"Jim - NN7K" wrote in message
m...
One consideration is that Magnetrons are NOT designed, normally for ccs
service-- they are designed for "PULSE" service, even the ovens- (duty
cycle of MUCH less then even .5) and, pulse transmissions arn't looked
favorably on many of the microwave bands any more.

Sorry if I am repeating stuff that you may have seen before, but I have
noticed things that do not seem to add up. For example the paper at:
http://www-personal.engin.umich.edu/...aes_tps04.pdf;
Shows the CW output of the magnetron spectrum as -10 dBm. Since for the
output is monitored via a 30 dB coupler, and a 30 dB attenuator I would have
expected the amplitude to be 0 dBm. It is hard to imagine such testing
being done with a grossly overheated magnetron. Having observed the output
of a microwave oven on a spectrum analyzer, it did appear to be pulsed,
although the spectrum was more characteristic of a frequency hopping (or
swept) signal.

Frank