Re: BIG solid state devices...

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

Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>

Hi Jim,

At 01:03 PM 7/13/2002 -0700, you wrote:
>> >
>> >For the IGBTs, switching time might also result in significant power
>loss..
>> >Rise and fall times at 500-1000 Amps look like a few microseconds.  The
>data
>> >sheet shows 1-2 Joule per pulse loss for 500-1000 Amps collector current.
>> >At 100 kHz switching rate, thats 100-200 kW in the switching losses...
>>
>> They key is that we are NOT switching at 100kHz.  Were switching at 120
>> BPS.  A spark gap is just a switch we turn on every 8.333mS and back off
>
>excellent point...
>But, since the bias on the devices will be reversing, does the device know
>that it is staying on?  Sure, the gate drive stays hot, but, I don't know
>enough about how the internals of these things work to be sure that they
>don't think they should turn off when the collector bias goes away.  After
>all, they are sort of like a FET driving a BJT.. when the collector bias
>goes away (voltage reversal), then the base drive dissappears.. Of course,
>the charge carriers probably stay in the area (after all, they DO turn off
>slower than they turn on), so when the base drive comes back on, the
>junction's already loaded up. Hmmm..

As far as I could tell in my tests.  There was no problem at all.  The
IGBTs and the fancy reverse diodes are real special, operate very fast, and
the manufacturers play all kinds of tricks to "fix" the problems you
mention.  They seem to work great.  I do wonder if that is true for the
newer really large "rev A" IGBTs that might not have such details corrected
yet.  I actually have to slow them down some with gate resistors to keep
them from switching "too" fast and damaging themselves.

I do think they are best for coils operating less than 100kHz where the
speeds of the IGBTs is far faster than needed.  If a coil runs at say
500kHz, then things may get messy.  If you look at the voltage waveforms in
some of the 500kHz tests in the link below, you can see a little crossover
glitching.  I remember I fixed that, but I seem to have forgotten how (I
probably read the data sheet or something ;-))

in:

http://hot-streamer-dot-com/temp/2001-04/SScap2.jpg

I get up to 150 amps at 500kHz.  The tiny voltage glitches get sort of lost
when your have say 1000 volts too ;-))  What is important is that the IGBTs
hold together and run cold ;-))

>
>> you may want to see the scope pics at:
>>
>> http://hot-streamer-dot-com/temp/2001-04/SScap1.jpg
>>
>> This is just one little 1200V 150A IGBT I was running into a dead short.
>> Even though I was going through all fiber optic triggering and such, the
>> times are very fast and the voltages and currents are very high too.  They
>> really are wonderfully well behaved for this stuff.
>>
>
>Then, there is the master oscillator - Class E power amplifier approach..
>Someone at work was saying that we should take a look at that... The TC is
>the tank.
>

It's not quite class E as Sokal would have meant since the switch operates
at a far different frequency than the carrier...  Ref US Patent 3919656.
"Ours" is far better and more efficient :-)))  That's why Tesla patented it
73 years before Sokal's class E (but Tesla didn't have IGBTs and Sokal had
FETs, so we use class E... (also true CW operation))  Maybe a chance to
patent something "new" in all this aside form the fact we publish every
detail all over the planet ;-))  But really, all the hard work is being
done by the folks that make these wonderful IGBTs!!! 

Cheers,

	Terry