kb8tq@n1k.org said: > If the main use is feeding test gear (and not direct synthesis) an Rb may do > pretty well. Most instruments assume a dirty reference signal and clean it > up internally. What's the bandwidth on the typical cleanup PLL? How well does that match the noise from a Rb? -- These are my opinions. I hate spam.

Hi > On Sep 23, 2017, at 9:39 PM, Hal Murray <hmurray@megapathdsl.net> wrote: > > > kb8tq@n1k.org said: >> If the main use is feeding test gear (and not direct synthesis) an Rb may do >> pretty well. Most instruments assume a dirty reference signal and clean it >> up internally. > > What's the bandwidth on the typical cleanup PLL? How well does that match > the noise from a Rb? It’s like any other PLL with noise involved. You look at the noise of the Rb, the noise of the particular cleanup oscillator, and your system requirements. In some cases that comes out to a few seconds. For Time Nut grade with a really good OCXO you are at some pretty long numbers. If the Rb starts at 2x10^-11 at 1 second and goes down by square root tau: At 100 seconds you are at 2x10^12. Still not as good as a really good OCXO. At 10,000 seconds you might get to 2x10^-13 (but probably will not). Your OCXO likely will climb out of the “parts in 10^-13” range before 10,000 seconds. Lots of fun. Bob > > > -- > These are my opinions. I hate spam. > > > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there.

Hi, On 09/24/2017 02:30 PM, Bob kb8tq wrote: > Hi > > >> On Sep 23, 2017, at 9:39 PM, Hal Murray <hmurray@megapathdsl.net> wrote: >> >> >> kb8tq@n1k.org said: >>> If the main use is feeding test gear (and not direct synthesis) an Rb may do >>> pretty well. Most instruments assume a dirty reference signal and clean it >>> up internally. >> >> What's the bandwidth on the typical cleanup PLL? How well does that match >> the noise from a Rb? > > It’s like any other PLL with noise involved. You look at the noise of the Rb, the > noise of the particular cleanup oscillator, and your system requirements. In some > cases that comes out to a few seconds. For Time Nut grade with a really good > OCXO you are at some pretty long numbers. > > If the Rb starts at 2x10^-11 at 1 second and goes down by square root tau: At > 100 seconds you are at 2x10^12. Still not as good as a really good OCXO. At > 10,000 seconds you might get to 2x10^-13 (but probably will not). Your OCXO > likely will climb out of the “parts in 10^-13” range before 10,000 seconds. > > Lots of fun. Once up on a time, you could order the OSA cesiums width different oscillators, essentially to buy into different balances between oscillator cost and performance. The time constant is adapter with regards to oscillator type, as the trade off shifts. I can't recall any other cesium or rubidium that did this, but HP upgraded oscillators at one point for this reason too. Cheers, Magnus