[time-nuts] 10811's

Charles Steinmetz csteinmetz at yandex.com
Mon Nov 17 22:49:40 EST 2014


Bob wrote:

>As best anybody can tell, the improved phase noise and ADEV 10811's 
>came out of a screening / select process. There is a finite chance 
>that a "normal part number" 10811 could be as good as or better than 
>a part from one of the "better part number" versions.

That is particularly true now that they all have decades of 
aging.  The specs were written for new units with fresh crystals, 
which are nearly always less stable than well-seasoned crystals 
(e.g., you will often find annual drift specs given separately for 
"first year" and "subsequent years").

At this point, I think it is essentially random which individual 
10811s are most stable (and, of course, it also depends on what you 
mean by "more stable" -- is the oscillator that drops to 1e-12 by 1 
second and stays below 5e-12 through 1000 seconds more or less stable 
than the one that is 2e-11 at 1 second, drops to 1e-13 from 20 to 100 
seconds, and is back up to 2e-11 at 1000 seconds?).  I certainly 
haven't noticed any pattern with the 10811s that have been through my hands.

I know -- everybody, label all of your 10811s with your name and 
address, box them up, and send them to Tom with a return shipping 
label and sufficient postage.  He can run each one for PN and xDEV 
(after burning each one in for a week, of course), publish all of the 
data, and return them.   ;-)    We're bound to have at least a few 
thousand between us.  Don't forget the ones in your counters and 
other equipment!

Best regards,

Charles





More information about the time-nuts mailing list