02-28-2012 01:46 PM
@Kevin
Thats pretty much what I am doing now actually. Except I am doing it with the Quad Decoding of 8192 pulses. All the processing is done via software and the Encoder sampling is sync'd to the sample clock of the DSA.
I was looking for a hardware based solution so I can decouple my sampling loop from the speed monitoring. Since all timing signals are available in hardware, why not use hardware control? In my original post, I mentioned that I have a complete working system. I am looking to improve on the design by eliminating software processing of the signals. By doing this, the timing is embeded in the sampling interval and data can be normalized. I believe this will yield a more accurate and repeatable result for the process due to flucuations in the closed loop motor control.
02-28-2012 01:50 PM
Based on Katie's info, parts of my suggestion definitely won't work. If the DSA can export a sample clock signal, other parts of the idea might still work.
Modifications would be:
- use the DSA sample clock to capture encoder samples
- configure the encoder for X4 for better position resolution. Continue to reset to 0 on Z-index.
- you'll almost definitely be interpolating to find the 0-position crossings because the DSA sample isn't likely to coincide with the z-index
-Kevin P
02-28-2012 02:07 PM
@Kevin
Yep thats exactly correct. 0-position crossings have to be interpolated.