I know it would require a decent oscillator to fully appreciate the exact division. About 100ppb accuracy to be comparable to a change of 1 to the divider, right? There seem to be a few 100ppb oscillators that aren't
prohibitively expensive. And there is the odd possibility I pick something used but interesting off from ebay... Also aging is a (progressive) absolute offset? So the ability to make small adjustments to the division factor would come handy to offset that. Anyway 10MHz sources
do come in more interesting varieties.
My custom clock idea is to have 64 "seconds" to a "minute", 64 "minutes" to an hour (regular), 24 hours a day. or in octal that is 100 "seconds" to a "minute", 100 "minutes" to an hour, 30 hours a day, or simply put 300000 "seconds" a day. I call it my octoclock,and I think it makes great sense
It's also fairly intuitive to translate to into the regular base 60 clock.
But I also want it to be general enough to play around with other clock divisions schemes. The lacking factor 2 in my particular case is what has been bugging me the most.
I think I wouldn't want to muck around with building my own PLL at this stage.. I got myself a couple of CDCVF25084 to play with. They are a bit pricey, and they don't want a too low input frequency. I'll put those in the list.
I think the kind of jitter with my vibrator-multiplier rises wouldn't be really bad since it's only every other pulse that is a wee bit off. Adjustment would be needed, probably using an oscilloscope. I think that is something acceptable in a device obviously aimed at enthusiasts
10MHz input -> Xor gate with a small RC network between the 10MHz input and the other input of the gate to produce ~25ns of phase lag = 20MHz output (You could even use the propagation delays thru the other gates in the xor package to get you the delay you need)....
This sounds really interesting. I need to experiment a bit with that. Just the kind of idea I was hoping for, something I would never have thought of myself