[Amps] load capacitor ratings?

[R.Measures]

On Aug 30, 2004, at 5:37 PM, Will Matney wrote:

> Rich,
> Exactly. I talked to the owner of Oren Elliot (Orens son Stephen) 
> several years ago about this and what he thought the rating should be. 
> He had several thoughts on this but told me that most were using a X2 
> rating the same as the coupling capacitor for tune C.

-   A x2 Vdc Tune-C (6kV) is what the TL-922 uses and the result is 
frequent loss of the bandswitch - which has a 5kV withstanding ability. 
  It is better to have a Tune-C flashover than a bandswitch flashover.

> He recommended using a load C with a voltage rating some higher than 
> the peak voltage rating across the load, generally having a 50 ohm 
> load as standard. I think he figured by a sometimes varying load, it 
> could take more voltage to create the same power across a mis-matched 
> load. Of course this could sure be the case and is something I would 
> not have thought of at the time. Not all antennas are a perfect 50 
> ohms or close to it. His capacitors are tested in open air with a 
> hipot tester. They had a jig made up with a nut driver handle made 
> onto a coupling to turn the rotor with. They applied the hipot and 
> turned the rotor through its rotation to catch any arcing. My concern 
> was the physical size of air variable capacitors and what was the 
> least size I could safely get by with. He did mention that one was 
> available where the plates had a "nylon dielectric" to insulate the 
> stator plates raising a .016" air gap to 1100 V (X2).

Will --
-  Nylon is a good DC insulator, but it has the worst D 
(dissipation)-factor I know of.  Another lossy RF dielectric is Delrin 
-- which is what Elliot used in their roller-coils a few years back.  
MFJ used one in their "3kW" tuner.  When this tuner was tested in the 
ARRL Lab, it caught fire with 900w of AØ/NØN.  An easy fix is to make 
the roller-coil form out of G-10, G-11, or ABS.

cheers
...