[time-nuts] PRS10 MDEV
chris at chriscaudle.org
Sun Oct 14 22:48:21 EDT 2018
On Sun, October 14, 2018 9:07 pm, Chris Burford wrote:
> I've collected several data sets over the last few days from my RFS
> and I'm puzzled by what I'm seeing.
Frequency stability measurements must always compare one device against
I do not recall that you mentioned what is providing the reference
frequency for your measurement. Is it possible that your reference signal
has some odd behavior that falls in that time range? It reminds me of the
behavior of a PLL crossing over between the noise behavior of the local
oscillator and the noise of the reference signal. Are you by chance using
a GPS disciplined oscillator for reference, which would have effectively a
PLL which crosses over noise behavior from the local oscillator at short
time scales to the GPS receiver behavior at long time scales?
For that matter a rubidium standard would have an OCXO providing the
output frequency, but it would be locked to the rubidium transitions at
long time scales, so even with a perfect reference signal you may still
see PLL crossover behavior in the output of a rubidium standard.
More information about the time-nuts