[time-nuts] Lucent RFTGm-II-XO / Lucent RFTGm-II-Rb / NT BW50AA

[Bob kb8tq]

Hi

The trick is that a good 125 MHz OCXO has better phase noise at 100 Hz than a 
multiplied 5 or 10 MHz OCXO. By the time you get to 30 Hz, usually the lower frequency 
multiplied up wins. By the time you get to 3 Hz, the much higher Q HF OCXO always 
wins. (Yes, you could have a great 125 MHz and an utterly trash 5 MHz where this would 
not apply. For the OCXO’s used in the big old telecom GPSDO’s it does apply).

This also *assumes* the noise floor in your phase detector and noise floor in your dividers
is not limiting what you do. That *probably* is true for most of the parts you would reasonably
decide to use. Again, something that is worth measuring on the device you build up. 

Now - in *your* application, how much does 30 Hz phase noise matter? Maybe it is a big
deal, maybe it does not matter at all. If it does not matter, then indeed do the GPSDO 
straight up with the 125 ( or 250 or …. ) MHz oscillator. It very much depends on what the 
system you are designing up requires. 

Bob

> On Feb 12, 2019, at 10:34 PM, jimlux <jimlux at earthlink.net> wrote:
> 
> On 2/12/19 4:41 PM, Jeff Blaine wrote:
>> OK.  I would still be interested to see if anyone has plots on the close in noise profile of the Lucents.  The XO part has an Efrom OCXO and I can assume it's similar to a generic OCXO.  But the RB is an unknown - I have no idea what to expect from that.
>> On a separate task, one of these days I would like to build up a set of 125 & 500 Mhz sources for lab and gear use disciplined by a GPSDO 10Mhz feed.  The current source is a 125 Mhz homebrew OCXO and a pair of doubler/amp/filter sets.  I assume they are pretty clean close in but the frequency moves around a bit more than I would like as the OCXO part of the equation is nothing special.  They drive some SDR and DDS sig gens here in the shack so the PN is a factor as well as the frequency accuracy.
> 
> what about disciplining the 125 MHz OCXO?
> or a really narrow PLL?
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.