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

Re: OLTC update - primary IGBT loss



Original poster: "Paul Nicholson by way of Terry Fritz <twftesla-at-qwest-dot-net>" <paul-at-abelian.demon.co.uk>

Terry wrote:
> Wonder what happens if one refires 180 degree off from the present
> ring...  That should send some shock waves up the secondary ! :o))

Oh, what madness is this!

I modeled the situation where: just as the primary energy is
momentarily zero at the peak of the first beat, we re-stock the
primary cap with another bangs worth of energy (eg by switching
to a second cap bank - one that you charged up while the 1st bank
was going through its first half-beat).

The normal ringdown, without quenching, is 

 http://www.abelian.demon.co.uk/tssp/tmp/tfoltc-c1.gif

and the double-bang event is in

 http://www.abelian.demon.co.uk/tssp/tmp/tfoltc-c2.gif

The second cap is switched in at 50uS.  Following this, instead
of the secondary energy ringing down, it does another cycle or so at a
slightly higher Vtop.  Note that the Ipri and Vpri don't exceed 
their original values, so no extra device stress.  The follow up bang
roughly doubles the number of cycles that the secondary could be
extending streamers over.

No doubt some fun could be had modeling various timings, but you can
explore that better with a lumped circuit sim than I can with a
distributed model.

> I need to try wrapping another coil to see if that affects
> anything like Antonio suggested.

Well if you do, ping the coil before and after so we can see how much
it changes.

> I did not retune the coil for the added probe C since I didn't
> want to fiddle too much

Yeah, that's fine.  I don't care what the tuning is, so long as it
stays put.  It's a major ball ache to recompute everything if something
is shifted.

> I think I have much higher coupling!  ~0.25

Indeed.

> This coil likes to arc to the primary at about 8% up the secondary
> coil.

That would coincide with the location of highest radial field strength
shown in the gradient animation.

> Is a pin point OK?

Depends what we want to know.  My particular interest is in witnessing
the expected 26kV/cm surface gradient in action, and to observe the
extent of the topvolts clamping action.  For that we need a predictable
breakout surface, so a point is out.  I can manage with a small ball,
or a rod with a smooth hemispherical tip.  The rod or ball must be
vertically above the axis of the coil/toroid so that the whole system
has cylindrical symmetry.  Single shot only for now.

But then if we simply want to know whether streamers evolve over 
part of a cycle, a couple of cycles, or several consecutive bangs,
then any old breakout point would do I suppose.  Don't need a topvolts
measurement for that, just the current to the breakout point, along
with Ibase.  We might find that we can go a long way just on that
basis, eg we can integrate the breakout current to see how much charge
is upheld.  We can tabulate upheld charge with observed streamer length.
We can also look for signs of streamer evolution preferentially on
one polarity.

Either way, tons of stuff to go at.
--
Paul Nicholson
--