RE: Flyback Transformer Boards - Any interest ? ? ? ?

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

Original poster: "Craig Fuller" <cbfull-at-hotmail-dot-com> 

In my experience, removing the shims made the circuit work much better.
Leaving them in keeps the core cooler by preventing saturation, but (I 
think) it wastes power through stray inductance.  My transistors (2 x 
MJ11032 in a push-pull design) stayed cooler longer, and the output was 
much better.  This also reduces the resonant frequency which in turn makes 
the transistors even happier.  The only problems I found is that the lower 
frequency tends to make the core sqeal rather loud, and the core tends to 
heat up.
I'm not sure what you mean about the output not beign sinusoidal because 
this circuit seems to produce a beautiful, near perfect sinusoidal output.
There is also a wonderful application note for electronic fluorescent 
ballasts that is basically describing these very circuits (take a look at 
figures 8 and 9)

http://us.st-dot-com/stonline/books/pdf/docs/3706.pdf

Maybe this would be a good piece of reference material to help you decide 
how to design the circuit.  I personally have been trying to convert my 
self-starting, self-tuning, push-pull circuit to an analogous version using 
a half bridge.  I found another application note that does just this, while 
also reducing the complexity of the circuit by using a complementary pair.
What I mean by reduced complexity is that no center taps are needed, and 
only a single resistor is needed for startup. That note is no longer posted 
but I have a pdf of it if anyone is interested.



>From: "Tesla list" <tesla-at-pupman-dot-com>
>To: tesla-at-pupman-dot-com
>Subject: RE: Flyback Transformer Boards - Any interest ? ? ? ?
>Date: Thu, 02 Oct 2003 21:44:29 -0600
>
>Original poster: "Wall Richard Wayne" <rwall-at-ix-dot-netcom-dot-com>
>
>Steve,
>
>That's how a flyback is suppose to work.  The shim prevents magnetic
>saturation.  It's almost impossible to saturate air.  Think of the air gap
>as acting like a huge magnetic "spring" which releases a large amount of
>energy rapidly after being compressed.  There is a huge HV peak in the
>secondary.  The secondary discharge of a flyback is not a smooth sinusoid
>waveform but a sharp spike with a D/C bias.  Flybacks are not just plain
>ordinary transformer action.
>
>There is an advantage to push pull flyback mode, but it takes making a
>center tapped primary.  Only one side of the primary is active at a time.
>Frequency, dwell time and push-pull configuration can be achieved with a
>variety of inexpensive PWM driver chips.  BUT, keep the shims intact.
>
>Dan, if you design a dual transistor (push-pull) flyback driver, then those
>who want just a single transistor may use just one side.  And, those who
>want push-pull can use both sides.
>
>
> > Wow, i did not know that!  Even when i roll my own flybacks, i leave the
> > gap spacers in the core.  Hmm, maybe i should test this push-pull setup
> > again and see how it goes with a gapless core.
>
>Good idea.  Let us know how it works.
>
> >
> > Steve
>
>--- Richard Wayne Wall
>--- rwall-at-ix-dot-netcom-dot-com
>