On Jan 3, 6:05*pm, "Michael A. Terrell"
wrote:

Windmill wrote:
Long ago, magnetrons (of a different design from the ones still used in
microwave ovens) were the only way to generate high peak powers, and
the oscillation frequency of a magnetron wasn't very stable.

* *Not that they weren't stable. *They were powered with unfilterd DC,
which caused a wideband output. *They are operated as a self excited
oscllator when used in a microwave, and as long as the thisng is in
band, the frequency or bandwitch doesn't really matter. *Have you ever
read the MIT Rad Lab series of declassified W.W.-II books, or Slotnik's
RADAR Handbook?

Isn't the frequency determined by the dimensions of the cavity the
energy is dumped into? Or does that tune it too broadly?