[ARC5] What's With 2,000 Volt, 50-Watt Xmtrs??

[David Stinson]

I've been looking at some transmitter diagrams 
that use tubes like the RCA 803 and its kin.
Take the RCA AVT-12:  This is a simple, late '30s
xtal-osc-power-amp transmitter intended for 
small aircraft which delivers about 50 watts out on phone.
The high B+ for this thing, supplied by an external 
dynamotor over a long run of cable, is like 1700 volts.
Why?  RCA designed similar rigs about the same time
that delivered that level of power with 500-600 volts.
Get a GP and start thinking about 2000+ volts to 
deliver less than 100 watts.  That kind of voltage
in the dirty, greasy, gas-vapor-y environment 
of a 1930s-40s aircraft seems...ummm....
not well thought-out (read "stupid and suicidal").
Other period equipment demonstrates doing so was unnecessary.

Whenever someone otherwise brilliant pens a design
that makes as little sense as creating the 803 to deliver
modest power with Tesla-esk voltages, I smell money
at the root of it.  Did RCA tube designers build the 803 and
other tubes like it to get around someone's patents?
That would certainly fit-in with all the radio patent fights
in the 30s. How many different ways can you make a tube,
anyhow?  They'd have to do something different to avoid
royalties to someone, like design a 50-watt circuit 
with 2200 volts of B+. 

OK, People-Smarter-Than-Me:  What do you think was the 
motivation to design low-power-out circuits with 
dangerously high B+ voltages?

73 Dave S.