[time-nuts] GPSDO control system
csteinmetz at yandex.com
Sun Mar 23 06:28:57 EDT 2014
>1. The difficulty with disciplining a local oscillator to a GPS signal
>is due to variations in the received GPS signal and the LO.
I'm not sure I'd call that "the difficulty" -- it is the task of a
GPSDO to discipline the local oscillator *at time scales where the
GPS is better than the oscillator*. The oscillator wanders, and the
GPS signal has noise and jitter. The oscillator controls stability
at small to medium tau, so improvements there require a better
oscillator (nothing you do on the GPS side is going to improve
overall stability below tau = 100 seconds or so). If you have an
exceptionally stable oscillator, this will be true out to tau = 1000
or even beyond.
>3. The gain of the system, in degrees of phase angle at 10 MHz (or
>higher) per microvolt of control signal, is fairly constant in a
The gain function is a free parameter -- it can have pretty much any
characteristic you desire, if you are sufficiently creative to
realize it. In practice, of course, it will be constrained by the
criteria for stability.
>4. The power supply for the device providing the control signal cannot
>be regulated to the accuracy required of the system, and so is a source
The control amplifier should have considerable PSRR, so this should
not be a material source of error. The reference for any DAC in the
system may contribute error. But that is all inside the control
loop, which is driven by a phase comparator of one sort or
another. Ultimately, the errors in the inputs (oscillator and GPS)
and the phase comparator should dominate the overall performance (if
not, you have left something on the table).
>5. The principle source of environmental variation is temperature.
>Humidity and barometric pressure are not significant. This may not be
>true of the received GPS signal due to atmospheric variations.
You will have to test the parts you propose to use to see if they are
affected by humidity or barometric pressure. Some researchers have
reported sensitivity to one or both for crystal oscillators and
voltage references, among other parts. Except for the oscillator,
GPS, and phase comparator, that is all inside the control loop.
More information about the time-nuts