Re: OLTC update

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

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

Hi Malcolm,

At 08:28 AM 8/29/2002 +1200, you wrote:
>Hi Terry,
.............
>
>Each IGBT is discharging an individual cap isn't it? If that's the 
>case, then the only ways of alleviating the stress on the devices is 
>to start paralleling them with others or reduce the capacitance 
>associated with each.

Yes, each IGBT has it's own cap to control current sharing.  I could use
twice as many IGBT/cap sets with caps of 1/2 the value and cut the current
each IGBT sees in half.  However, I designed a "tester" today that can run
a single IGBT up to destruction that simulates my real coil's current
waveforms.  So I can find exactly how much current they can take with my
coil's pulse pattern.  I know they can take "more" than the data sheets
say.  I am just not sure if it is 10% more or 5000% more...  IR just gives
guaranteed specs under just about any condition.  If one wants to run them
higher in some particular application like mine, "I" have to do the work to
find the specs in that case...  The fewer IGBT/caps needed make the thing
cheaper.  I guess an "E"MMIGBT array :o))  In our non-"life or death" hobby
Tesla coil application, we are allowed to run them at "destruct level - 1%"
:o))  Some folks I have talked with say that since the die are only
slightly above room temperature, the duty cycle is low and slow, the life
time can be "only" hours, the application is not-"critical" and the
oscillations divide the stress between the IGBT and the anti-parallel
diodes...  all add up to being able to drive the IGBTs at "wildly high"
peak currents...  Nobody knows how high.  I plan on finding out :-)))  They
do warn to be careful of Ccg shoot through but that is mostly a high
collector dV/dT problem not a giant current problem.  I have some really
fancy TVSs if that turns out to be a problem but I can still slow the
switching down too which is probably easier if needed.  I don't think it is...

>
>     Re the coupling, Eddie Burwell has it. Could you post a url for 
>an oscillogram of the primary ringing alone please? I don't have time 
>to wade through all the urls you post unfortunately (my problem, not 
>yours ;)

Cool!

http://home.hiwaay-dot-net/~eburwell/

I have not seen this before!  Looks like Eddie is using an SCR to do
something similar to my coil.  I will check this out!

I am planning on changing the ring up pattern drastically tonight.  So
don't worry too much about "todays" patterns since they may be all
different "tomorrow" :o))

Cheers,

	Terry
 

>
>Regards,
>Malcolm
>