impact of phasenoise on demodulation

[afz23]

We often see phasenoise being represented in dBc/Hz or deg rms.
I am not very sure ,how to use these phasenoise values to find out whether
the achieved phasenoise at a transmitter output is sufficient for an authentic
or error free demodulation at receiver end? I believe these values are more
stringent for say 8-psk than bpsk demodulation.:?:

Experts please throw some light on this.

please consider some examples to explain this matter.

 

[chuckey]

Any phase noise on a carrier will be put straight will be present if the carrier is demodulated using any phase sensitive method. So with the 8-PSK, each modulation step is 45 degrees, so if the phase noise is 25 degrees (> .5 X mod phase step), the demodulator cannot detect if it is step1 + 25 degs, or step 2 - 25 degs. With BPSK, the phase difference between the steps is 180 degrees, ie phase 0 or phase 180 is sent, so if putting in a +- 25 degrees, still gives a large band of phases which can be detected as a 0 phase or a 180 phase.
Frank

 

[afz23]

Thanks chuckey for explaining this,
What will happen,if I have a analog
Phase modulated signal with modulation
Index of say 1 radian(57 deg).,if phase
Error due to phase noise is 20 deg rms,
How it will effect phase demodulation
Practically or mathematically can u please
Explain this.

 

[chuckey]

If your demodulator produces 1V for 57 degrees, i.e. 57 degrees is actually in phase with the local mixer, corresponding to Sin 90 degs (1!). If the phase noise is 20 degs rms, this corresponds to +- 10 degs rms, or Sin 80 degs (which is the same as Sin 100) = .984 V, so the output signal will have a 1 - .984 = .016 rms noise.
Frank (best guess?)

 

[afz23]

That's great explaination chuckey!

I am generating a microwave carrier using
PLL synthesizer, which has phasenoise resulting
in undesirable residual PM of this carrier.

As this carrier is phasemodulated by a baseband
PSK signal,I have a concern due to this phase error
due to phasenoise of carrier.

I want to find out the value or specification of this
phasnoise generated phase error which can
create unacceptable errors at the demodulation
end.

Please tell the parameters ,I should consider
to arrive at this value as the datasheet specs
of 2 deg or 5 deg rms phase error are
very stringent and difficult to meet.

Can some one help in simulating this
concept to get better understanding?

 

[chuckey]

Its really complicated because at the receiver end there will be another phase locked loop (or two) which is phasing up the local phase reference to check the incoming PSK phase against, which involves the lock up time v. the noise bandwidth on the various loops. I got involved with this many years ago with buying a Frequency Synthesizer as a substitute for the reference oscillator in broadcast colour TV transmitters. The colour sub-carrier has a phase specification on it. In the end we actually used one and measured all the demodulated TV parameters and it proved to be OK.
FYI , This was in 1982 and we should have bought 23 X crystal oscillators at £700 each ~ £16,000, the synthesizer cost ~ £2,500. The additional benefit was that it supported other transmitters during a period when a big aerial array caught fire and put off its TV transmitter.
Frank