Kaizer DRSSTC I

WARNING!: Working with electricity is dangerous, all information found on my site is for educational purpose and I accept no responsibility for others actions using the information found on this site.

Read this document about safety! http://www.pupman.com/safety.htm

Introduction

Building a Dual Resonant Solid State Tesla Coil have been the ultimate goal since I started experimenting with high voltage apparatuses 3 years ago.

A DRSSTC is the modern day topology of driving a Tesla coils taking advantage of IGBT technology, pulse rated capacitors and a very low inductance primary circuit layout.

Considerations

Due to IGBT technology the resonant frequency have to be lower than what is normally acceptable in a regular solid state Tesla coil. The higher frequency the more demanding is the current delivering ability of the gate driving circuit and switching losses will stress the bridge more than good is.

Specifications

Bridge	4x IXGN60N60C2D1 IGBTs in a full bridge configuration
Bridge supply	0 – 260VAC through a variac, 2x 35A rectifier bridges in parallel and 2x BHC 3300uF 450V filtering capacitors in parallel. (each: ESR 39mOhm@100Hz, Z 27mOhm@10kHz and 53A Iripple@70KHz@50 Degrees celcius)
Primary coil	320 mm diameter, 10 mm diameter copper tubing ~ 28,27 mm², 9 windings. Tapped at 5,4 turns.
MMC	6 strings in parallel of 2 in series Cornell Dubilier (CDE) 942C20P15K-F capacitors for 0.45uF at 4000VDC rating and 81 A Irms.
Secondary coil	160 mm diameter, 605 mm long, 2200 windings, 0.25 mm enamelled copper wire.
Resonant frequency	Around 65 – 70 kHz.
Topload	127 x 620 mm aluminum flex tube with aluminium tape toroid.
Input power	130BPS, 9 cycles, 500A limiter: 2500W at 250VAC at 10A.
Spark length	Up to 1500 mm long sparks.

Schematic

Bridge section

Driver section

Same as Steve Wards universal driver version 1.3. Just made on single sided PCB without SMD components.

Construction

20th March 2009
Bought 30x IXGN60N60C2D1 from Digikey USA, import taxes etc. almost killed me

7th April 2009
Bought 60x 942C20P15K-F capacitors through dr.spark, again hello import taxes

15th May 2009
Bought heat sinks cheap from Germany

19th May 2009
Started converting Wards latest DRSSTC driver to single sided board

23rd July 2009
Started 3D designing the full bridge

It is bulky, on 2 heat sinks with capacitors between them, its too big and needs to be overhauled.

23rd August 2009
Redesigned the full bridge

The two IGBTs in the middle are turned 180 degrees to have the supply at one side of the heat sink and output on the other, it got compact, neat and only one overlap with busbar, I am very happy with this design.

24th August 2009
Made a spreadsheet to ease experimenting with different DRSSTC settings, this was also a try to collect some of the different theory and put it side by side, it might not all make sense

27th August 2009
Bought 50 meter of 10 mm copper tubing, two drain pipes 160 mm diameter x 1000 mm and 1,5 mm copper sheet.

Here is the collection of parts for the DRSSTC.

8th September 2009
Finished converting Wards latest drsstc driver to single sided board

12nd September 2009
Finished assembling the single side board ward drsstc driver

16th September 2009
Etched, assembled, tested and housed interrupter with burst mode.

19th September 2009
Cut out busbar from copper sheet and assembled IGBTs, heatsink, capacitors with busbar, home made 8mm brass spacers are used. Leads to the primary are 16mm^2 stranded 90 Degree celcius machine tool wire.

25th September 2009
Found a aluminum box and transformer for the driver, wound a GDT and started preparing decoupling capacitors, TVS and zeners for the bridge.

With all the wonderful theory about GDTs in the wiki, it would be a shame not to check it out instead of just going for a high permeability core with 10 turns on CAT5 cable. So here goes

Its a Epcos ringcore, material: N30, good up to 5MHz, Aemm^2: 95,89, AL: 5750nH

Inductance with 10 turns: L = AL * N^2 = 5750 * 10^2 = 575 uH

Peak current: Ipeak = (Vin * t * D) / Lmag = (24 * (70000/1000000) * 0,5) / (575 * 10^-6) = 1460 mA

Irms = Ipeak * 0,577 = 842 mA

Minimum number of turns needed to avoid saturation

t, 50% duty cycle = (1 / 70000) / 2 = 7,3*10^-6

Nmin = ( V x t ) / ( B x Ae ) = (12 * (7,3*10^-6)) / (0,2 * (95,89*10^-6)) = 4,6 turns

Current needed to drive a single 60N60 IGBT gate

I = Qc / t = (146*10^-9) / (1/70000) = 10,22 mA, including magnetising current, double this figure.

So it all seems to have overhead enough to drive a full bridge.

3rd October 2009
Started on contruction of the MMC

10th October 2009
Made the round platform plates for the coil to be built on, they were cut out from 19 mm MDF wood plates with a modified router, also shown in the picture, very neat for making circular cuts.

11th October 2009
Finished MMC construction, features a 80 mm fan that delivers 30 cubicmeters/hour of air.

5th November 2009
Debugged PCB design of the driver Forum thread link

Tested driver and interrupter on a small DRSSTC I put together just for testing purpose, blew the half bridge when I ran it in CW without feedback, I guess there is no way I could have treated that poor little coil any worse…

12th December 2009
Made primary form and winded the primary coil onto it, with a strike rail and mounted on the upper platform, took me 1½ hours just to wind the coil through the holes and also spend quite some WD40

The coil is a true helical coil with 4 mm steps between the supports

27th March 2010
It is far from satisfying to wind half a secondary to learn that you have used a ruler with 2 scales and you started with a 100 mm offset in the wrong direction…

The second try on the secondary was winded in 2 hours using my new coil winder rig, its 2200 turns of 0,25 mm enamelled wire on a 160 mm diameter pipe, winding length is 605 mm, its currently hardening its second layer of varnish till I get time to visit my parents again as its staying at their garage while getting varnished.

2nd August 2010
I made a fitting piece for holding the CTs as the cable shoes were too wide to go through the CTs

7th August 2010
Today I assempled some of the parts on the platform, its beginning to look like a DRSSTC! Also I painted it black some weeks ago!

9th September 2010
The topload is done, measuring 130 x 620 mm, made from aluminium ducting on a wooden form, smoothed with filler for metal and covered in aluminium tape.

17th October 2010
Final testing of driver features, got a 555 acting as a feedback at 70KHz while my signal generator is used to simulate over current input signal.

FIRST LIGHT! YIPEE!  See further down for demonstration video.

22nd October 2010

The layout of the electronics have reached their final state

3rd November 2010

I made a new interrupter with selectable BPS, either from 3 to 15 or 130 to 500. On time from 1 to 20 cycles and burst mode.

5th November 2010

First test run, the coil is still slightly out of tune. Achieving sparks around 120 centimetres.

After a few adjustments of no more than 5 centimetres on the primary coil at a time, what seems to be the sweet spot have been found. Sparks will now fly out at 145 centimetres.

A run was made with the breakout point going straight up, beautiful sparks and some very heavy sparks directly between topload and earth rail.

A final adjustment of the primary tap would be the end of the days testing, the variacs 10A fuse blew after the coil had been running for 5 minutes.

150 centimeter long sparks! Running from 250VAC in at 10A, 9 cycles ~200-225uS on-time, 500A limiter.

Conclusion

A years work have come to an end with a result I am very satisfied with and still I did not make a spark longer than my own height, which was one of my goals.

The metal filler used for the topload to smooth the surface is way too hard to sand down without damaging the aluminium tubing underneath. I will use ordinary wall filler if I use this method for topload construction again.

A full bridge of IXGN60N60C2D1 SOT-227 package IGBTs is slightly too small for a Tesla coil this size, I will upgrade with some heavier silicon, preferably a Powerex CM300-24H brick.

Demonstration

17th October 2010: First light

5th November 2010: Stress test