[ntp:questions] Very large jitter and offsets on GPS ref clock after upgrade to "p5"

David J Taylor david-taylor at blueyonder.co.uk.invalid
Wed Jan 11 07:29:12 UTC 2012


"unruh" <unruh at invalid.ca> wrote in message 
news:dJ0Pq.27842$kp3.26636 at newsfe13.iad...
[]
> Yes, I was talking about the width of the pulse's rising edge.
> It is the risetime that is of first importance for timing. The length of 
> the pulse
> only is important in how the serial port identifies interrupts ( and I
> guess the possibility of the pulse been smoothed to nothing if too
> short-- but that would show up in the rise time first)

Whether the rise-time is significant would depend on the accuracy 
required.  I'm assuming that the waveform is rising cleanly and without 
ringing, and that the receiver circuit has hysteresis.  I'm thinking 
microseconds, and a 90m length of should be able to preserve risetimes of 
that order.  If the OP is looking for sub-microsecond accuracy, then 
delays in the PPS signal may be more significant, and the delays in any 
TTL-RS-232 convertor circuits etc. would start to matter.

Cheers,
David 



More information about the questions mailing list