"unruh" <[email protected]> wrote in message
news:[email protected]...
[]
No, the Sure unit outputs a 100 ms wide pulse. I have not measured its
Sorry, it risetime width.
risetime. The stated accuracy for "time" is 20 ns, but the Web page
does
not state where that 20 ns is measured - it may be internal to the
chip.
I would expect it to be neither significantly better nor significantly
worse than the Garmin.
20ns is significantly better than 1000ns (a factor of 50)
But there is no statement in that document as to where that 20 ns is
measured.
The risetime, as stated, I measured to be about 5ns.
.. as are many logic level signals. They may be precise to 5 ns, but not
accurate.
And much of that error is due to the sawtooth error in the output. (As I
understand it, in order to make things simple, the pulse is emitted at a
the nearest zero crossing of the internal oscillator, rather than
exactly on the time. Thus the actual time of the pulse wanders depending
on the phase of the interal oscillator with respect to the actual time.
The oncore reports that difference so you can, post facto, correct for
that error in the output pulse time compared to the true second
crossing. The sure, AFAIK does not, so it has an inherent error due to
this. (sawtooth because the phase of the internal oscillator drifts with
respect to the true time and thus slowly increases until it reaches 180
degrees and then jumps back down to 0)
There are some plots here:
http://www.leapsecond.com/pages/MG1613S/
The sawtooth has an amplitude of about 60 ns peak-to-peak (according to
the text), which (for me at least) is negligible even in a FreeBSD/NTP
context.
Perhaps someone, somewhere has compared the absolute timing of the GPS 18
LVC, Oncore and Sure boards. It's all very well noting 50 ns jitter, but
if that edge is 200 ns out? Within 1 microsecond suits me fine, but
perhaps it's not good enough for others.
Cheers,
David
_______________________________________________
questions mailing list
[email protected]
http://lists.ntp.org/listinfo/questions
