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

RE: SSTC design procedure



Original poster: "Steve Conner" <steve.conner-at-optosci-dot-com> 

 >I take that means one being well-tuned to the other, or are you
 >hinting at something else? What happens to that fine tuning when the
 >sparks start flying?

Yes, the secondary and primary of DR/ISSTCs ought to be tuned to (about) the
same frequency. Then, the L/C ratio and primary-secondary coupling are
chosen to get the desired loading on your inverter. (ie the highest current
your devices can take without incinerating)

We do know that L/C and k have optimum values for power transfer, but there
is no formula for finding them yet, we just cut and try. (If some math
genius would like to derive one, we would be delighted.)

When the sparks start flying it damps the resonance, and lowers the
frequency, of the secondary. With a feedback driver, the inverter will track
this frequency change, so it appears that the _primary_ is going out of tune
instead. Tighter coupling, and detuning the primary slightly low, seems to
help maintaining power transfer under these conditions.

But again I think things are kind of at the hand-waving stage, and will stay
there until someone creates a good equation or PSpice model for secondary
impedance including dynamic streamer loading. Right now it seems everyone is
having too much fun making huge sparks and blowing up IGBTs, and can't be
fussed with boring stuff like modelling :)))



  > Other SSTCs that don't require a tuned primary circuit can be built
  > more arbitrarily.

This is correct. If you are using an untuned primary circuit, and a feedback
driver, you can pick a secondary and toroid at random, wrap 10 turns of wire
round the bottom (or 20 if you're running off 240V) hook it up to your
inverter, and light it up.

In untuned primary SSTCs, the primary turns and coupling determine inverter
loading and power transfer. Less turns gives more power, but more
magnetizing current, which is bad as it increases device losses. Tighter
coupling gives more power but does not increase the magnetizing current. So
it's best to use the tightest coupling possible, and reduce the turns until
your inverter is maxed out. (You can tell how heavily it's loaded by
measuring the peak primary current with a ferrite cored current transformer,
and comparing to the MOSFET/IGBT datasheet absolute max rating.)

Note: DR/ISSTCs have no magnetizing current because of the tuned primary, so
you can use as few primary turns as you want.

It is also desirable to go for a resonant frequency less than 250kHz because
it makes the inverter easier to build.

SSTCs with untuned primaries normally need a breakout point as they can't
generate enough power to break out a smooth toroid, the MOSFETs tend to blow
first :(

Steve C.