How master clock oscillators work

[EC8010]

Thanks for that. It's time for bed now, but I will read and inwardly digest tomorrow.

[Elso Kwak]

Originally posted by andrei

Take a closer look at the phase noise measurements of Linrad oscillator - they are exceptionally good. Elso, if you have measured the phase noise of your clock, I would love to see it, especially if you think they are better.


I guess I am enjoying an institutional subscription... Here is the direct link to the paper.

I only listen.....
Guido measures too......

[Elso Kwak]

Colpitts have superior phase noise according to the literature

[andrei]

Originally posted by Elso Kwak
I only listen.....

Even if you design your stuff by ear, there is no excuse not to measure the end result (even if it is just to try to understand why it sounds good). Unless you cannot do it, of course...

[Elso Kwak]

Originally posted by andrei


Even if you design your stuff by ear, there is no excuse not to measure the end result (even if it is just to try to understand why it sounds good). Unless you cannot do it, of course...

Oh there we go again.
Why, understanding that comparator B sounds better than comparator A?
And I don't have the money for that kind of equipment......

[EC8010]

Originally posted by andrei
Unless you cannot do it, of course...

Sadly, the equipment that enables jitter or phase noise measurement is horrendously expensive - that's why when I first looked into this I had to attack the problem from theoretical first principles, knowing that I would not be able to measure, only listen. I did come up with a circuit involving a four-quadrant multiplier for amplitude stabilisation and low impedance drive to the crystal but I never got around to building it properly.

[Bernhard]

According to some papers, jitter causes nonharmonic sidebands, so jitter could be detected indirectly with other equipment { [ ( ) ] }

[andrei]

Originally posted by Elso Kwak
Oh there we go again.
Why, understanding that comparator B sounds better than comparator A?
And I don't have the money for that kind of equipment......

There we go again indeed... You do not need expensive equipment to measure phase noise - read about it at Wenzel's website. I imagined that as a resident clock expert, you knew these things.

Originally posted by EC8010
I did come up with a circuit involving a four-quadrant multiplier for amplitude stabilisation and low impedance drive to the crystal but I never got around to building it properly.

Check out this oscillator circuit which features amplitude stabilization to produce high-purity sine wave. Maybe it's something like you had in mind. OTOH, high harmonic purity does not necessary imply low phase noise floor...

[poynton]

Originally posted by andrei


There we go again indeed... You do not need expensive equipment to measure phase noise - read about it at Wenzel's website. I imagined that as a resident clock expert, you knew these things.

True ... but you do need a reference oscillator !

[andrei]

Originally posted by poynton
True ... but you do need a reference oscillator !

No, just run two samples of the same oscillator - random phase noise from both will add. Even PLL probably can be omitted. Just feed two free-running oscillators into a mixer followed by a low-pass filter/buffer, and inspect the resulting beat spectrum with a sound card/FFT software.