[time-nuts] Low Jitter / Phase Noise 10kHz Signal

[Bruce Griffiths]

HagaaarTheHorrible wrote:
> Hi there,
>
> I hope this is the correct way to post here and that its not been covered before. I've read some threads about frequency dividers, but didn't really find all the answers I'm looking for.
>
> I'm trying to generate a 10kHz sine wave thats as clean as possible, as it will be a reference signal to measure jitter. The idea was to use a 10 / 10.24MHz Crystal Oscillator (since they are easy to get) and divide it down to around 10kHz (exact frequency is uncritical). So I've been looking for low phase noise frequency dividers, and now have a few questions.
>
> 1. Will this way result in worse or better jitter/phase noise specs than generating a 10kHz Signal from the start? (meaning, is the additional noise added by the divider worth more or less than the reduction of phase noise by the division ratio)
>    
The problem would lie in finding a 1kHz resonator with sufficiently high 
Q to achieve low phase noise and hence jitter.
> 2. If the suggested 10MHz->10kHz way is the correct one, would a conjugated regenerative divider be the best way to get to 10kHz or are there less noisy alternatives?
>
>    
Conjugate regenerative dividers are currently considered the lowest 
noise technique but using a custom mixer perhaps using diode connected 
BJTs (eg 2N2222's or 2N3904's) will be required for frequencies below 1MHz.
CMOS dividers can have reasonably low PN floors but a properly 
configured ECL divider can have lower close in phase noise.
You could try using something like a 74AC4040 driven by a 10MHz input 
signal.
However lower phase noise can be achieved by using cascaded decade (or 
divide by 16) dividers with input bandpass filters (reduces aliased 
input noise) for each stage.
> 3. I've read through several threads about the divider designed by David Partridge, havnt found any real specs for it though. Is there any measurement data available?
>
>    
Rough measurements indicate a cycle to cycle jitter of around 5ps rms or so.
Most of this is probably due to the input comparator.
Its difficult to measure such low jitter even a wavecrest 2075 (and its 
cousins) isnt really good enough.
Measuring phase noise at frequencies below 1MHz or so is outside the 
range of a Timepod or most mixers.
> Thanks in advance for any replies.
>
> Hag
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
>    
Bruce