Re: Recent s.s.t.c work

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Steve Conner <steve@xxxxxxxxxxxx>


> Only by feeding back from the primary can one ensure 
> power-transistor switching near the primary zero-current 
> levels.  (And only by employing a resonant primary can one implement 
> a feedback-oscillator with the primary as its resonant circuit--so 
> that takes care of the resonant-primary question.)  Q.E.D....right?

That's exactly right. I started off with secondary current feedback, 
but I discovered that the phase shift between primary and secondary 
currents, although it's nominally 0 or 180 degrees, depends on the 
streamer loading. I saw it wander through that full 180 degree range 
when I tried various loads. So this was something we all argued over, 
and primary current feedback won, since you are sensing the very 
thing you want to control.

There is one missing step in the "QED" proof though, which is: Why do 
we want a feedback oscillator using the primary as its resonant 
circuit in the first place? Put another way, why is zero-current 
switching a good thing?

The answer is that zero-current switching allows the IGBTs to work 
much more efficiently- the switching losses are reduced by about a 
factor of 5. It may also allow them to stand overloads better. It 
seems that most IGBTs can carry more current than they can switch, 
due to dynamic latchup.

Steve Conner
http://www.scopeboy.com/