[time-nuts] Testing frequency using NTP

[Steve Rooke]

2008/10/2 Poul-Henning Kamp <phk at phk.freebsd.dk>:
> In message <1231b6a80810012218t8f9b37y7395215c24229305 at mail.gmail.com>, "Steve
> Rooke" writes:
>>2008/10/2 Bruce Griffiths <bruce.griffiths at xtra.co.nz>:
>>
>>> Save yourself a counter and just divide the frequency down to about 1Hz
>>> and time stamp the 1Hz transitions with the Linux box.
>>> As long as you know the division factor its easy enough to calculate the
>>> frequency.
>>
>>I'd planned on a simple interface to the PC via using the parallel,
>>suitably strapped, as a basic output port. Now, I could use it as an
>>input but I'd probably have to poll the port which would be somewhat
>>inefficient [...]
>
> Linux should have the PPS-API from RFC2783 so that should be piece
> of cake to do.
>
> I know that it works in FreeBSD (guess who wrote it :-) where you can
> just:
>        cd /usr/sc/tools/test/ppsapi
>        make
>        ./ppsapitest /dev/lpt0

OK, interesting idea but I'm really not planning to use the PC to
count pulses. I have a Fluke 1912A Multi Counter which can count
pulses to 7 digits and I can allow that to overflow to get more digits
by changing the pre-scaller I plan to use on the front end of this. It
would be easier to use Linux to control an input gate to this
pre-scaller with a known good period of time.  Besides which, my quick
scan for PPSAPI/KIT only showed patches for the 2.2 and 2.4 kernels,
still I suppose I could use an older distro or even FreeBSD for that
matter (it's all Unix to me).

Cheers,
Steve
UNIX since SVR0
-- 
Steve Rooke - ZL3TUV & G8KVD
Omnium finis imminet