[time-nuts] GPS disciplined oscillators - how not to do it.
Dr Bruce Griffiths
bruce.griffiths at xtra.co.nz
Thu Apr 5 09:41:22 EDT 2007
Bob Paddock wrote:
> On Tuesday 03 April 2007 21:48, Dr Bruce Griffiths wrote:
>> An Australian Electronics magazine recently published a circuit for a
>> GPS disciplined crystal oscillator.
>> This particular implementation is the worst I've ever seen.
> What would you consider the best you have ever seen?
Probably Quartzlock's carrier phase disciplined OCXOs.
Although a few improvements spring to mind.
If we are only considering phase locking to a GPS timing receiver's PPS
output the HP58503 and its ilk weren't too bad although the double oven
OCXO implementation used is a bit of a kludge.
The Post was intended as a warning don't be tempted to build the circuit
as you will be disappointed.
The logic circuitry appears to have been designed by 2 people, one who
knew how to cascade counters correctly and one who didn't.
Whoever designed the PLL and its filters seems to be particularly clueless.
Using parts of a hex inverter for an oscillator and its buffers as well
as using the rest as buffers for other frequencies is also asking for
Its a shame that an opportunity for demonstrating the performance that
is achievable with a well designed crystal oscillator without an oven
The choice of a GPS receiver with a PPS timing accuracy more than 30
times worse than that possible with most good GPS timing receivers
More information about the time-nuts