Hi With OCXO A in environment B you may make it to step 6 in the time constant update table. With OCXO D in environment E you may make it to step 9 on your table of time constants. In the first case step 5 is probably a useful thing to flag, in the second case step 8 makes some sense. The problem is that if the environment changes, the final step will change as well.
All that of course *assumes* that the logic in the loop decision making does not let it go to far down that table. If OCXO A is still in environment B and it (somehow) gets to step 9, that's probably a bad thing. It likely needs the faster time constant to keep things in order rather than the slower one. Bob On Aug 30, 2013, at 10:49 PM, Bob Stewart <[email protected]> wrote: > Hi Bob, > > You've given me a lot to work with. Perhaps the best answer is to put in a > user parameter for how many seconds between updates to consider a lock > condition. For warmup, I had planned to put in a 5 minute holdover period, > but that could easily be user configurable, as well. > > Thanks again, > > Bob > > > > > >> ________________________________ >> From: Bob Camp <[email protected]> >> To: Discussion of precise time and frequency measurement >> <[email protected]> >> Sent: Friday, August 30, 2013 8:55 PM >> Subject: Re: [time-nuts] GPSDO PLL convergence question >> >> >> Hi >> >> A GPSDO is nothing more than a PLL that operates at 1Hz. If you use the same >> criteria as they do in some systems, as soon as the OCXO pulse is within 1/8 >> second (sort of 10%) of the GPS, the PLL is locked / converged / doing it's >> thing / functioning as it should. It's a time domain system, so time is the >> reasonable way to look at it. >> >> Looking at DAC voltage *assumes* that the OCXO is perfect. Consider a >> situation when the temperature is continuously climbing. The OCXO has a TC, >> the DAC will move continuously. The same thing applies to aging / warmup / >> what ever. >> >> Bob >> >> On Aug 30, 2013, at 9:36 PM, Bob Stewart <[email protected]> wrote: >> >>> Hi Bob, >>> >>> I'm not sure I understand what you mean by "45 degree phase error" in a >>> GPSDO. This isn't a situation where I'm at the right frequency but at the >>> wrong phase: at least not by my understanding of phase and frequency. >>> Maybe PLL is even the wrong term. I dunno, but it feels right for this. >>> Let's say I'm counting the number of OCXO pulses in each second, and in two >>> intervals 100 seconds apart I have a surplus of 1 count each. I think that >>> means my 10MHz OCXO is off by 1ppb. Is that right: the reciprocal of 1 >>> error count during 100 seconds times 10,000,000? And yet my DAC still has >>> about 100 steps to go before it reaches best solution. So, OK, what is the >>> "magic number" in ppb or ppt I should consider for convergence? Maybe I >>> should collect 24 hours of data and see what the average is for the time >>> between DAC changes and work from there. I suppose 5 counts from best >>> solution would be fine. By my math that would be 240ppt, or about 417 >>> seconds between DAC updates. I'm out of my depth on this project. >>> >>> >>> Thanks for your continued help. >>> >>> Bob >>> >>> >>> >>> >>>> ________________________________ >>>> From: Bob Camp <[email protected]> >>>> To: Discussion of precise time and frequency measurement >>>> <[email protected]> >>>> Sent: Friday, August 30, 2013 7:31 PM >>>> Subject: Re: [time-nuts] GPSDO PLL convergence question >>>> >>>> >>>> Hi >>>> >>>> Convergence in a PLL has no absolute definition. In some cases 45 degree >>>> phase error = all is running fine. >>>> >>>> I would simply track GPS time vs OCXO time and declare it ok when it's >>>> inside some magic number of nanoseconds. >>>> >>>> Bob >>>> >>>> >>> _______________________________________________ >>> 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. >> >> _______________________________________________ >> 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. >> >> >> > _______________________________________________ > 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. _______________________________________________ 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.
