RE: SSTC problems1 (no its not a PlasmaSonic ;-))

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Malcolm Watts" <m.j.watts-at-massey.ac.nz> 

Hi Sue,
         I did a switching device/diode product inventory of my bits
and pieces over the weekend and dug out some facts and figures on
power MOSFETs I have, some of which might assist:

On 29 Jan 2004, at 21:28, Tesla list wrote:

 > Original poster: sebastian gaeta <sgtporky-at-prodigy-dot-net>
 >
 > Hi David,
 >
 > I built the full bridge Feedback coil at:
 > <http://dave.golfbuddys-dot-com/hvguy/>http://dave.golfbuddys-dot-com/hvguy/
 >
 > My coil needs a lot of improvement though. If I had it to do over
 > again, (and I AM doing it over again!) I would use the same driver
 > that they use for their 10 KW design, even for the P460s because it is
 > more rugged, and the parts are more available. I saw a scope picture
 > of someones gate waveform using that design, and it even appears to be
 > a faster switcher than the one I am using. If you look at my picture,
 > the gate drive transformer is not in a good place because the back
 > fets are too far away from it. My new one will have the GDT in the
 > middle of the heatsinks and as close to each fet as possoble. I also
 > put 10 ohm resistors in series with each gate with a schottky across
 > each one to speed up the fall times.
 >
 > I rewired my bridge rectifier to do half wave instead of full wave
 > because my mosfets get too hot. And half of them get hotter than the
 > others so my duty cyle stinks too. I played around some more last
 > night by powering the H bridge with 12VDC, and probing around with a
 > scope. You need something like 100 volts on the antenna to give steep
 > enough rise and fall times to create a decent enough square wave with
 > your 1N34 'chopping diodes'. If the feedback voltage is too small, the
 > duty cycle is uneven, and one set of mosfets work harder than the
 > other set. Just something to keep in mind if the same thing is
 > happening to anyone else. Rewiring the bridge was the last thing I did
 > before I went to bed last night, so I haven't even tried it yet with
 > half wave rectification. This should force things to run cooler even
 > if my circuit is messed up!
 >
 > A properly switching MOSFET should never even get warm because it
 > should never be dissapating power. When the current is max. the
 > voltage should be 0 and when the the voltage is max there is no
 > current! Power dissapation = 0 = I am doing something wrong! The
 > "beast09" does not look like a typical MOSFET switching waveform
 > though! If you look hard and long enough, you can almost make out a
 > square wave buried in there!

MOSFET "on" resistance is never zero although some types get close.
If it is the IRFP460's you are talking about, they had a somewhat
higher channel resistance than I would have expected so you will
still get conduction losses proportional to drain current squared.
In fact, the D-S voltages at high drain currents can easily exceed
Vcesat in similarly rated bipolar devices. The lack of DC base
currents (needed to drive bipolars) makes IGBTs a popular choice for
this reason.

Malcolm