Story of my SSTC / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Story of my SSTC

Move Thread

LAN_403

Kizmo

Wed Oct 26 2011, 08:47PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Mark I - The beginning (Started 09.02.2008)

Almost four years ago i got bitten by Solid State coil bug and after some googling i decided that Steve Wards MiniSSTC was the best way to go.

I bought some magnet wire and components from local electronics dealer including terribly expensive IRFP460s (4.5eur each). Also 2.2kVA variac and Tektronix 2235 oscilloscope were found for reasonable price.

I carefully copied Steve Wards circuit on cheap stripboard and then wound about 950 turn 110mm diameter and 300mm long secondary with 0.315mm wire. Few layers of polyurethane was applied. My first bridge was full bridge because i was looking for some power from the beginning.

I got almost all components done but then suddenly lost interest when i had troubles with finding correct phasing for GDT and i didnt get that damn thing to oscillate. At this point the project was put back on the shelf.

Mark II - Hey it works! (Started 10.2.2009)

Time passed by and i made some smaller projects like flyback drivers and spark gap coils. Then i found this project again. It didnt take much to get it going with all knowledge i gained from 4hv forum and irc friends

Coil was operated from heavily filtered DC. 2.2kVA Variac was the limiting factor in this kind of operation. As you know CW sstc:s are real power hogs :) I popped couple H-bridges with bad gate signals and overheating. Finally all mosfets i bought for Mark 1 were dead and this thing was put back on the shelf.

Mark III - Discovery of the high ripple input voltage (Started 1.11.2009)

During dark winter days i tought i give this thing another go this time with half wave rectified DC. Starter oscillator was used in form of 555 oscillator connected through 47k resistor and 100n capacitor. It wasnt very reliable under load but i still managed to get fairly decent output

Then i stepped it up with Half wave voltage doubler and really tortured those poor mosfets with DC voltages over 450VDC.. body count went up but so did sparks

Two years of playing with SSTC. Most of these were salvaged mosfets (IRFP460 and Ixys 24N50)

This coil seemed to love large toploads so i just piled lot of misc junk on top of the coil. Then i ran out of power capacity in my apartment (2.2kVA Variac + 16Amp outlet) and once again this thing was put back on the shelf.

Mark IV - Far too much power (Started 22.10.2011)

Almost 2 years passed and things changed quite a abit. I moved to another city and when collecting stuff from my old apartment i once again found the original Mark I driver card and secondary coil. This time i had plenty of power available, lot of good Fairchild 44N50 Mosfets and 3 phase variac.

Bridge was put together on huge heatsink with copper heatspreaders. Idea was that electric insulation pad was between heatsink and copper piece and mosfets were sitting on bare copper. This is only picture where you can see the almost 4 year old driver card i made back in 2008. Clas Ohlson provided nice 300mm diameter stainless steel ball for topload.

This time i had commercial 0.1V/A wide band current transformer and Rigol DSO. What you see here is inverter output current during balls-to-the-walls run when i maxed the variac. The coil was still operated with Half wave DC but this time i used half waves from three different phases in order to create 15ms long lump of power at 50Hz repetation rate.

60.2A RMS

Green is conventional 50Hz half wave and Blue is close to what i had. I used combination of transformers and 3 phase variac to create that.

And i made small invention too. Running the coul with half wave DC on bridge is impossible if you have self resonating driver and no interrupter or starting oscillator. I dont like interrupters since they make loud noise and CW sparks have more current. And starter oscillator was proven to be unreliable and caused lot of mosfet failures when the coul was loaded with ground arcs. But the solution is very simple and effective!

Whole idea is that bridge rectifier, T1, C1 provide stable supply of filtered DC that keep the coil alive all the time. D1 prevents your high ripple input from variac/mains + D2 from being filtered by C1. I did just that and it works perfectly. I dont have starter oscillator or interrupter and no matter what waveshape i use it just keeps going. Initial starting of the coul is done by poking the antenna with finger.

And result? Well - lot of dead mosfets. I had some feedback issues and bad habit of cranking variac way too high. But results were pretty spectacular. I had to move the stainless steel topload further away from primary because it got so hot it melted the plastic screw.

Average power input was 10-12kW

Stupidly high power with insanely high coupling results insane flashovers

Today the H-bridge popped 3rd time and i think thats it. This is the end of the almost 4 year journey of this driver and secondary. Both of them served me well and last few days have been very interesting. Practical limit of this topology is pretty much where Mark 3 ended.

Despite all flashovers and accidents the secondary still works and i think will retire it and controller card in form of small table top demo coil that can sit on my shelf as reminder for ever.

Future plans? DR... :)

Marko

Wed Oct 26 2011, 09:29PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi sir,

+1 for all the epic SSTC's. Funny, I never thought of using a transformer contraption like you did for keeping the coil running, I might steal this idea for my mega SSTC if I end up not being bothered redesigning my driver...

In the last video, what kind of input waveform are you using? The spark looks pretty CW to me. And are you sure it's 12kW?

This is what I dared to run up to, with a full bridge of HGTG30N60A4D igbt's... (3 phase, fullwave rectified so almost CW)

A great way of preventing your mosfets/igbt's from dying is to use a current limiter - a CT on the bridge output, with a comparator and a NE555 monostable to disable your gate drivers will keep you safe from practically any overcurrent accident... You can see the circuit tripping in my video as I crank up the variac, limiter was set to 60A (peak).

I was a bit puzzled to see the input current reach 16A per phase, at 300V DC input, and I thought something must be wrong with my measurements there despite I was using moving iron meters only.

Regarding feedback, you are very right about the Steve ward's 74HC14 scheme sucking. A while ago I was naive and used a feedback CT directly into inputs of UCC's (with a few diode clamps) which worked surprisingly well for CW - in this coil I added a TL494 as a starting oscillator but it still didn't do the job of starting the coil with full or half wave rectified input.

And a few times I tried using other gate drivers than UCC's this way, even buffered with HC14's always turned out to be a disaster and exhibited extremely bizarre behavior... I wouldn't recommend anything like this anymore.

If you want to drive a coil reliably from a weak signal from a CT or antenna, you need a high gain amplifier with properly biased input... preferably some sort of a fast comparator. This is what Steve ward's first coils seemed to use and they were apparently quite reliable.

The comparator is so sensitive that it may start on it's own from noise or require only a very weak starting oscillator signal.

I'm actually designing a ircuit of this kind based on LM119 and IXDD414's, with OCD ofcourse.

I'm not exactly sure of the fate of my coil though, whether it will end as QCW or I'll just try to ramp it up as CW even more.

This topology is certainly nothing but pointless inverter torture as huge reactive power needs to be processed - a high impedance series resonant primary, or perhaps LCLR topology would be great if it could be made to work, also you wouldn't need to use as extremely high coupling that invokes flashover destruction.

Marko

Kizmo

Wed Oct 26 2011, 09:34PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Marko wrote ...

In the last video, what kind of input waveform are you using? The spark looks pretty CW to me. And are you sure it's 12kW?

It was little bit over 12kW actually and im pretty sure its correct since i measured current from DC rail with TrueRMS clamp and calculated rms voltage with rigol dso

And that last video, im still using my half wave rectified 3 phase contraption so the input is nowhere near flat line dc :)

Marko

Wed Oct 26 2011, 09:56PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Dung, is it possible then that my coil is already using 14kW then? Still finding it hard to believe. Need to place the ammeter onto DC link instead.

What kind of DR upgrade do you intend? A CW or a pulsed system?

Marko

Kizmo

Wed Oct 26 2011, 09:58PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

Completely different coil, almost 2 meter tall CM300 drsstc :)

Pieznice29

Thu Oct 27 2011, 03:59PM

Registered Member #3918 Joined: Sun May 29 2011, 08:42PM
Location:
Posts: 34

Damn, that's impressive. Great work, the results were amazing. I love those flame like discharges into the air.
Shame about all those MOSFETs that went pop in the process...

Mads Barnkob

Thu Oct 27 2011, 05:08PM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

I like the long story of its life and brutally beaten to death in the end, might as well go out with a good show.

I feel like digging my old SSTC out from the shelf and rebuild it, I think its blown if I remember correct... It goes back to May 2009, so its just a young one ;)

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob