Do not underestimate the power of the brick. V2

Kizmo, Sun Sept 23 2012, 08:07PM

Just little bit something i wanted to share with you guys.

Inspired to by this thread: Link2


My victim was Semikron SKM150GB128D which i got from German ebay for 20 euros.

Datasheet: Link2


I wanted to see what kind of power levels and general abuse brick like that take. So far it has been very interesting.

Coil spec:

160x500mm secondary, about 1550 turns
100x490mm topload
flat pancake style primary, k=0.16
667nF tank capacitor
370VDC bus, 5600µF bus capacitance
Single phase 32amp feed

Secondary unloaded resonant frequency was measured to be 115kHz, primary was tuned to around 104kHz.

First test setup looked like this. Later i added some more capacitors in series and made nice sealed metal box for driver. (Whole thing was put together from scrap parts in two days, it sort of shows in some places cheesey )

1348429593 599 FT0 Done


First test at my student apartment, seriously limited by 10A breaker, room and noise.


2nd test later at other place where noise, room or power are not problems..


As video title said, i didnt have Over Current Detection in use. I did scope primary current and during air streamers it was around 1.5kApk and during heavy ground arcing it went way past 2kA. Coil seems to be making scariest sparks at around 300-400bps and 100-120µs on time.

Once i chose primary components i knew this thing would need serious amounts of primary current for impressive spark output. Low impedance primary (Zprim 2.22ohm), half bridge and relatively low input voltage made sure i was not wrong. (its hard to get any sparks out of that thing with less than 500Amps in primary!)

Coil ran just fine at 3-4kVA input power from wall for quite some time. Then primary cable that was leaning against bus capacitor mounting strap induction heated that steel strap to a point where cable insulation failed and it shorted primary lead to house ground. And it of course blew rectifier bridge and fuses from electrical panel.

Seems that everything else survived. Maybe next weekend we will see series-parallel bus capacitors for 3 phase power (565Vbus?) Maybe OCD will be added too..


Re: Do not underestimate the power of the brick. V2
Killa-X, Sun Sept 23 2012, 08:22PM

Nice job man, No need to build a fancy coil to get a fun toy :)

Hope to see more limit pushing! Maybe until utter destruction? How much will it take, hehe :D
Re: Do not underestimate the power of the brick. V2
dex, Mon Sept 24 2012, 09:07AM

Kizmo wrote ...

I did scope primary current and during air streamers it was around 1.5kApk and during heavy ground arcing it went way past 2kA. Coil seems to be making scariest sparks at around 300-400bps and 100-120µs on time.
Does this mean that "heavy ground arcing" occurs before on time is over?
Re: Do not underestimate the power of the brick. V2
Steve Conner, Mon Sept 24 2012, 09:21AM

It certainly can depending on your coil setup. Classical Tesla coil theory assumes that the energy transfers quickly into the secondary, and the streamers and sparks appear after the energy transfer is complete, but the DRSSTC doesn't work like that. The streamers are being "fed" with energy from the DC bus by the inverter while they grow, and maybe develop into a ground arc. If the ground arc gets too heavy, the inverter may shut off (Steve Ward style) or go into a pulse skipping mode to limit current (my style) or ignore it and carry on (Kizmo's :) )

In fact I would argue that good performing classical coils didn't work like the classical theory either. Once you have a decent amount of bang energy, it's not possible to transfer it "quickly" without causing racing sparks.

Good luck Kizmo on your quest to burn up every part except the IGBT. tongue
Re: Do not underestimate the power of the brick. V2
Kizmo, Mon Sept 24 2012, 10:16AM

Actually this style of driving is called Dalus mode (at least in chatroom..). Dalus has 20kVA low impedance liquid cooled dual h-bridge drsstc (can be found from projects board, search for drsstc named as "resonant rise").

I just find it cool how long on times and high peak currents produce fat and "hairy" ground strike sparks..
Re: Do not underestimate the power of the brick. V2
dex, Mon Sept 24 2012, 10:34AM

Thx for the info.
Re: Do not underestimate the power of the brick. V2
Intra, Mon Sept 24 2012, 05:44PM

post please model number of this brown round caps for mmc?
Re: Do not underestimate the power of the brick. V2
Kizmo, Mon Sept 24 2012, 07:32PM

Caps im using are 4µF 600VAC/1000VDC/1600Vpeak GTO snubber capacitors. 6 in series for 667nF. Rated RMS current is 70A and god knows how high the pulse rating is. I dont remember model number. I have to say they are awesome caps. I pushed that coil to a point where primary cable coatings softened and caps were stone cold :)
Re: Do not underestimate the power of the brick. V2
Intra, Tue Sept 25 2012, 04:52PM

Please. Could you cast a look on this caps or you very far from it? I'm really very important to know model number, because I'm looking for High RMS Caps for my Guitar Interrupter.
Re: Do not underestimate the power of the brick. V2
Kizmo, Wed Sept 26 2012, 03:14PM

Intra wrote ...

Please. Could you cast a look on this caps or you very far from it? I'm really very important to know model number, because I'm looking for High RMS Caps for my Guitar Interrupter.
PM sent
Re: Do not underestimate the power of the brick. V2
Kizmo, Sat Sept 29 2012, 08:52PM

Aaaaand we have results. Well, sort of. I finally killed that IGBT. But it did not fail because i had too much power running through it.

I cranked the bus voltage up to 400VDC (3phase supply). I kept pushing it harder and harder finally getting almost 4x secondary lenght sparks to grounded objects, OCD was set to 2kA and the OCD led was blinking constantly. When playing music, input power peaked close to 6kW (i was measuring rectifier output current and voltage..). Then, BZZZKRZ..BLAM *sparks flying from base*

What i found was not something i expected to see:

I got little bit carried away with power and sort of forgot that my primary cabling was not up to several kW range cheesey Cables got really hot and sagged, insulaation failed and tank capacitor got shorted to inverter output.


1348951155 599 FT144858 Fail


When i saw this i knew instantly that it was game over. Gate discharge diode was blown in two pieces.

1348951466 599 FT144858 Fail2


The IGBT wasnt very happy

1348951671 599 FT144858 Dead2


Oh well. Another item to the wall of fail.



But this does prove that these igbts can handle insane punishing when everything is set properly. If i ever get similar IGBT this test will continue.
Re: Do not underestimate the power of the brick. V2
Gabriel35, Mon Oct 01 2012, 08:17PM

Pretty interesting Kizmo, keep the test up!
Re: Do not underestimate the power of the brick. V2
Thomas W, Mon Oct 01 2012, 09:10PM

thats awsome, i wish i could get myself together and build one myself hehe, pity about the IGBT faliure.

--Thomas Williamson
Re: Do not underestimate the power of the brick. V2
Bennem, Tue Oct 02 2012, 07:39AM

Did you use a phase lead driver for this?
Re: Do not underestimate the power of the brick. V2
Kizmo, Tue Oct 02 2012, 09:25AM

Nope. Still dont know what is wrong with my UD2.


It is switchign ~560ns late but since the bridge was so heavily snubbered and decoupled its not very big of an issue. Except that HF noise...
1349169921 599 FT144858 560ns Late
Re: Do not underestimate the power of the brick. V2
brtaman, Tue Oct 02 2012, 07:59PM

Great work, Kizmo! :)
Re: Do not underestimate the power of the brick. V2
Kizmo, Fri Nov 02 2012, 08:59PM

And our story continues. This coil doesnt need dark room for sparks to show cheesey

Coil is now rebuilt with SKM200GB-128D full bridge

Bus voltage in this clip is 250VDC (dips below 200 during long notes..) and current meter has scale from 0 to 15A with 5A steps



This run was limited by floor and ceiling. I really did not want any sparks to ceiling as all ccfl fixtures and their power wires are there.

If i ever want to push even more power through this thing i really have to redo primary to get higher primary surge impedance and then crank the bus voltage to 500+V

Still, cool
Re: Do not underestimate the power of the brick. V2
Steve Conner, Fri Nov 02 2012, 09:53PM

Neat! :) What are you using for the DC power supply?
Re: Do not underestimate the power of the brick. V2
Kizmo, Fri Nov 02 2012, 10:10PM

That table with meters is just 3 phase isolation transformers and 15kVA 3phase variac. Followed by 6-pulse rectifier and more panel mount meters
Re: Do not underestimate the power of the brick. V2
Steve Conner, Fri Nov 02 2012, 10:27PM

No shortage of DC power then, I guess! :-o
Re: Do not underestimate the power of the brick. V2
Kizmo, Fri Nov 02 2012, 10:31PM

Yeps but judging from the heating of my single 5600µF 420V bus lytic i really need to modify this coil to accept 500+V bus voltages.

There are some serious currents now cheesey
Re: Do not underestimate the power of the brick. V2
Kizmo, Sat Nov 10 2012, 02:39PM

Its getting pretty awesome


1352558228 599 FT144858 350vdc2






Sure sparks this long could be done with less power but length is not only thing i want to see cheesey
Re: Do not underestimate the power of the brick. V2
zilipoper, Sun Nov 11 2012, 12:33AM

Rock n’ Roll Racing the best!!! amazed
Re: Do not underestimate the power of the brick. V2
Josh Campbell, Sun Nov 11 2012, 07:21PM

Love it! Great work!