[Amps] fully-submersed oil cooling -

[Steve Wright]

On 11/04/15 01:52, Dr. David Kirkby (Kirkby Microwave Ltd) wrote:
>
> I have never used it, but I believe FEMM could allow you to determine
> the extra capacitance at DC, which might be useful to know.  I suspect
> that it is fairly negligable compaded to that inside.
>
> [....] but I have tubes with dissipstions from 250 W to 5000 W here. 
> I could measure C than stick in a bucket of water and measure it again.
>

Hello David,

Many thanks for your engagement.

It seems the extra capacitance in question is relative to the metal
surround of the tubes, and not relevant in (near) free space.  My
suggestion is, the bucket of water experiment would net us little.  I
might be wrong however, and if I am proved wrong, I am better off.

K of the transformer oil is about 2, so the rhetorical question is, will
this K of 2 double that extra capacitance that the metal surround
added?  Theory suggests the answer to that question will be yes. 
Perhaps we can use this theory as a given?  Would it make that extra
10pF into 20pF just by adding the oil?

I am persuaded the amp will operate.  In any case, if it kills any
opportunity for 6m then I accept that.  If I lost 10m as well, then
maybe that is the basis for further experimentation.  If, in the worst
cast, it gets converted to low-band amp and lives its life like that,
all is not lost.  All the oil bits can go in the bin, and that aint bad
either.

Original paragraph that sparked this discussion:  Lost author, apologies..

>     ### The problem with oil immersion...which is used on some eimac
>     tubes per the
>
>         Eimac catalog is..... if you install a metal tub surrounding
>         the tube... you now have
>         stray C from the enclosure to the the anode....which can
>         affect the PI net on the
>         upper bands.   RF parts used to sell a metal chimney..with a
>         teflon ring on top, for
>         the 3CX-3000A7 tube..for air cooling.  It added 10 pf of C.
>         Tube is 24 pf from anode to grid, measured on a wooden bench.
>         And when inserted in the socket, rose to 33 pf.
>         That extra 9 pf is caused by the proximity of the lower anode
>         fins to the chassis below.
>         Once the metal chimney was installed, it rose to 43 pf.
>


Steve
ZL1BHD