[sdiy] Software to log frequency every x seconds?

[Paul Perry]

Gee, my $15 Q&Q watch from a stall at Victoria market has been better than
10ppm  over the last six months. I guess my wrist must have a pretty
constant temperature.

As for variation in oscillator freq, I wouldn't write off the possibility of
short term random variations. At lower frequencies (that is, 1Hz and down,
corresponding to periods of 1 sec & more) what is called 'flicker' noise
becomes a problem. And its amplitude increases inversely with frequency.
Hence the alternative name, '1/F noise'.

Where it comes from is a bit of a mystery - but it is analogous to eddying
in fluid flow & the fewer impurities there are in and on the semiconductor
structures, the less there is.
Whether it is a 'bad' or even a 'good' thing is something else, of course.

I certainly agree that a computer crystal is the last thing you would want
to use as a reference, but a standalone crystal clock of high precision is
not a big ask.

paul perry Melbourne Australia


-----Original Message-----
"Who checks the checker ?"
A crystal clock will drift with temperature, maybe as low as 10ppm, maybe a
lot more.
So, how stable is the 'reference' computer's timebase ?
Time is the most difficult parameter of all to measure accurately, hence
theoretical drift values on atomic clocks.