[time-nuts] precise external reference time for a PC

Hal Murray hmurray at megapathdsl.net
Sat Jul 12 14:07:13 EDT 2008

> for an upcoming experiment I am looking for precise external time
> reference, preferrably an atomic clock that I can hook up to a
> computer running a Unix variant. 

How good a clock do you want/expect?

If you are using a computer, you have a layer of jitter from things like 
interrupts and cache misses.  PCI adds more jitter from things like DMA 

How stable is your thermal environment?

Have you experimented with ntpd and a simple GPS unit?  That turns the 
problem from hardware to software.  I'd expect you could get the system clock 
to within a few 10s of microseconds.  That may take luck and work.  Quirks 
will vary from OS to OS and from release to release, both of OS and ntpd, and 
maybe from motherboard to board.

These are my opinions, not necessarily my employer's.  I hate spam.

More information about the time-nuts mailing list