Hi to all, Very new and green at this .. I am trying to understand more about Allen Deviation, a subject about which I , for practical purposes know 'nothing'.. I have spent a lot of time digging on internet and have read many articles , inc W.J Wriley publications, the Stable32 files and user manuals, dug in the TimeLab files, etc. I understand the concepts somewhat, but some of my practical results are beginning to seem too good to be true.
I built a GPSDO with an HP 00105-6013 OCXO, covered in some previous mails on this forum so I won't go into the detail, but I have finally got the thing working well, 'stable' or so I believe/thought. I log the output of the TIC, in nanoseconds, and use that file to generate an ADEV plot. After 10-12 hours, it looks good - was approaching 10 minus 12, and that is good enough for my purposes. So I left the unit running and logging - after 32 hours, it approaches 10 minus 13,.... The line continues downwards - I do not see much indication of the upwards turn I see in all other ADEV plots. I have not managed to understand the mechanism behind this upturn.. The attached - Rub_Adev.gif is just an example of what I mean - this image courtesy of LeapSeconds.com, of a rubidium source. A plot of my oscillator is in JN_1e-13.gif. That show that after approx 142K 1sec samples, the Adev is around 1E minus 13...!! and not showing signs of turning up yet. This is surely not true? If not true, what should I be looking at to understand what is going on? My TIC measurement resolution is for all intents, around 0.25ns, but I suspect noise makes it no better than 2ns. Why does my plot keep going down, below minus 13? Why is it going down that far? My GPSDO cannot be that good? Is there a means of determining the ( minimum?) number of phase samples needed to give a sensible indication of the Adev value? Before I ask more questions, I need to discover the extent of what I don't know, so I don't ask too many foolish questions! Thank you! regards Joe
_______________________________________________ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com and follow the instructions there.