[time-nuts] True Position GPSDP + Rb X72
kb8tq at n1k.org
Wed Sep 20 10:21:51 EDT 2017
> On Sep 20, 2017, at 8:39 AM, Attila Kinali <attila at kinali.ch> wrote:
> On Tue, 19 Sep 2017 18:09:16 +0000 (UTC)
> Stephan Flor via time-nuts <time-nuts at febo.com> wrote:
>> Hi everyone,I'm new to this list, and I wanted to get the opinion of some
>> mere experienced people on my project idea.I just got a Symmetricom X72 Rb
>> oscillator, and I have a True position GPSDO on the way. I wan't to have a
>> self calibrating 10 MHz signal to synchronize my hobby rf lab equipment.
>> My idea is to make a GPS disciplined Rubidium oscillator, by removing the
>> crystal oscillator ans substituting the X72 in the GPSDO circuit.I would just
>> have to make a op amp circuit to change the 0 to 4v correction signal from
>> the True Position, to a 0 to 5v signal for the X72.
> Unfortunately, this doesn't work as easily as you thing.
> The control loop inside the GPSDO is designed for an OCXO. This means
> that its loop constant is in the 10s to 100s of seconds, while for
> an GPSDRb you would need 10'000s to 100'000s of seconds. The other,
> more criticial issue is the sensitivity of the tuning input. Most
> OCXOs have something like 1-10ppm of tuning range. The tuning range
> of an Rb is usually 2 to 3 orders of magnitude smaller. In a normal
> PI loop, this shouldn't be an issue. But you don't know what the
> internal logic does. So at least be careful with that.
> The more common way to build such a system is to do a phase comparison
> between the 10MHz output of the GPSDO and the Rb, then steer the Rb
> using a seperate control loop electronics. One easy way to do this
> would be using a uC with a capture compare unit, at least one with 16bit,
> better with 24bits and more (e.g. STM32F4xx). Divide both 10MHz outputs
> down to roughly 1kHz-10kHz using 74LVC161. With this you can measure
> the phase differnce between the two with about 12ns precision (limited
> by the uC, not the dividers). Average over 2-3 days and you are good.
An alternative is to add an offset oscillator to the mix and look at beat
notes. It does not have to be ultra close. An 11 MHz offset signal would take the
12 ns down to an effective 1.2 ns. Odd frequency OCXO’s sell cheap on
> For additional precision, you can use an TDC7200 to measure the time
> difference with around 50-100ps precision, at which point the GPSDO
> noise and the stability of the Rb will be the limiting elements.
> Attila Kinali
> It is upon moral qualities that a society is ultimately founded. All
> the prosperity and technological sophistication in the world is of no
> use without that foundation.
> -- Miss Matheson, The Diamond Age, Neil Stephenson
> 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