Experiments with a (DR)SSTC2

After all the fun I had with the small pretty SSTC I decided to make a more powerfull one. As I still had a few of the IRFP460's I made the SSTC2

Primary: 30T 2mm copper on a 4" PVC former. (runs at 20T) & 1.5nF MMC
Secondary: 11" 0.225mm enamelled copper wire around 1500T on a 3" cardboard former.
Switching: 4xIPFP460's in full bridge, GDT drive, from 240V supply, bridge rectified and smoothed.
Driver: Secondary base current feedback, Pulse width and repitition rate controlled,

The whole lot was designed using my SSTC calculator, this ment that I couls change values easily and see what the effect would be on the coil. Afer I got the bugs out of it it did work very well.

The Driver, built into a small alu case to prevent interferance from the working coil. The Secondary base current transformer is on the LHS, and there is two variables for the pulse width and the repitition rate on the RHS. I drive the GDT from a full bridge of  IXY GDT drivers these are the two un-heatsinked TO220 case's on the LHS.

Circuit diagram

The Driver consists of 4 IRFP460's, each transistor has its own gate resistor and a pair of back to back 15V zenners to protect the gate from over voltage. This picture shows the construction of the GDT very well, 10T of 5 wires, each twisted together to prevent stray inductance, each winding is then split out, 1 to each mosfet and one from the driver..

The Bridge sits on an old pentium 3 heatsink, with no fan, It won't run CW, but with interruped drive, it hardly gets warm at all. Each IRFP40 is insulated from the heatsink by an insulating (but heat conducting) pad.

And of course more power....

I slowly wound up the voltage, BANG!!!, one full set of mosfets died.

I replaced all of the MOSFETS and tried again. This time with a home made current transformer around the primary feed. This showed 13A peaks with about 200V in, so I added a few more turns to the primary. I now had 7A peaks (and I'd fixed a bug in my SSTC calculator) with 20T on the primary.

I went right up to full rectified mains in 340V.

11" sparks. & Hot 'n' jucy ones too..

At full power, I stated to get some primary-secondary sparks, so I swapped my sphere topload for a 8" diameter toroid made from 4" ducting, it cured the problem.

The coil seams to run quite reliably now, no more mosfet deaths (Yet!!), I need to secure the components and build a 12V PSU, as im running on batteries at the moment.

After some lower power experements with salt on a breakout, I tried with this coil, Too much pulse width, and the arc was too strong to see the effect, but with less power the effect of the salt is clearly seen,
 

I tried the coil with a primary capacitor making the SSTC into a simple DRSSTC (dual resonant SSTC).

By calculation a 17nF cap should make the primary resonant with the secondary.
I have 15nF at present and the performance tails off with any increase or decrease in capacitance, so I suspect the cap is about the right value. 

The results were really quite impressive, I have increased the max spark length to air by about 60% I need to see what happens at full power. I'm not sure if the fet's will stand the large current pulses.
 

One way to find out though..

Raising the voltage over 200V blew half the bridge.

After further investigation this was due to the increased current draw of the FET's with the series capacitance.
Using my current transformer, I confirmed that at 100V I was getting a draw of 10A, where as without the series capacitance I would only get a peak draw of 7A.

The question is now what to do next?, It probably means a change of the bridge from FET's to IGBT's to allow for the increased current.

Experiments with a (DR)SSTC2

A Slight Diversion.

Whilst deciding if I aught to change to IGBT's I had an idea. I was looking through my workshop and found a large bulb that I'd bought ages back, and the SSTC was sitting in the corner un-used.

So....

I added a 1000W 240v bulb to the toroid to give some Plasma globe effects.

Pictures below..

Running at low power

Click Image for better picture

Running with more power

Click Image for better picture