Re: LO Phase Noise and wideband front ends


Hi Chris - I remember once being taught about the low pass characteristic of kTB noise - as it must have because a flat characteristic would imply indefinite power at indefinite bandwith. 
A simple approximation has the cut-off at f=kT/h - somewhere in the Terahertz range. Google "Johnson-Nyquist noise"
73 Ole, OZ2OE

Den 14. juni 2016 kl. 22.33.34 +02.00, skrev Christopher Bartram cbartram@... [ukmicrowaves] :


Hello John
> Are you sure that amplifier noise is AM noise? I always thought that
> amplified thermal noise, which this is, was an equal mix of both AM
> and PM noise.

I suspect that this is a situation where on one level we're both right!
Of course, thermal noise will have both varying phase and amplitude
components, and at a quantum level, I believe thermal noise has a
low-pass characteristic. (Physicists please comment - I'm just a
horny-handed engineer... ) But when I say thermal noise or AM noise, I
refer to noise defined by kTB, which has no practical frequency
dependence. Phase noise, though, is generated by the action of an
oscillator, and can be thought of effectively as angle modulation of
that oscillator by a noise source.

It's possible to measure the amplitude and phase noise generated by an
oscillator independently - amplitude noise is surprisingly easy to
measure. Phase noise, though, is primarily dependent on the oscillator's
amplitude (kTB) noise and the Q of its resonator, and it's measurement
requires a bit of subtlety. I could give you a reference to Leeson's
classic paper on oscillator analysis and phase noise if you are interested.

The term 'phase noise' tends to bandied about without much thought -
that's true of both amateurs and many professionals - and the real
meaning of the term has become muddied in many people's heads.




Join to automatically receive all group messages.