RE: 48kW DRSSTC

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "Hooper, Christopher AZ" <christopher.az.hooper@xxxxxxxxx>

Hi Steve,

What I mean by slice card is; all components are placed on a card for
that channel of the network. It does not have to be a network card, but
a component of a torpedo (round card - worked for Rockwell a long time
ago) or a front end of a receiver or the backend of a power amp. This
card is plugged into a cage/rack. Since the failures on my DRSSTC's have
been the IGBT's due to spark making contact with the primary, or the
driver card losing its feedback, or the OCD (Over current detention
circuit not set correctly) the poor IGBT's fail. It is a pain in the
middle of the demo to have to pull the DRSSTC apart, pull the dead IGBT
off the board, white goop a new IGBT and reconstruct the dead DRSSTC and
start the demo again.  You are correct that this is more costly due PCB
fabrication and card cages. Money is not a concern for me. I just want
to bring the coiling of DRSSTC to a more run time and less bench time
repairing blow bricks and such. So the IGBT's will have there own card
(x 4) on a slab of cooper with traces for the inputs and outputs routed
to edges, the edge will be plugged into a heavy Viking connector and a
stabilizing bar will be attached to the top of the card. The heat sink
will also be part of the board. There will be four cards, each with its
own IGBT with heat sink and plugged into a very simple base. I will also
build two extra cards with IBGT (CM600) with heat sinks sitting in a
nice ESD package. A little feedback circuit will be designed to drive a
LED, this LED will be visible from the side of the coil. I push my
DRSSTC to melt down as always try to beat my mentor (Dr. Ward) on his
spark length. We started this on tube coils @
http://users.cableaz.com/~chooper/images/sparks_1.gif  200 years ago,
went to SSTC's and now DRSSTC's. Why make a new design, just copy and
make beter....he he.

Anyway, if I am pumping 25 foot sparks and the coil dies and the fuse
blows, I bet it will be one or more of the CM600's. The little feedback
circuit will indicate which IGBT (hope only one) is dead, I will walk
over, pull the card, pop in a new card, power up and keep pumping 25
foot sparks. I hate pulling things apart with 20 people around me, so
this will make DRSSTC more up time, easy to fix. When things are quite,
I can take the dead card, lock the garage, open a beer and peel the
melted CM600 off the card. I hope only two extra cards will be need for
Teslathon, he he. I forget not to make things more clear, ugggggg.

Thanks for the SCR input for the controller as this is the first phase
of our project. The 48KWatt coil is still in the dream stage, as got to
get this 5KWatt DRSSTC done for W.W.T in Jan (I Hope, but have two hot
ones running now - 800 watt and 2kwatt ), then the 48Kwatt will start in
March time frame. I am really, really excited, I have already talked to
Dave (D&M) to build the 3 foot secondary, as his winding machine can
take up to a 5 foot secondary if I want.

Best rgs,
christopher

-----Original Message-----
From: Tesla list [mailto:tesla@xxxxxxxxxx]
Sent: Thursday, December 01, 2005 1:38 PM
To: tesla@xxxxxxxxxx
Subject: RE: 48kW DRSSTC

Original poster: Steve Conner <steve@xxxxxxxxxxxx>

 > This
 > is what I want to do with the 48Kwatt coil, all on
 > slice cards, one
 > fails, pull, drop in new and keep sparking!

So what exactly would you be thinking of putting on
slice cards? Would you build, say, ten 5kW converters
instead of a single 50kW unit, and combine the power
from them? Do you intend to do that with the
rectifier, or the HF inverter, or both? It sure would
be interesting to see how that turns out!

I thought about doing it, but I realised I would end
up spending so much on PCB fabrication and card cages.
It seems a lot more economical to just use one set of
the biggest IGBTs I can get my hands on (which would
be CM600s I guess) on a huge heatsink, with an array
of honkin' big electrolytic caps strapped to them.

If I really get power crazed I can always build
several of these large bridges and combine the power
output from them into a single Tesla resonator, in
some way that I've yet to figure out. But even four
CM600s, or two of the CM300 half bridges for that
matter, can probably produce bigger sparks than I have
room to fire off.

You might be interested in the DRSSTC driver I
developed. I'll be honest, I have a vested interest in
getting as many people to use it as I can. The more
people use it, the more bugs I'll find ;-)

It was done with museum type coils in mind, so it has
all sorts of protection features that would come in
real handy on a big coil. Finn Hammer and I have built
7 DRSSTCs with this driver circuit, and we have had no
IGBT failures in service yet.

I'm currently working on some add-on gate drive
amplifiers for it that will allow you to drive almost
any number of IGBT bricks. At the moment I'm wondering
whether to put desaturation and gate drive failure
protection on these add-on boards. It's theoretically
nice, but it adds enough complexity and extra places
for EMI to get in, that it might actually end up
reducing reliability.

There is also a plug-in "solid state variac" board
that will drive SCRs for voltage control of a
rectifier or doubler, but I only have a prototype of
that built on perfboard.

The schematics etc. are on my DRSSTC page-

http://www.scopeboy.com/tesla/drsstc/

Steve Conner