Dennis, > Ugh -- let's try Google Drive...
I was able to download them, thanks. All but the 100uSec one seemed to be rather stable. > That is what I'd thought, but if you've looked at that simple program I > used, apparently the "run longer" was taking place between the > > t += > > and the > > t - time.time() I don't think so, as that "t +=" is relative and depends on the end time of the last sleep - just as it should I might say. And that means that that "run longer" can have been taking place anywhere between the end of the last sleep and the "t - time.time()" in the new one. By the way, on my scope the "wrap around" time seemed to be quite long, in the neighborhood of 80 uSec Alas, the drift-compensated sleep did not seem to make the output any more dependable - though I hope it means that I do not need to re-tweak the delay when running on another Pi than my current one (a 3B+). Regards, Rudy Wieser -- https://mail.python.org/mailman/listinfo/python-list
