[time-nuts] Re: James Miller/Brooks Shera 10 MHz GPSDO

[Bob kb8tq]

Hi

The specific version of the 10811 involved has a very wide tune range
on the EFC. When they blocked access to the mechanical trim, they 
boosted the EFC range. The circuit is basically still the same, they just
changed some values to take it up above +/- 1 ppm ( on most examples).

Since this is a locked circuit ( a GPSDO ), long term drift and even medium
term issues likely get taken out by the lock process. Just how quick things
go from “this matters” to “taken care of by lock” depends a lot on exactly
which of the many versions of this GPSDO is involved. 

On many GPSDO’s the lock bandwidth is in the “couple hundred seconds” 
range. Go to ~5X that and all you should see is the GPS. Go to  ~1/5 of that
and the oscillator is all that is involved. 

As has been noted a lot of times here, the correct lock point is determined 
by the noise of the OCXO and just how good a PPS you get out of your specific
GPS module. Things like sawtooth correction can make a difference. So 
can the “era” the module came from. Since neither the OCXO or module 
have a “single number” sort of noise process, it’s very much a measure and
see sort of thing.  Simply put, there are a number of versions for a very good
reason ….

Bob

> On Aug 31, 2022, at 5:47 AM, Dan Kemppainen via time-nuts <time-nuts at lists.febo.com> wrote:
> 
> Hmmm, What is the temperature coefficient of coin cells?
> At one point I hooked a AA battery to high resolution voltage data logger. The AA battery made a pretty good thermometer.
> 
> Maybe the temperature coefficient of the battery combined with the low EFC sensitivity wouldn't matter. Personally, I'd be concerned adding anything else to the system that could tune the crystal frequency.
> 
> Dan
> 
> 
> On 8/31/2022 3:31 AM, time-nuts-request at lists.febo.com wrote:
>> Pretty clever approach to establishing an offset. But doesn't it matter how
>> the vari-cap is hooked up in the oscillator circuit? If the veri-cap needs
>> a negative voltage to work that may indicate the cathode is at ground and
>> the positive voltage would forward bias the vari-cap. Just thinking about
>> the possible impact.
>> Though the cost of experimenting is $0. No harm.
>> Regards
>> Paul
>> WB8TSL
>> On Tue, Aug 30, 2022 at 1:22 AM Mervyn Thomas via time-nuts <
>> time-nuts at lists.febo.com> wrote:
>>> Stephen,
>>> 
>>> An idea I may try -thanks for your response.  The current involved in
>>> the EFC line to the oscillator is extremely small-button cells would
>>> last years.
>>> 
>>> Merv
>>> 
>>> On 29/08/2022 8:42 pm, Stephen C. Menasian via time-nuts wrote:
>>>> Merv,
>>>> 
>>>> If I understand your problem, the 10811's VCO control input voltage has
>>>> drifted above the output range of the GPSD driver board. If the VCO
>>>> control terminal current is low enough, you might be able to quickly
>>> solve
>>>> this problem by putting 1 or 2 Lithium Ion Coin cells in series with the
>>>> control lead. If only a few microamperes are involved, this solutin might
>>>> last several years.
>>>> 
>>>> Stephen Menasian
>>>> 
>>>> On Mon, 29 Aug 2022 15:56:52 +0800
>>>> Mervyn Thomas via time-nuts<time-nuts at lists.febo.com>  wrote:
>>>> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com