[time-nuts] OCXO Oven design (was: E1938A phase noise improvement)

[Attila Kinali]

On Sun, 26 Dec 2021 13:03:24 -0500
Bob kb8tq <kb8tq at n1k.org> wrote:

> The TC of the E1938 as reported in Rick’s papers is way past what
> a “normal” single oven will do (like by several orders of magnitude).

Yes, but the oven itself is also a much better design than
almost anything out there. Actually, I think that most
single oven OCXO designs out there are rather crude[1], when
it comes to temperature control. Just using the lessons
learned from the E1938 design (i.e. to deliberate control
where the temperature gradients will lie) would improve
many of the OCXO design out there. And that's something
that could have been done without infringing upon the E1938
patents at all.

The only other oven design, that I am aware of, that does
something similarly is the 860x inner oven design from
Oscilloquartz. But, as they never mention that they control 
gradients, I am not sure whether they were aware of that issue
or whether that's more an outcome of design optimization.

While we are at it: One of the key points of the E1938 oven
only mentioned in passing[2] is that lateral heat flow,
i.e., heatflow parallel to the designed equi-temperature planes,
needs to be zero. If it isn't zero, then the equi-temperature
planes will bend towards the upper or lower ends of the cylinder
and there would be a net temperature gradient over the crystal.
The two ways how the E1938 achieves this is by using "finite
guard rings" with a gap between the main oven mass and an
additional heater ring along the rim of the oven to compensate
for the heat lost there. Both are very effective, though not
the only ways to achieve this.

If this "keep the parallel heat flow at zero" would be applied
to more OCXO oven designs, that would already improve them quite
a bit, without the need to go full E1938.

			Attila Kinali

[1] E.g., the FE-405's inner oven are two resistor next to the
can of the crystal (see attached picture stolen from Bert Kehren).
There is not even an attempt to control the gradients or make
the temperature over the crystal related to the temperature at
the thermistor.

[2] "The theory of zero-gradient crystal ovens", by
Karlquist, Cutler, Ingman, Johnson and Parisek, 1997
http://www.karlquist.com/oven.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
-------------- next part --------------
A non-text attachment was scrubbed...
Name: FE-405B_Osc-3.jpg
Type: image/jpeg
Size: 525297 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20211226/ad806094/attachment.jpg>