Sam Goldwasser wrote:
Chris writes:

Nah, it was a half a stick, well within the pre-programmed range, so
that's not it. In any case, some bit of foil of just the right size
or shape can lead to gradients high enough to get sparks, but this was
not like this. This was some large fraction of the entire current
output of the transformer, like the fireworks you see on youtube when
people have pulled the guts out of the microwave and using it just to
make an arc.

This whole thing doesn't sound right. Regardless of whether the microwave
oven is attached to Earth ground, the return path for the high voltage
IS the chassis of the microwave oven. It would be almost impossible for
that to be disrupted as the magnetron and HV transformer are bonded to
the chassis. So, it's highly unlikely that the HV current (not the
microwave energy) would want to jump from the oven to an external ground.
Looking at the typical schematic, it's hard to come up with any sort of failure
mode where such a potential difference would appear between the chassis
and ground.

I'm not saying something very peculiar didn't happen. Just that an
explanation relating to the HV power doesn't make sense. I'd quicker go
with some combination of load (butter), dirt, and other factors affecting
the microwave distribution.

It could be possible that some part of the door, possibly one of the
hinges has lost it`s bonding to chassis earth and a potential is
building up there then striking to the nearest earthed metal. it`s
unlikely tho in my experience.

There was a model of Phillips oven that could build up a charge on the
thin aluminium cover over the lamp leading to quite spectacular arcing
around the rivet which held it in place and acted as a swivel. A poor
connection there was indicated by quite strong leakage of microwave
radiation through the plastic menu strip on top of the oven.

Ron(UK)