[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: drsstc - catastrophic failure! And Thor



Original poster: "Stephen Conner by way of Terry Fritz <teslalist-at-qwest-dot-net>" <steve-at-scopeboy-dot-com>

At 07:57 29/04/03 -0600, you wrote:
>Original poster: "jimmy hynes by way of Terry Fritz <teslalist-at-qwest-dot-net>" 
><chunkyboy86-at-yahoo-dot-com>
>
>Hi,
>
>One of the coolest things about OLTCs is that they are so simple. 
>Designing a resonant power supply would ruin that aspect of it.

Thanks for bringing me down to earth...

>You could easily convert it to a boost converter instead. If I remember 
>correctly, the IGBT you killed was because of this "problem", when you 
>weren't quenching at the first notch. With lower charging inductance, 1kv 
>should be easy.  The peak current would not rise too much either.  unless 
>the inaccuracies are really bad, you could just add a few extra IGBTs that 
>you would have used in the resonant power supply.
>
>Do you know what is causing the inconsistant charging voltage? how bad is it?

Due to imperfect tuning, and voltage induced back into the primary from the 
secondary, the tank cap never discharges completely. There is usually about 
150V left in it. So instead of recharging to 2*Vdc, it charges to 
Vdc+(Vdc-V left over). Unfortunately, on the first bang of a run, the tank 
cap _is_ empty and therefore charges to the full 2*Vdc. So if I set it to 
be near the limit in normal operation, the first bang would blow it.

It works the other way too: pulling an arc to ground detunes the secondary 
badly, which makes the charging voltage fall even further. It's interesting 
to think that these effects may all be at work in conventional coils too.

The boost converter action is useful, but it depends on the on-time, 
therefore adjusting the quenching changes the charging voltage too. The 
lower the inductance, the more sensitive it'll be. That's an interaction I 
could be doing without. I could always design for "more than enough 
voltage" and use transzorbs to clamp the excess.

>Are you planning on using the same secondary and toroid?

I think the new IGBTs will deliver around 4J/bang and I doubt that little 
4.5" x 9" secondary would survive long. I'm assuming that the 600A bricks 
can be overdriven to at least 6000A pulsed, however I don't know if this is 
possible. They might not behave the same way as discrete IGBTs. At least 
I'll have a bit of fun finding out!

Steve C.