[time-nuts] Generating a stable 26MHz and 19.2MHz from 10MHz

Magnus Danielson magnus at rubidium.dyndns.org
Sat Jul 2 22:53:38 UTC 2011


Don,

On 07/03/2011 12:29 AM, Don Latham wrote:
> Hi Magnus: I goofed, it's actually 5.12 MHz, the driving crystal
> standard for the ICOM 260A 2 meter transceiver synthesizer. I want to
> use it with a transverter for moonbounce, hence would like to start out
> with a solid reference frequency. Might even be overkill. See, I knew it
> could be done with (almost) some 7400's :-)
> Thanks for the hint! The thing to do then is to pick off some of the
> existing xtal frequency, and as you say count to the common denominator.
> A little varicap in the crystal circuit can then be driven by the phase
> comparator...
> Should not be too hard to do...
> Don

So true. 10 MHz and 5,12 MHz is a little different. You need to divide 
by 125 to knock the 5's out of 10 MHz and get 80 kHz and then a divide 
by 64 from the 5,12 MHz if you choose to go the PLL route.

It should not be too hard.

For relative low feedback division rations it is fairly easy to achieve 
decent results.

Keeping the comparator frequency high ensures that the loop filter can 
do a decent enough job to keep the comparator frequency and overtones 
out of the sidebands without being forced into tight PLL bandwidths.

For a step-up PLL of low ratio the loop filter can be very simple and 
always lock up without too much difficulty. Even a passive lag can do at 
times. As always, shaping up things with an active PI loop filter is 
better. It all depends on the requirements.

Just make sure that you do not get a high Q cause that is a real downer.

I'm not saying that PLL is the best way to do things at all times, but 
for some applications the performance needed and applicability of a PLL 
makes it a fairly low hanging fruit, so just go for it. I guess that for 
your application it could very well work good enough.

Cheers,
Magnus



More information about the time-nuts mailing list