[time-nuts] re multiplication to measure phase noise

[Jeffrey Pawlan]

Yes, you can mulitply and I do that as well. But this a part of the measurement
techniques and not really part of the discussion at hand. Besides, the
multiplication scheme adds noise and you must calibrate that to be accurate.
If they do not have a means of phaselocking and comparing two relatively equal
oscillators at its fundamental frequency, then it does not matter how many times
they multiply it. Using a millimeter wave spectrum analyzer does not really
change this inaccuracy as the noise floor of the spectrum analyzer goes up with
each range multiplication so it does not allow you to measure a 10811 even at
40GHz.  You can't get "something for nothing".  You can definitely benefit from
the multiplication ONLY if you know exactly what the multiplier contribution is
and then use a phase detector at the mm-wave frequency and thus are getting
back to DC/baseband and not reading this on an RF spectrum analyzer.

I do not know whether your time spent measuring the phase noise of mm-wave
sources was for ham radio or for work. If it is for a very important work
project, then ask them to invest in the Poseidon (spelling?) oscillators from
Australia. They are using machined sapphire cavities suspended in liquid helium
as the resonant structures. These are in the 10-12 GHz range so you would need
to then multiply them up. There are lots of papers in the IEEE UFFC
Transactions.

73,

Jeffrey Pawlan