[time-nuts] NavSync CW12 and CW25

[David]

(time-nuts Digest, Vol 49, Issue 14)

Ed, a couple of rather delayed answers to your CW25 thoughts:

> but if you changed the output frequency of the CW25 is it 
>possible that you might find a 'sweet spot' that would give
> you a cleaner signal

I've not experimented with that, it affects the simplicity of the synthesis
design and it may not be that important. The PLL that locks the OCXO to the
'GPS' reference will have a very narrow bandwidth and operates a little like
a LPF to minimize the transfer of 'GPS' phase noise sidebands & spurs to the
OCXO. Finding a useful "sweet spot" frequency needs an appropriate way of
dividing the OCXO down so the frequencies match in the PSD. In my system the
initial loop is a 0.3 Hz bandwith, this is rather loose to get things going.
It's now working and given some free time (Ha!) the loop bandwidth can
probably come down by about a factor of 10, this may further clean the
output signal.  

The most important consideration is what is happening very close to the
carrier rather than noise and spurs much beyond 1kHz from carrier which will
be heavily attenuated by the PLL.

> Did you divide down the output from the CW12?  

Yes. I initially thought about a discrete system and did not quickly find
the phase / frequency PSD I wanted (MC4044). I looked at modern monolithic
synthesizers and was put off many due to the need for some form of
controller to load and control them. I do have a bag of (long obsolete) PICs
which I use for this sort of thing but additionally many of the modern
monolithic synthesisers are not specified for frequencies as low as 10MHz
although they may work well. About this time my raking through the recycling
buckets found a few MC145152FN2 parts which are parallel load so just needed
some hard wire links to set the signal and reference divider ratios, easy.
The limitation in this application is that the reference divider is limited
to ratios of 2^N, so in my toy N=7, the division is 128 in both paths. 

Contact me off list if you want one posted to you, I suspect they have been
off the market for years.

At the moment the OCXO is happily locking down and vastly cleaner than the
CW25 output, given time I want to modify my 60kHz (MSF) time signal receiver
before I review the PLL in the GPS.

Regards
David