[sci.electronics] magnetron tubes

ESV@psuvm.psu.edu (05/08/91)

Ok here's the situation, I have a microwave oven that I does not
work.  I opened up the case and checked the filament on the magnetron
tube; it turns out that there is a low resistance path from both terminals
of the filament to case ground, which is also where the cathode of the tube
is connected.  The resistance is so low that my meter cannot measure it, i.e.
less than .1 ohms.  It seems to me though that if the tube were to go bad,
the filament would be burned open, not shorted to ground.  At this point
I hypothesized that perhaps metal from the filament had plated the envelope
of the tube sufficiently to create a path from the cathode to one of the
terminals of the filament.
I spoke of the situation to an EE friend of mine who said that he heard
of television repair people using large capacitors and about 300v to
blast the plated metal off of television tubes that had a similar problem.
The procedure is to wire both terminals of the filament together and charge
up a 300v capacitor (like a photo-flash cap.) and discharge the capacitor
from the filament to the cathode/ground and blast away the metal with
several of these pulses.
Now, since my strong suit is computers and not microwave technology, I thought
that rather than donning my microwave-proof suit and go mucking around with
semi-high energy devices that I don't understand, I'll put on my flame-proof
suit and consult the collective wisdom of the net.
So here goes:
Is the above about discharging a photoflash cap across this thing pure
bunk or a possible solution?  Do magnetron tube normally die this way or
do they do other more interesting things for death spasms?  If I do try
the above method, do I need to worry about emmisions from this thing
being extremely hazardous to my health?  I intend on not removing the
tube from the microwave so if it does start emitting lots of interesting
stuff, it should be directed into the microwave cavity and not at me.

Well, thats all for now, what do you think?


Andrew Coll       ESV@PSUVM.BITNET
                  ESV@PSUVM.PSU.EDU