[time-nuts] Weird GPSDO behavior

Bob kb8tq kb8tq at n1k.org
Fri Sep 29 11:50:28 EDT 2017


If a simple GPS outage makes the GPSDO go bonkers, there is something 
else involved. Noise jamming or flying saucers over the antenna should just shut
down the receiver. When it locks back up again, the disciplining should 
resume. If it goes into a death spiral that pretty strongly suggests that the
math isn’t handling things very well. Indeed the math may never have been 
tested for this case so it becomes “that’s the way it is”…..

20,000 seconds is errr …. 5.55 hours. A normal loop could take 10 tau to settle 
to a reasonable level. That would put “bump recovery out around two days. The
sort of checks and tweaks that make sense on a 10 minute tau likely are not the
same for a 5 hour loop … It could indeed be a timeout timer saying “not settled 
in a day? -  reset …”. 


> On Sep 29, 2017, at 11:42 AM, jimlux <jimlux at earthlink.net> wrote:
> On 9/29/17 6:13 AM, Rob Kimberley wrote:
>> I'd go with a power surge as it's so regular at 8AM.
>> Rob
> Delivery truck with Jammer, as suggested by Graham also.
> There could be a RFI burst from something like a streetlight or storefront display turning on/off.
> Have you analyzed the timing of the outages?  What sort of distribution do they have? (I hesitate, but this is time-nuts... what is the ADEV of the outage times<grin>)
> If the timing is *very* regular - that implies its controlled by a clock
> If the timing slowly drifts later (if you're in the northern hemisphere) then it might be tied to an astronomical timer (e.g. sunrise) or an actual light detector.
> If the timing is sort of random, but around a median, that implies some sort of human scheduled activity.
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

More information about the time-nuts mailing list