Re: Pushing the IGBT Envelop

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

Original poster: "Greg Leyh" <lod-at-pacbell-dot-net> 


>
>Original poster: "Antonio Carlos M. de Queiroz" <acmdq-at-uol-dot-com.br>
>[snip]
>The lossless design ideally recovers all the energy, at least if some
>of it is removed from the output capacitance at the instant when the
>maximum voltage is reached. It would also recover most of the energy
>left unused by a load connected to the output when the output voltage
>is close to the maximum. But the driver would have to operate with
>reverse current for this, and be active for the full output "beat".

That's a reasonable scenario, as non-quenching sparkgaps almost exclusively 
operate in that
fashion.  Solid-state switches can, as well.

>What is ALF?

It's a theoretical coil that I would like to see built.  The physical scale 
of the coil is set at the
maximum size that still allows shot-to-shot buildup of the arc while 
retaining a relatively short
operating duty factor.  Both of these are necessary for a good arc-length 
to power ratio.
I'm estimating that a break rate of at least 250PPS is needed for good 
shot-to-shot buildup,
yielding a shot period of 4mS.  Here I'm also defining a short duty factor 
as <10%, in order to
maintain a reasonable power multiplication.  This allows about 400uS for 
the beat envelope
to couple the power to the secondary.
The largest coil design I've come up with that can couple the power in 
400uS has a secondary
that's approximately 96ft high and 22ft in diameter.  This coil has a Fo of 
about 7500Hz.
Another factor of two in scale can be had by operating two identical coils, 
in antiphase.

I would imagine the finished twin coil system appearing something like this:
http://www.lod-dot-org/720L5M.jpg
Here's some preliminary specs, although they're terribly dated at this point:
http://www.lod-dot-org/alftech.html