False Peaks with Threshold Peak DetectorI

[johnsold]

Paul,

I do not have a lot of experience with the current DAQ cards, so I cannot give a direct recommendation.

If you tell your sales rep you want to sample 12-14 channels, he will very quickly forget scopes. The remaining issue is how nearly simultaneous the sampling needs to be. It is easy to say, "Sample all channels simultaneously." But for the kinds of measurements you want to make that does not seem to be necessary. You need to get enough samples on each channel to have good resolution in the transition regions, but with the noise and variability you seem to have, sample by sample simultaneity would be meaningless. I looked back through the thread. I don't see where you ever specified the sampling rates you are using or the actual transition times of your encoders. What units of time apply to the horizontal axis of the Actual Form.jpg for example?

Lynn

[Pablop]

Lynn,

I'm using two simulataneous channels at 100 MS/s. The simulataneity is required to check phase (edge separation) between the two channels. For some of our higher count encoders, the data rate is 800 khz, so that equates to 125 samples per cycle. Based on 1 cycle being 360° electric, the resoultion of my phase measurement is 360/125, or 2.88°e. That's kind of chunky if the tolerance I'm testing is say ±10°e.

Channels that are not simultaneous would induce further error in phase measurement.

[johnsold]

Paul,

It looks like you do need the speed!

Lynn