[time-nuts] Optical Distribution of 1PPS and IRIG
kb8tq at n1k.org
Mon Jul 14 18:12:51 EDT 2014
There are indeed modules that will move a pps over fiber. They are not cheap. They also are not real good bang for the buck if you need millisecond level timing. Without guidance on what we are trying to do here it’s very tough to come up with much. NTP over a symmetrical route will indeed do the job (to some level) without using any additional fiber at all. 1588 over a dedicated pair (virtual or otherwise) would do a better job, and still probably be cheaper than any of the dedicated fiber solutions.
On Jul 14, 2014, at 3:25 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> maflynn at theflynn.org said:
>> I am working on a project where I need to transfer 1PPS approximately 120
>> meters between buildings. I cam borrow a pair of 62.5 fibers from IT/
>> telco to do so.
> Google for >fiber IRIG< gets plenty of hits.
> Raw PPS is a nasty case for fibers. You need some sort of modulation/encoding, or derive it from the IRIG signal.
> The dynamic range that the receiver in a fiber setup can see is huge. All the modules I've worked with include AGC. That needs a signal to listen to. I haven't seen any with a time-constant slow enough to work with 1 PPS.
> If I wanted to send a PPS over fiber, I'd probably do something like this:
> send the PPS using Manchester encoding.
> that assumes you have a 20 MHz source clock synced to the PPS
> at the receiver, use a PLL to make a 20 MHz clock, and use that to decode the data
> The received PPS will be several clock cycles late. You can undo that with with simple digital logic. Just delay it by 20,000,000 - N cycles.
> Now it's only late by the fiber speed of light delay. (plus various digital prop delays and clock-out delays)
> These are my opinions. I hate spam.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts