Sue's Plasmasonic, WAS: Terry's DRSSTC

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "Sue Gaeta" <sgsparky@xxxxxxxxxxx>

I ended up sticking a TVS across the power supply caps. The extra fuse 
seems to work OK in this case since it's not in the way of the primary (the 
heavy RF circulating currents should be confined to the MOSFETs and 
primary). This scheme probably wouldn't work in a DRSSTC though. I have a 
bunch of little caps instead of one big one similar to your arrangement. 
Each of the 4 HV caps are 820uF at 250 volts. I am not going to go above 
140 volts on the line so these are good. When the coil is running, there is 
no hint of power supply hum in the spark, but when I tune a receiver in to 
237KHZ I can hear some hum. No problem, it's a Plasmasonic, not a pirate 
radio transmitter :-)

What the heck, I'll put up a pic.
<http://community.webshots.com/photo/17872929/222513836QQTBSE>http://community.webshots.com/photo/17872929/222513836QQTBSE
It all fit so nicely in that chassis! I relocated the L.V. filter caps so 
that I could get some airflow to the driver chips that were hiding behind 
the caps. I also stuck an extra 4,700uF cap in there because the fans draw 
an amp at 12 volts.
I didn't want to use "a ceiling fan" over the heatsink because having a 
solid cover over the top of the board will keep the potato chips, screws, 
and whatever else that could be "death from above" outta there!

That thing sounds really sweet! Thanks to Dan for making those boards 
available to us!
My first light song was "Too Much Time On My Hands". My family thought it 
was appropriate! I was only drawing 8 amps off the line at 100 volts in, 
and those big heat sinks got warm after a 3 minute run. I guess the 
heatsinks work :-)
I know there are people that like to look at boring waveforms (same darned 
thing 236,000 times a second gets boring after a while), so I uploaded 
those too. The first two is a side by side comparison of two different 9A 
driver ICs (no difference).
<http://community.webshots.com/photo/17872929/222523787zDEZNl>http://community.webshots.com/photo/17872929/222523787zDEZNl
Things get a little more interestng when I compare different output FETs, 
and gate drive scenarios.
<http://community.webshots.com/photo/17872929/222524761KHBcnH>http://community.webshots.com/photo/17872929/222524761KHBcnH
I finally decided to do away with the schottkeys since it's a little easier 
on the driver, and I have all that dead time anyway. I will never run this 
thing single ended mode since it would require some butchering, and I will 
use my feedback SSTC circuit to do the big streamer thing.

Richardson seems to have 942Cs listed cheaper than the last place I looked. 
Thanks.

Happy Thanksgiving,
Sue

Tesla list <tesla@xxxxxxxxxx> wrote:
Original poster: Terry Fritz

Hi Sue,

At 01:24 PM 11/24/2004, you wrote:
>Me Too!
>
>I am glad to see that there is someone else is still working on these things.

I guess there are "two" of us now :o)))

>Someone commented that there has been a lul in the solid state work, but I
>have been quietly working here in my lab too. I finally put the finishing
>touches on my Plamasonic I PCB that I started two years ago, and also
>finished stuffing the two resonant driver boards that I got from Dan.

I takes time for the rest of us that work like snails to catch up...

>I found two nice beefy 24 volt center tapped transformers on Monday that
>will work well to power both the driver circuit and two muffin fans!

I just got a 2 amp one off DigiKey for $20! They don't give them away
anymore unless you try hard. They have to make room for lots of safety
labels these days... Since a few have flamed up in the past, I guess I
can't complain for a little more safety ;-))

>
>I drilled out some holes in the chassis for the Plasmasonic, but the drill
>walked a little in spite of my making divots first. The problem is that I
>have a hand drill with no press, so I will have to move a few holes today 
:-p.

Making small holes first and then working up by steps works pretty good. A
variable speed one (even cheap from Harbor Freight) helps a ton too...

>
>I built the high voltage rectifier PCB, but I changed my mind and want two
>fuses instead of just one. I want one ahead of the bridge rectifier and
>bulk caps in case something shorts there, but I also imagined another
>scenario. What if a MOSFET shorts? - All the energy that is stored in the
>big bul k caps will discharge across that short and KABOOOOOOOOM!!! (Dave
>Sharpe says it so much better than me!). So now I want another fuse
>between the bulk caps and the MOSFETs.

Problem is that circuit also carries the full primary current unless you
can bypass most if it out with big poly caps local to the IGBTs. Fuses may
not do well in the primary path but maybe they could do OK if they were
just the right kind. Low RMS current with high peak current. But they may
not help incase of an IGBT failure still... Just got a ton of parts
sitting here, so I will ponder it now... At work, MH circuit breakers do
well here, but they are not cheap (except on the used market ;-))...

>Your line filter sounds like it's really top notch, lots of good
>protection and all too! I don't have anything between my line cord and HV
>supply except a fuse :-o. Maybe I need to re-think things a little better.
>At least I have a comme rcial line filter on my built from scratch SSTC.
>That's a great idea you have there - using the same psu/filter assembly to
>power each coil. That would save on parts and space consumption.

I would at least drop in some MOVs. They could sink a cap over voltage
event and keep it off the line. My variac box also has MOVs and line
filters ;-)) But I like over kill, especially with a house full of
computers...

>
>I have also been playing with the driver a bit and swapping out different
>MOSFETs and IGBTs. I had these discontinued 47 amp devices from ST that I
>thought would be perfect for my coil, but the gate capacitance is
>tremendously high, and the waveform was awful, and the driver supply
>current was above an amp!
>I swapped to the IXYS 48 amp ones and they were much better. I dropped the
>gate resistors to 4 ohm and added schottkys, and the results were better
>still. Beautiful wav eforms, and the supply current was only 0.6A. People
>who complain about their driver chips getting hot should experiment like
>this. Even different brands of devices with close to the same ratings can
>be considerably different.

Lots has changed in the last few years and "new" devices are far far
better! Unfortunately they demand higher prices too...

>
>I think you are right on the Poly filter caps at the IGBT bridge. I think
>Dan himself had one on his DRSSTC that he demo'ed at the Teslathon, but it
>isn't in the schematic.

;-)))))

>
>I should have the Plasmasonic one up and running by tonight.
>My DRSSTC progect is going to start out as a SSTC for now until I can find
>those capacitors at a reasonable price, or in someones garbage :-).

Mine are left over from the OLTC, but no extras here... Might have to call
Rell.com anyway...

>I still have a ways to go yet on t hat project. I still have to wind the
>1:2 GDTs, and wind the much lower Fres secondary. No more cheating and
>using my medium sized SGTC secondary for all my solid state excursions any
>more!

I think it would be cool if the DRSSTC drive could be sort of a "drop in"
replacement and work on existing coils... More to think on there...

Cheers,

Terry


>
>Good luck on yours Terry!
>Sue