Hi Ray, It seems that some of the "old" techniques have been forgotten,
buried in 1950s vintage ARRL Hadbooks :-)
The technique is know as the three signal method. Beat the 10MHz from your
counter with the 10MHz WWV (we dont have HF standards broadcast in Europe
any longer) but put the BFO on. As you get close to zero beat the result is
a swelling and fading of the BFO note. This can easily allow calibration to
better than 0.1Hz at 10MHz. This is really only as far as it is worth going
because the random phase shift due to the ionospheric propagation will be
worse than that. The S-meter idea will work to some extent but is prone to
disruption from fading levels of the WWV signal, the three signal method is
not affected.
If you happen to have a local good standard then the way to do it easily is
to use a scope and trigger the scope from the standard, and time adjust the
display of the "unknown" for minimum "slip". Then by timing the phase slip
with a stop-watch it is possible to get calibrations to a few parts in 10^11
(if your standard are counter timebase are that stable) I have a 10MHz
source locked to a UK BC station at Driotwich (198kHz) whose frequency is
measured twice a day by NPL and posted to their web site. You can do the
same very easily with WWVB on 60kHz. The LF signals are much more phase
stable in daytime than the HF signals are. You may have to divide the 60kHz
down and do the comparison at 10Khz which can be a bit time consuming,
though there are ways round this.
Alan G3NYK
http://www.btinternet.com/~alan.melia/freqmeas.htm