Original poster: Terry Fritz <vardin@xxxxxxxxxxxxxxxxxxxxxxx>
At 01:12 PM 8/25/2005, you wrote:
......... But you guys tell me. What are your thoughts.
I'm open ears.
I have not done a lot of real math on this, so the numbers and
speculations are just "pie in the sky" right this moment...
But assume that IGBTs fail due to over current, over voltage, and/or
over power... Probably a good assumption :o))
But consider the secondary power arc to ground... The secondary
suddenly has a low resistance (lower loss) path to ground... The
"power" has to go somewhere"... It goes to the streamer, secondary R
losses, primary R losses, and IGBT losses... (maybe "lost" IGBTs ;-)))
The streamer losses probably go way down even though the arc "looks"
brighter... I^2R, but R to ground is many orders of magnitude less
now... I^2 probably can't keep up... Power gets thrown "elsewhere"...
Secondary pure R losses are higher, but still not very significant
overall... The secondary coil is still far from significant heating
effects even with a couple of thousand watts input...
Primary R losses... I have not noted a primary coil/cap ever getting
red hot, or even warm in this case...
IGBT losses... Loud explosion with hot white flames... Hmmmmm!!! A
weak spot!!!
But, I think it is a "primary circuit Q factor" thing... When the
secondary shorts to ground through an arc - voltage, current, and
power "buildup" in the primary loop seems to be the "killer"... Just
a "little" primary resistance will drop the primary Q like a
rock!! I guess I find it hard to imagine that "skinny wire" has more
R than IGBTs... But IGBTs are the masters of "low R"
silicon... Added resistance her might place a "ceiling" on primary
loop over stresses...
So when the coil is making pretty streamers, primary R is just not a
big deal. But with a shorted secondary, primary R might make all the
difference in the world as a place to find "low Q" to protect the
IGBTs... Primary to secondary coupling must be a BIG factor here too....
There may be a "Primary Q, Secondary Q, and Coupling" equation out
there, somewhere, that would govern all this... I think it is a "two
hump" thing so things would get funky...
"That Teslathon" is probably going on right now, so interesting to
here the "results"???...