[time-nuts] GPS noise reduction

Bruce Griffiths bruce.griffiths at xtra.co.nz
Sun Apr 6 09:54:41 EDT 2008

Magnus Danielson wrote:
> From: Bruce Griffiths <bruce.griffiths at xtra.co.nz>
> Subject: Re: [time-nuts] GPS noise reduction
> Date: Sun, 06 Apr 2008 14:30:52 +1200
> Message-ID: <47F835DC.50404 at xtra.co.nz>
>> METHOD 2:
>> If one uses an ADC to sample say a 1MHz sinewave (produced by dividing 
>> down the OCXO frequency and low pass filtering) on the PPS edge then 
>> subnanosecond resolution combined with the ability to make software 
>> corrections of the sawtooth error is possible.
> A refined method is to use a sine and a cosine, since these have their worst
> slope errors when the other has its best. But you can get far on just a sine.
> A classic method is to count time in a coarse clock and then produce an error
> signal which is integreated into a capacitor and then either converted into a
> longer pulse using a much lower discharge current. Another method is to use an
> ADC and then only clear the capacitor. Either way allows for easy access to
> x10 to x100 increase in time resolution. Really a handfull of stuff being not
> that exclusive and esquicit. Reaching for 1 ns to 100 ps resolution is doable.
> Cheers,
> Magnus
Hej Magnus

For the purposes of locking to the sinewave zero crossing a sinewave is 
more than sufficient as long as the sawtooth correction corresponds to a 
small part of the sinewave.
For example a maximum sawtooth correction of 15ns is equivalent to less 
than 0.1 radians of a 1 MHz sinusoid.
The sinewave nonlinearity for a 15ns correction is less than 130ps.
Locking occurs to the sinewave zero crossing in just the same as when 
using an analog phase detector with a sinusoidal phase characteristic 
(eg a diode double balanced mixer with an unsaturated RF port ) in a 
conventional PLL.

However if one wishes to measure the phase error rather than lock the 
sinewave zero crossing to the PPS signal then having  quadrature 
channels (sine and cosine) is useful in extending the range over the 
full cycle.

Suitable ADCs are relatively inexpensive.


More information about the time-nuts mailing list