SSTC (Not) Working Correctly

[Mads Barnkob]

So basically a capacitor in the primary circuit doesn't effects the primary resonant frequency,and the important thing is to drive it at the secondary's resonant frequency?

Not if there is a huge mismatch between inverter impedance and capacitor reactance, then it will just behave as a DC blocking capacitor, which is only there to protect the MOSFETs from a saturated primary coil (preventing explosions).

[Laci]

I calculated my designs properties.At 150kHz,330VDC rectified mains,the 3×0,22uF caps have 23 ohm reactance.So if I remove one from the three,the current and the arcs size should increase?

[Mads Barnkob]

I calculated my designs properties.At 150kHz,330VDC rectified mains,the 3×0,22uF caps have 23 ohm reactance.So if I remove one from the three,the current and the arcs size should increase?

In theory yes.

But you are only going to be blowing up MOSFETs when you try to run a MOSFET inverter for a SSTC as a DRSSTC.

As the primary current goes up, the losses in a MOSFET versus a IGBT is very different!

A MOSFET has a fixed on-resistance and this is the root cause of heating problems with high currents, the fixed voltage drop of a IGBT makes it have a lot less conduction losses than the MOSFET, at high peak currents.

[Laci]

The oneTesla interrupter is done on a breadboard.Here are two short videos about it running.I was changing the on time,to see if the audio quality is better at lower on times or not.The popping sounds are existing at me too,and also at high on times the higher tones gets distorted,like in the Pirates of Caribbean.
*Running at a halfbridge of 2xIRFP460.

https://mega.nz/#!EWBkgADC!LDEw2XMfGtdSZRNl6JwAd5hNmV-WHBNajizALh813NI

https://mega.nz/#!MbZRBZQI!Vikm4im2t40Qc0l12D8tJiDiLgcfKILILvruWl-_BJI

[Laci]

Since I got 10 IRFP460,why not rebuild a SSTC driver?!I made a full bridge of them,more less accordingly to Mads recommendation,with thick wire and without perfboard.Also made a full bridge rectifier with 5 caps in parallel for a total of 1200uF.I put them into a nice enclosure,with full/half wave rectification(120/60Hz) and a switch,which bypasses the 330ohm resistor responsible for slowly charing the caps.
Actually I am sad,because 5 transistors and 4 TVS diodes died in a row.Two of the transistors died violently and three of them silently,as well the diodes.The 440V TVS diodes always died when I played midi with high on time from the interrupter.The first transistor died because of a spark to ground (which will probably be a new post),then the next two pair simply exploded at low bps,then I stupidly after replacing the transistors touched the "snubber caps" which of course got connected to the gate of one mosfet.Silent death.Then replaced the transistor,but not the protecting zener diodes which killed the new mosfet...It's a really bad idea to work with these kinda things when you are tired...

[Laci]

While being busy with school,I had a bit of time to continue this "never ending" project.Made a new IRFP460 bridge again on perfboard,but with 2 sides(halfbridges)on a heatsink to make the replacement easier.
Another important thing is that I got a new grounding.Instead of the mains ground now I have around a 50cm "U" form steel sinked in the ground and connected to the bottom of the secondary through around a 7m long,really thick 3in1 cable with the wires paralleled for lower resistance.
The driver circuit is the UD1.3 with antenna feedback,without current transformers.The interrupter is the oneTesla with pretty low on times,probably the default settings,for DRSSTC.The whole driver is poorly insulated from the tesla coil,not sure if that's a problem or not.
The main issue is the current drawn.Using my newer primary with the thick wire(6 turns)the rms current is not enough to blow a 1A fuse and the arcs are not really good,maximum 10cm(1/5 of the secondary).The midi sound is already too loud.Don't get me wrong,not for me,but for my neighbours.
I assume that the low on time of the interrupter prevents the arcs to grow,because with such a thick and just 6 turns primary the current flowing into it should be way to higher even at SSTC operation.How else can I increase the arcs lenght?Any tips would be appreciated.
Also the main motivation is the Kaizer SSTC II and oneKone's SSTC at the arcs to secondary size,which is around 1:1 in both cases,right?

[Mads Barnkob]

Are you using a DRSSTC UD1.3 with antenna feedback to drive a SSTC? I never heard of that before and maybe you have some timing issues with coil being driven off resonance or the very low on-times for a DRSSTC is too little for a SSTC.

In music mode you should expect shorter arcs, than what you can get when pushing it in interrupted mode, as you have more charge on the topload.

It seems like you have metal enclosures on everything, just get them grounded to the same ground rod as your secondary coil and you should avoid too much interference. But just a metal enclosure in itself will help a lot.

[Laci]

Right,the signal is received from an antenna.Not sure if it makes any difference,but it works this way too.With current transformer feedback the oscillation is not starting on it's own so I tried this.Also the first set of transistors has blown up.Some weird phenomena were happening with the coil.The arcs were not so big so I uploaded the code to the arduino with the lookup table scale 10 for longer on time,but without longer arcs,just higher current drawn.The arcs stopped when I was pulling an arc,moving the antenna-no effects.
Just as an experiment I removed the topload and added a anorher smaller coil to the top of the secondary.This doubled the size of the arcs and the current and sadly killed the transistor at the second midi song with the maximum on time(~350uS).3 transistors died and the zener diodes at one side.I assume that the high magnetic field got into the GDT and without grounding it easily got to the transistors.

Since the inverter is screwed again but I have the IGBTs,I wonder if this inverter design is good for a DRSSTC or not?
By the way thanks for the reply again,Mads!

[oneKone]

you've mentioned that with CT feedback the oscillation will not start, did you try changing the phasing of the CT? also is the driver self made or bought? the antenna feedback could be causing some feedback errors.

another thing (sorry i have not re-read your thread) are you using a tank/resonance capacitor? if so how have you tuned your primary to the secondary? 

[Laci]

The driver is self made.I think that the phasings were shifted,but I was not paying much attention to this,because I thought that using antenna is the same thing.Not sure if current feedback works with SSTC.

I was using capacitor with the primary in the past,but that was unstable,simpy due to the antenna feedback.Some mosfets died that way too....

[Laci]

After two days running at 180V,two IGBT died when I connected the inverter to rectified mains.Probably because I just hooked up a capacitor to make it double resonant,or because the GDT flashed over(I noticed this yesterday).Anyways as last attempt I replaced the failed halfbridge with a capacitor voltage divider and it is working now for 3 days!Well,it is not working now,because the driver has an issue;a smoothing tantalum capacitor shorted out and I have to replace it,but also upgrade a bit on the design and use a transformer instead of the ATX PSU.
I tried several things with this inverter,so far so good,it works like a charm expect the flashovers from secondary to primary even after decreasing the turns on the conical primary.
Yesterday I was looking for those fake QCW interrupters to create long and straight sparks and ended up with a combined Kaizer VTTC staccato controller and loneocean's RSSTC3 interrupter.I made this circuit on a breadboard,it was not working as it was supposed to,but probably due to bad connections on the breadboard.While testing it produced 40cm straight arcs to ground,which I think is really good.Sadly at this point the capacitor blew up and I couldn't make a video of it.

I have a video with 8 primary turns and with the complete conical primary.With that amount of turns doesn't exist flashovers,but with 4-5 turns the flashovers were significant.
I don't know why is the inverted working better with 2 transistors instead of 4!?In comparison with the fullbridge now it has +-325V/2 so like +-162V across the primary instead of the +-325V.Double the voltage,double the size of the arcs,if the inverter is working correctly,right?

https://mega.nz/#!ZeJVnKoR!FvYfJYxrESm7BmyOdQYU9mSWFapb9-rF1G6Ag1iQua4

[Laci]

Here comes some more updates!After a month the coil is still going strong,it has no problems and produces decent sparks.I suppose the full bridge blown up by an arc-over from the secondary to the primary,with the halfbridge the arcs flow through the capacitors instead of the transistors,blowin' them up.
A couple of pictures taken by my friend:

[Mads Barnkob]

After two days running at 180V,two IGBT died when I connected the inverter to rectified mains.Probably because I just hooked up a capacitor to make it double resonant,or because the GDT flashed over(I noticed this yesterday).Anyways as last attempt I replaced the failed halfbridge with a capacitor voltage divider and it is working now for 3 days!Well,it is not working now,because the driver has an issue;a smoothing tantalum capacitor shorted out and I have to replace it,but also upgrade a bit on the design and use a transformer instead of the ATX PSU.
I tried several things with this inverter,so far so good,it works like a charm expect the flashovers from secondary to primary even after decreasing the turns on the conical primary.
Yesterday I was looking for those fake QCW interrupters to create long and straight sparks and ended up with a combined Kaizer VTTC staccato controller and loneocean's RSSTC3 interrupter.I made this circuit on a breadboard,it was not working as it was supposed to,but probably due to bad connections on the breadboard.While testing it produced 40cm straight arcs to ground,which I think is really good.Sadly at this point the capacitor blew up and I couldn't make a video of it.

I have a video with 8 primary turns and with the complete conical primary.With that amount of turns doesn't exist flashovers,but with 4-5 turns the flashovers were significant.
I don't know why is the inverted working better with 2 transistors instead of 4!?In comparison with the fullbridge now it has +-325V/2 so like +-162V across the primary instead of the +-325V.Double the voltage,double the size of the arcs,if the inverter is working correctly,right?

https://mega.nz/#!ZeJVnKoR!FvYfJYxrESm7BmyOdQYU9mSWFapb9-rF1G6Ag1iQua4

Great that you get all the small errors sorted out and keep the coil running!

Half the primary turns does not nescasarily mean double the current, you also have some stray inductance in wiring and bridge, but you are pushing much higher currents through the transistors and it most likely what is blowing them up. While you have two transistors in series conducting the half the voltage, the current through them is the same as if a half-bridge was driving the same load, but the power dissipation is lower.

Great video with a good clear MIDI like sound reproduction.

You should challenge your photographer friend to get crystal clear single spark shots and he will soon realize that it requires one hell of a camera