For example I finished tuning a 6-cavity 10 GHz uW BPF that was so
amazingly sensitive to the adjustment screws position and contact to the
case it is hard to explain. Even the slightest change in the electrical
bonding of the tiny screws and their position changed the filters
properties drastically.
In the past I worked on certain sensors for nuclear SCRAM systems that
were hypercritical, to say the least. All sorts of training, and
certification through the company I worked for, the USGS, AEC/NRC,
traceability that would make your head spin... One calibrator was a
battery, power switch, and two wire-wound 10 turn pots. It took a week
to document any work on it, and the stack of paper was about an inch
thick.
Critical mechanical parts were secured with Loctite 380, you could place
a drop, it didn't migrate, and could be broken loose with an amount of
effort when service was required. I personally used it on some RF
connectors, and antenna fasteners. Pricey here, I'd shop around. I got
a couple bottles as free samples:
https://www.ellsworth.com/products/adhesives/cyanoacrylate/henkel-loctite-black-max-380-instant-adhesive-1-oz-bottle/
It's pretty strong, I'd try a pinhead sized drip across a couple of
threads for starters, if you go this route.
The ultimate end-adjustment was a single PCB mounted potentiometer that
would calibrate the sensors within one ten-thousandth of a G. That was
held in place with GLPT, brushed across the top of the body, and the
flush slotted shaft, filling in the slot, and slightly adhering the
shaft to the body to prevent changes in adjustment due to vibration, or
temperature fluctuations, but could be broken loose later with a
screwdriver:
https://www.jensentools.com/gc-waldom-10-9002a-red-glpt-insulating-varnish-glyptol-2-oz-bottle/p/388ae050
Kurt
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|