[time-nuts] Advise on building a DIY GPSDO?

[Chris Caudle]

On Mon, April 4, 2016 11:18 am, Tom Holmes wrote:
> The preceding questions always come up when a newbie
> comes up on the list wanting to build their own GPSDO.

It is good to make sure the person actually wants to build a GPSDO.
A few years back I was in a similar position, and my answer was I actually
needed a high accuracy reference to check frequency accuracy of some
clocks, and the recommendation was just get a Thunderbolt.  I didn't know
about surplus GPSDO's available so that was good advice for me, I didn't
actually >want< to build a GPSDO myself.  It would have been fun, but the
end goal was more important to me.

So I think really the first question should be do you want to build a
GPSDO, or do you want a high accuracy time and frequency source for its
own use?  Because if you want to own a GPSDO to use, building one yourself
is probably not the way to go.  Get a surplus unit, or save up a little
more and get a new Jackson Labs.
If you just like to build stuff and a GPSDO is the particular stuff you
want to build next, then go ahead and have fun.  Chris Albertson's arduino
based design is probably as good a place to start as any, it should be
cheap and I think the pieces are easier to find than that CPLD based
design that relies on having a GPS that can output 10kHz instead of just
PPS.

-- 
Chris Caudle