[time-nuts] PM-to-AM noise conversion

Bruce Griffiths bruce.griffiths at xtra.co.nz
Thu Jan 17 19:45:43 EST 2008


John Miles wrote:
> Hmm, I didn't know that.  Wouldn't that effect have suppressed the 2nd lobe
> almost entirely on the baseband measurement, though, rather than making it
> appear 6 dB too high?  I should probably repeat that measurement to make
> sure of what I'm doing.
>
> Much appreciate the advice, by the way!
>
> -- john, KE5FX
>
>   
John

In a linear system the corresponding USB and LSB sideband component
amplitudes are identical so that on translation to baseband by the phase
detector all the even harmonics cancel and the odd harmonics add.
However in a non linear system the USB and LSB components will not have
equal amplitudes. Assymmetric filter passband response will also
attenuate one sideband with respect to the other.

If the phase detector LO phase shift isnt exactly 90 degrees wrt to the
carrier the even harmonics no longer cancel exactly.

In an I + Q demodulator FM USB and LSB components have the following
characteristics:

I channel:

for odd harmonics (of the modulation frequency) USB and LSB components add

for even harmonics (of the modulation frequency) USB and LSB components
cancel.

Q channel:

for odd harmonics (of the modulation frequency) USB and LSB components
cancel

for even harmonics (of the modulation frequency) USB and LSB components add.

Bruce




More information about the time-nuts mailing list