[time-nuts] Software Sawtooth correction prerequisites?

Tom Van Baak tvb at LeapSecond.com
Fri May 11 20:38:51 EDT 2007


> My first post...newbie...be gentle...
> 
> I spent the last several evenings reading the archives and saw mention of 
> sawtooth error correction in software. Since the corrections to be applied 
> are on the order of 1e-9 seconds it would seem that the phase detector 
> outputs to which these are applied must be similar in resolution.

Right. When a correction is way smaller than the signal one
usually doesn't bother with a correction. On the other hand,
some might argue that every little bit helps.

> That would seem to require a pretty hefty phase detector and a pretty 
> substantial computing resource. Doesn't sound inexpensive. Is there a way 
> around this?

I would agree. A 1 ns phase detector is usually more than a
single IC. Not sure what to say about the computer resource,
though. The sawtooth correction is usually obtained though a
low-speed (e.g., 9600 baud) serial message from the GPS
receiver. Most any $2 microcontroller can handle a task like
this.

When someone finds a cheap single-shot 1 ns TIC-on-a-chip
please let me know.

> OTOH, the Dallas Semi delay line pushes the "computation" out into the input 
> of the phase detector at the cost of a $17 chip (1k qty) which in small 
> quantities is likely $50 or so. This would seem to allow for a wider variety 
> of phase measurement techniques.
> 
> Do I have this right?

Yes, although it depends on the application.

If you are using GPS, for example, to monitor the long-term
performance of a cesium standard then you already need a
high-resolution TIC. Sawtooth correction is useful but it doesn't
make more than 1 ns difference if it's hardware or software.
Sometimes using two serial ports and correcting in software
is fine. Other times it's more convenient to use a box like the
CNS-II where it's all nicely done in hardware.

If you hare thinking of a GPSDO you need a phase detector
so you have to ask yourself if hardware sawtooth correction
with a good phase detector is better than an excellent phase
detector with just software correction. You'd have to do the
math to decide which was better.

The other factor that someone can comment on is what effect
time averaging has on all this. Can a coarse phase detector do
the job if you average long enough? Most GPSDO don't need
to make decisions every second. If you have a good OCXO
and tweak the EFC only once every 100 or 1000 seconds, can
you get away with a cheaper phase detector design?


All of this reminds me -- how many of you would be able to, or
would like to, do a GPSDO simulation? I have or can get days
of M12+ raw data, sawtooth data, and OCXO data. It should be
quite possible, then, to model a GPSDO.

Once you have this virtual GPSDO you can instantly see what
effect something like sawtooth correction has. Or see what
effect making the phase detector 5x better or 5x worse has on
the output. Or see what happens when there is a frequency
jump in the OCXO. Or the effect of a temporary loss of GPS
signal, etc. Or vary the cross-over point to see how the ADEV
of the VGPSDO changes. Or test different PID algorithms. Or
compare the effect of a cheap vs. good OCXO, etc.

/tvb

> tnx
> 
> jim miller
> ab3cv 





More information about the time-nuts mailing list