[time-nuts] Passive maser update 7 (cavity measurements)

[cdelect at juno.com]

Ed,

I've added our correspondence here to maybe get additional advice! :)

Ed: This is awesome to see. ItÂ’s great mind food.
 
The first question I have is whether the cavity resonance is really
shifting from the gas within, or whether itÂ’s simply being crushed by the
lack of air pressure.
 
If you were to put it into a 55 gallon plastic drum with a V band clamp
and simultaneously pump both down, and my theory is correct, the pressure
change shouldnÂ’t cause frequency change. You wouldnÂ’t be able to take all
of the atmosphere out, but even a 2% drop ought to be doable ought to be
easily observable. Letting air back into the drum would then show the
pressure shift. Or you could even pressurize the drum slightly and see
the shift increase.
 
My rationale is that 100 microstrains while seemingly tiny is not
completely insignificant with the precision you require.



Corby:   Thanks for the thoughts! 
It is not caused by pressure difference as the cavity is entirely within
the vacuum chamber.
 Vacuum inside and outside so there is no pressure difference to affect
the measurements.
 My main problem now is that before the varactor repair the cavity was
around 800kHz high at 22 degrees C.
 With that case as the temperature of the cavity increased to the target
of 50 degrees C the cavity frequency dropped to the proper value.
 With the varactor repair at 22 degrees C I am way to low, almost 900kHz,
so as the cavity heats it heads way from resonance! :(
 So I'm going to try a different varactor and add a tiny chip of alumina
to see if reducing the cavity volume would increase the frequency by the
proper amount.
 Probably am looking at several or many tries to see if this works.



Ed"
Using a series capacitor that has higher Q than the varactor would reduce
its capacitance, and tuning range. YouÂ’d need an inductor thatÂ’s open
circuit to the cavity frequency, which is easy.


What was the varactor youÂ’re using, and in what package style?


Corby: 
 Attached is a PIX of the varactor "loop".
 The electrodes on each side of the gap are flattened slightly.
 Originally the varactor was a tiny black epoxy blob on a tiny ceramic
base.
 No way to determine its part number. :(
 In the PIX you can see a tiny wire from the blob to the other side
electrode.
 Also there was a tiny rectangle of alumina epoxied across the gap on the
opposite side to mechanically anchor the loop ends.
 I removed the shorted varactor and also the alumina rectangle.
 I installed an SOD882 cased SMV1405 (2.67pf @ volts and .63pf @ 30V.)
that I had attached two tiny wires to.
 Next plan is to install an SC-79 cased diode that is about half the
capacitance and to also add back an alumina stiffener.
 Other thought is to wind a tiny wire around part of the loop to add
volume???
 Then I could unwind wire till I get to the spot I need.


Ed: 
Before worrying about any of that, perhaps rotating the coupling loop so
it's interacting less with the fields might be good.

I don't think changing the loop size without knowing whether it's close
to resonant frequency
If the thread screws in to stop against cavity wall, add a metal washer
as a shim that's only part of a thread thick.
Ultimately a -210 ceramic pill or a -240 ceramic disk package varactor
would be good in a newly fabricated stud mount with a tempered spring
steel hook to clamp it in compression (silver plated of course).
I get it you want to make it operate, but it would suck the tear one lead
of a beam lead diode.
Or maybe you could harvest a connector pin of the right diameter from a
D-sub connector or something to serve as a socket to hold the diode.
Less coupling is a good thing ... if you shorten the loop there's no way
add back length chopped out.

Corby: 
I think maybe adding that shim might be the way to go. After the next
varactor is installed and tested that might be the way to go!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: varactorloop.jpg
Type: image/jpeg
Size: 11254 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20230210/6f5dfc9e/attachment.jpg>