Original poster: Terry Fritz <vardin@xxxxxxxxxxxxxxxxxxxxxxx>
Hi All,
I played with computer models today and had a few changes to the concept.
http://hot-streamer.com/temp/OL-DRSSTC-02.gif
1. I think bi-directional transorbes can replace the Zener/Schottky
pairs so only one device is needed instead of four on the IGBT gates.
2. The old trigger circuit did not work well, so I added a whole
new trigger circuit. It is a MOV in series with a SIDAC and a
current limiting resistor. The SIDAC could fire at say 20V and make
a nice sharp current pulse to start the system.
The separate trigger current loop wire is simply strung though the
CTs along with the primary wire.
I think to begin, one should just charge Cdrive (name changed from
Cresonant since that caused confusion with Cprimary) with a DC power
supply. The resonant charging thing is sort of trick and the other
issues need to be worked out first. The trigger circuit could also
be easily fiddled with for testing.
One problem modeling suggested is that it does not like to turn
"off". Once the oscillation is started, it tended to start pulling
current directly off the AC line. I think fiddling more with Lblock
can solve that.
Modeling did show that it "loves" to run and found no basic problems
with the topology.
I should not that this is all theory and has never been tested in hardware...
Cheers,
Terry