[TCML] SISG Repaired / Question about timing resistors

["caspina" <caspina@xxxxxxxxx>]

Hello,

My SISG is repaired.
Turned out that 7 of the 15 IGBT's broke down (it's a 15-module SISG,
1000V/module).
Fortunately, when IGBT's go to Walhalla they leave a short behind,
protecting everything else on the module.
To be sure I replaced all 15 IGBT's anyway.
Since it's a triggered SISG each module has also a SCR. They too are still
ok.

I think the IGBT's died because of over-heating, too high current and too
high duty cycle (playing music).
I also think they didn't all go at the same time, but one every now and
then.

The reason?
First I thought my charging choke arced internally (it's a DC-Resonant
charging coil),
but the choke is still ok.
Fact is that for my last runs I changed tank cap from 100nF to 125nF.
Maybe this was too much for the IGBT's... Resonant frequentie is 125 kHz.

So for the moment I use only tank cap=100nF and a big toroid (Fres=104kHz
now), and the coil sings again like a nightingale :^)

Because of this crash I plan to build a new secondary, a 10" instead of the
6" I use now, only to lower the resonant f, also lowering the current
through the SISG.


I have a question concerning the timing resistor of the SISG (the famous
R4).
Most use 4K7 like in the schematic of Terry Fritz.
I've been experimenting with it and use now 'only' 220 ohm for R4, in my
opinion the IGBT's don't need to be 'ON' longer than a few cycles (4-5).
I'm I doing wrong using such a small value?
Could it be bad for the SISG if there is no break-out and IGBT's are shut
off this early? (It happens occasionally).
I would think the energy is trapped in the secondary system and does not
harm the primary system...



-Michael,
 (from Belgium)
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla