[time-nuts] Low phase noise affordable GPSDO

[Bruce Griffiths]

The PN of the output amplifier in the thunderbolt will limit the PN nois of the 10Mhz output to a value significantly higher than the actual PN at the output of a low PN OCXO.

Bruce

> On 21 June 2018 at 09:26 Attila Kinali <attila at kinali.ch> wrote:
> 
> 
> On Wed, 20 Jun 2018 14:27:35 +0000
> Ethan Waldo <ewaldo at healthetechs.com> wrote:
> 
> > I've tried my best to scan through the archives to see if this has already
> > been answered, so apologies if I missed something relevant.  I am going to
> > be experimenting with >10Ghz microwave frequency grounded coplanar waveguide
> > phase combining/cancellation/detection across a number of generated
> > signals.  That being said, I am in the market for a GPSDO that has the least
> > amount of phase noise.
> 
> Phase noise is usually not the strong traits of GPSDOs.
> That said, I guess a Trimble Thunderbolt or Nick Sayer's GPSDO with
> the OCXO option would probably fit your bill.
> 
> Alternatively, take any of those two GPSDO and replace the OCXO
> by one that has as low phase noise as you need.
> 
> 
> > A secondary factor in my choice is how many 10Mhz references the GPSDO
> > can provide to sync all my signal generators; not as concerned about what
> > amplitude is provided.
> 
> You are looking a bit naively at this. The noise floor of modern,
> low-noise OCXO is limited by the Johnson noise of the 50R impedance.
> Ie the higher the output signal level, the lower the noise floor (in dBc).
> Hence, just using some passive splitters will increase your noise
> levels. The 1/f^a noise levels are unaffected by this, though. 
> 
> I would recommend going for a seperate distribution amplifier in order
> to keep the noise at a minimum. If you are fine with DIY, go for the 
> design by Anders Wallin.
> 
> 
> 			Attila Kinali
> -- 
> <JaberWorky>	The bad part of Zurich is where the degenerates
>                 throw DARK chocolate at you.
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.