortunately this did not work.

Do you have any suggestions?

[scwoods]

I have tested this a little further and have found that the output pulse is generated but each counter decrement is the result of one full cycle of the A/B stream. I would expect, given the counter is set to X4 decoding, that each transition should cause the counter to count down by one, not after 4 transitions. Also, is there a way to have the Z reset occur on any A/B combination, not just one of 4 possible ones?

[Kevin_Price]

I'm not certain you'll be able to get exactly the behavior you're hoping for.

1. You should be able to get the counter to decrement with each transition of A or B, making 4 decrements per full cycle.  In LabVIEW, one of the inputs for creating the position measurement channel is for specifying quad decode mode -- x1, x2, or x4.  There must be a similar call available in the direct DAQmx calls you seem to be using, but I don't know anything about the specific syntax.

2. No, unfortunately there ISN'T a way to perform z-index reload regardless of A,B state.  You have to specify a particular combo.  In your app, you'd also need your z-index reload value to represent an integer multiple of A,B cycles.  (If counting with "Ticks" for units, you'd need the z-index reload value to be a multiple of 4.)

-Kevin P.

[scwoods]

Thanks for the information. You are correct, it is close but it isn't exactly as I need as it requires my prescale to be a multiple of 4, and if that were the case I would use only one channel (A) of the encoder stream to begin with.

Instead I'll feed the encoder into a quadraturizer card and use its X4 output and direction bit to feed a 6601 counter to generate the required pulse train. Unfortunately it means more hardware. Maybe the next software revision will address this limitation.

Cheers...