LO phase noise is almost always what limits the noise floor at close-in offsets, because of the narrow RBW (either analog or digital) typically used at those offsets to keep the carrier out of the measurement.
Occasionally a high degree of RF attenuation might raise the equivalent front-end noise floor high enough to dominate the LO's inband noise, but that's not usually what happens. Certainly not if you're using a strong carrier to characterize the analyzer's PN floor. Phase noise at 100 Hz would likely be dominated by the reference, either the original 10 MHz reference or a higher-frequency crystal oscillator locked with a very low bandwidth. As in Bruce's example, if the first IF is 2 GHz and the LO is tuned to 3 GHz to receive a 1-GHz signal, there's a 50 dB penalty on a 10 MHz reference (20*log(3000/10)). A decent 10 MHz OCXO is good for about -150 dBc/Hz at 100 Hz from the carrier, so you would expect to see about -100 dBc/Hz at 100 Hz from the carrier on an analyzer built with this sort of architecture. That's a good match for what you see in the PSA data sheet, as well as the FSU67's. The 8560E portables are about 5 dBc/Hz worse. -- john, KE5FX > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Behalf Of Luis Cupido > Sent: Friday, August 15, 2008 4:34 AM > To: Discussion of precise time and frequency measurement > Subject: Re: [time-nuts] Designing and building an OCXO and GPSDO > > > Bruce, > John, > .... > > And at smaller offsets like 100Hz and less ? > Shouldn't the improvement be even bigger ? > Closer to the carrier we are dealing with bigger signals > so the ADC issues like resolution should be less important, > and the limiting factor should really be the the > phase noise of the LO's and etc. > Am I right ? > > > Luis Cupido. > ct1dmk. > > > > Bruce Griffiths wrote: > > > Luis > > > > The R+S FMU36 has a phase noise floor of around -143dBc/Hz (offset > > > 10kHz) with a 10MHz input. > > Whereas the R+S FSU67 has a phase noise floor of around -133dBcdBc/Hz > > (offset = 10kHz) with a 640Mhz input. > > > > There is a definite improvement at lower frequencies but not quite as > > much as one might have expected. > > > > Bruce > > > > _______________________________________________ > > time-nuts mailing list -- [email protected] > > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > > and follow the instructions there. > > > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
