[time-nuts] u-Blox ZED-F9T block diagram or timing knowledge available here?

[Bob kb8tq]

Hi

For a rubidium, the sawtooth error may or may not be as big a deal as for an 
OCXO. With an OCXO your loop time constant is likely below an hour. Indeed
it may be a couple of minutes. Bridges do pass through that sort of loop and mess 
up the output. 

The math:

10 ns hanging bridge at 1 second is 10 ppb
at 100 seconds it is 0.1 ppb
at 1000 seconds it is 0.01 ppb

If you are bothered by 0.1 to 0.01 ppb bumps in the frequency of your OCXO then indeed 
the bridge will be noticed.

Bob

> On Jan 6, 2020, at 9:56 PM, Forrest Christian (List Account) <lists at packetflux.com> wrote:
> 
> For a few grand you can get an off the shelf fs740 or something similar,
> probably with a rubidium timebase.  But I get the impression this is more
> of a learning experience.
> 
> One thing which it took me a while to wrap my head around is that for
> frequency, sawtooth error generally dissappears for a lot of definitions of
> reasonable accuracy.  You're averaging for such a long period that what
> seems like horrible accuracy is very usable.
> 
> Hopefully some of the more experienced time nuts will correct the following
> if my math is faulty.
> 
> For instance say your gps 1PPS source is accurate to +-1uS.  Over 1 second
> this is 1e-6.  But over 1000 seconds this is 1e-9.  Over 100000 seconds
> you're at 1e-11.   For a 10MHz clock adjusted using this long of averaging,
> this error is only +-.0001Hz
> 
> Of course this is somewhat simplified and the devil is in the
> implementation details.
> 
> As I'm sure others will point out, if you're using certain receivers, then
> sawtooth correction data is available to you which will enable further
> correction, which makes the internal clock frequency which the 1PPS edges
> are aligned to less of interest as the receiver outputs both the 1PPS and
> data indicating how far the receiver believes the 1PPS edge is out of
> alignment with ideal.
> 
> On Mon, Jan 6, 2020, 6:07 PM Tim S <tim.strommen at gmail.com> wrote:
> 
>> Hi, first time poster, just getting into the time rabbit-hole.
>> 
>>    I'm looking at building my own 10MHz double-oven + Rubidium standard
>> for home-lab use, and I wanted to investigate GPS disciplining.  I read
>> some remarkable work using Jupiter engines and a simple N/1000 and PLL, but
>> with those receivers becoming much more outdated and higher precision GPS
>> receivers now fairly cheap - I thought I'd try something new from uBlox to
>> not repeat prior efforts.  I've gone all through every resource I can find
>> online and on the uBlox forums, but there is a general lack of public
>> information about the time and clock features of the u-Blox F9 engine.
>> 
>>    Suffice it to say, there is a section in the integration manual where
>> they elude the 1PPS signal is set to the closest one of 1023 edges - which
>> seems to suggest a possible sawtooth phase noise creation.  Talking with a
>> person "clive1" on the uBlox forum it sounds like the internal GPS engine
>>>> MAY<< now running at 384MHz there sounds like a possibility to align a
>> 1PPS edge to a much finer resolution clock which is closer tied to the GPS
>> solution and thus less digital divider phase noise might be possible.
>> 
>>    Does it sound like I'm way out in left field here?  Anyone have the
>> luxury of more insight into the construction of the uBlox F9 GPS engine?  I
>> don't mind spending a few grand on some factory new components to get a
>> decent 10MHz standard, but I'm less interested in doing so if I'm not going
>> to understand what's going on inside.
>> 
>> Thanks in advance for responses,
>> 
>> -Tim Strommen
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.