"John Smith" wrote in message
...
... if I was to attempt to jam that broad of range of freqs, I would
couple it to a waveguide which exhibited resonance on that whole spectrum
(feed the center of circular guide/"modified horn" with shortest point to
edge of circle resonant at 2.4 ghz resonant, longest point to edge of
circle resonant at 800 mhz) , feeding the magnetron a ramp or triangle dc
voltage/current of 1 mhz freq--is going to generate harmonics until the
cows come home... I would feel like Dr. Frankenstein when the neighbors
showed up on my door step with their pitchforks and scythes!!!

Warmest regards,
John

Not sure how the magnetron would respond to a ramp, but at some amplitude
the magnetron would cease oscillations. The fact is that the harmonic decay
of a sawtooth waveform is slightly higher than a rectangular pulse. A worst
case analysis would involve a rectangular pulse with a finite rise/fall
time. If for example the magnetron were pulsed with a square wave, 50% duty
cycle, at 1 MHz, with a rise time of 10nS, the 1,600 th harmonic (i.e.
lower sideband at 800 MHz) amplitude is about -50 dBC. This result is
obtained from the solution of the products of two Sa(x) functions involving
the rise time, period, and pulse width of a symmetrical trapezoidal
waveform. If you have a burning desire to know, I can vary the parameters
of rise time, pulse width, etc. to see how the spectral shape changes.

These sidebands, in the vicinity of 800 MHz, are therfore of significant
amplitude. I am, however, fairly certain that 800 MHz is below the
waveguide cut-off frequency. I will try and measure the dimensions of a
typical oven waveguide to determine its cut-off frequency. The other factor
is the Q of the magnetron cavities how would this effect the banwidth of the
spectrum? Since the magnetron is coupled to the waveguide from a probe in
one of its cavities, I would think that the higher sidebands would be
significantly reduced. Possibly you have made measurements on such a
system.

The math, at least, does indicate the potential of building a very wide band
jammer -- probably not a good idea!

73,

Frank