[time-nuts] 74AC gates phase noise
dk4xp at arcor.de
Thu Feb 25 23:12:28 UTC 2010
Rick Karlquist wrote:
> I think I could approach this noise floor by taking a 10811
> and bringing the signal directly out of the grounded base
> buffer, and then readjusting the ALC to double or triple
> the crystal current. The biggest unknown is at what point
> the crystal sustains damage. The 10811 is definitely not
> optimized for phase noise floor. The designers were well
> aware of this. But phase noise floor isn't everything.
And, if a crystal is rejected for phase noise at 10 or 100 Hz,
that does not mean that it could not be used as a post filter
for offsets > 1KHz. It even could be somewhat damped so that
its own phase noise does not kill the close-in phase noise performance.
And we could hit it with more power. If it runs away by 100 Hz, so what?
All we would expect from a second crystal is another 10 dB far-off.
And I think there are a lot of leftover crystals; the really good ones,
close-in, are few.
Did I forget a show stopper b4 I try it?
Magnus, I havn't forgotten about you. I don't think that
my locking scheme will work for your beacon application, but
the VCXO might. Maybe I'll have to give up that
"no microcontroller, no FPGA, no CPLD, just programming resistors" hybris.
More information about the time-nuts