Picked this project back up this afternoon after couple of weeks, narrowed the problem down but not solved.
Refresh: My ONCORE UT+ module appears to have very extreme jitter (on on order of 100 uSec, not nSec random variation between leading edges of successive pulses) in the 1 PPS output. I am using the 1 Hz signal, divided to 0.05 Hz at 50% duty cycle (10 seconds high, 10 seconds low), to gate my frequency counter's reference oscillator back into the same counter running in pulse-count mode so I can adjust the OCXO bias on the counter for calibration. 10 seconds count-up, followed by 10 seconds idle where I can record the measurement, reset the counter, and adjust the OCXO bias. I've now built a simple 1 Hz pulse generator using a 32768 Hz crystal and CD4060 divider. Using that signal instead of the ONCORE UT+, my counter is behaving as expected (that is, each count-up session results in the same result, +/- a few counts as would be expected with a simple 1 PPS signal.) This tells me the gating circuit and interface to the counter's reference output and count input are probably OK. When using the ONCORE UT+, the counts vary by 10000 or more from one 10-second capture to the next (each count-up period would total 130,000,000 counts if the ONCORE UT+ was working properly and the reference OCXO was properly adjusted.) I am not yet able to determine why the UT+ 1 PPS signal has such severe jitter. TAC software tells me 7-8 satellites acquired/tracked once cold start has completed (10-15 minutes.) Is there something I'm missing about UT+ care and feeding? There are no RF or switching digital signals close to the UT+ module (the gating circuit is a couple of inches away, both mounted on the same perf board, with common ground and power supply connections. Now that I've ruled out the counter/gating circuit interface, I begin to suspect the UT+ but don't really know what to expect in terms of normal/abnormal operation. Any advice at this point? Thanks, Dave _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
