[time-nuts] Re: E1938A phase noise improvement

[Attila Kinali]

Hoi Rick!

On Sun, 26 Dec 2021 08:36:06 -0800
"Richard (Rick) Karlquist" <richard at karlquist.com> wrote:

> The E1938A has both an ALC loop (common in many oscillators) plus an AFC 
> loop, which is exclusive to the patented bridge stabilized circuit.  I
> am thinking that maybe the AFC circuit was causing the hump.  That
> theory could be verified by disabling the AFC and seeing if the hump
> goes away.  The AFC is not necessary just to make the oscillator play
> for noise testing.

May I ask a heretic question? Does the AFC actually improve
stability?

I looked at quite a few OCXO circuits (that are still in use)
over the past years. Of them, the E1938 is the only one that
tries to compensate changes within the oscillator at all
(besides ALC). Yet there are many other oscillators that acheive 
similar stability values as the E1938.

You hav written a nice description[1] of how the oscillator does
work and why it should be better, but there are no comparisons
of the ciruit with and without AFC. Or even a standard Pierce
versus the Balanced Bridge. Which for me raises the question,
whether the additional complexity is justified. Especially given
that the bridge circuit halves the available Q and thus increases
1/f^2 and 1/f^3 noise.

			Attila Kinali

[1] "A New Type of Balanced-Bridge Controlled Oscillator",
by Richard Karlquist, 1999
http://www.hpl.hp.com/techreports/1999/HPL-1999-6.pdf
-- 
The driving force behind research is the question: "Why?"
There are things we don't understand and things we always 
wonder about. And that's why we do research.
		-- Kobayashi Makoto