Kapanadze Cousin - DALLY FREE ENERGY

[d3x0r]

hi everyone,

It seems strange i get better light output on my 220volts 60watt bulb upon disconnecting one of the two wires from toroid to IGBT transistor collector.

The maximum output for me is achieved using 1.07Mhz for kacher coil which means my PWM circuit is automatically divided by 60 at 17.83khz at 47% duty.
Total input power =55watts  at 30volts.

1.03uf capacitor from toroid 3 turns to 25 turns bifilar.

I'm curious why 1 IGBT transistor produce better output than 2 transistor.Next stage i will reduce duty cycle from 47% to 23% and use 2 transistors at 23% duty.

I'm using the connection to kapanadze coil base on attachement.Note there is no grounding at 3 turns in this version.

Does it matter which IGBT?

[Jeg]

Hi mag, merry Christmas.
It is strange indeed, if it works better after correcting again your pwm frequency. Because when you disconnect one mosfet then the low KHz frequency drops to half of the value than when both mosfets are in operation. By the way, nice looking job mag, well done.

[magpwr]

Does it matter which IGBT?

hi d3x0r,

Both IGBT is working fine.

But strange thing is if i connect to both IGBT the total Amps drops a little and so does the bulb output.

Duty reduce to 23% from 47% but still single IGBT produce better output than dual IGBT.

Both IGBT is driven by TC4452 (12Amps Mosfet driver) supplied by 7818 (18 volts)voltage regulator.

The 12-0-12 winding with center tap is one direction only.

I think time for me to swap wiring to 28 turns winding  maybe it will solve this problem.


Merry Christmas and a Happy New Year everyone.

[NickZ]

  I disconnected the jumper from the 3 turn coil's tuning cap to the grenades 168 turns output coil Didn't make any difference.

  I still can't light a small ac bulb from my Kacher's output, either.
My Kacher's output may be too weak to power a small 20w ac 110v bulb.

  Now I'm back to using using the IRFP260N mosfets, and I'm also back to my normal circuit's output, which is able to light several 110v, 100w bulbs.
However I can only do it for a few seconds, or my fets will melt down in less than a couple of minutes. One of the fet gets untouchably hot in less than 5 seconds, the other fet stays relatively cool, to just warm. This happens even without running a load of bulbs on the circuit, (no load).
I need to find a solution to this overheating issue which has been my main problem for over a year now.

[magpwr]

hi everyone,

Today i've tried various wiring combination to achieve the maximum brightness on the 220volts 60watt which was on the verge of destruction.
I am able see the filament becoming white even though my bulb is coated.


I have attached the wiring from toroid or yoke to the kapanadze coil to achieve this effect.

Take note my duty cycle is mere "23%" just by connecting full bridge rectifier from 28 turns to bulb with 20uf 400volts PIO capacitor do not even give brightness as close to the new wiring which i have done.

I have just blown one IGBT(internal short) due to insufficient cooling at 100watt total input power at "23%" duty.(Take note Ruslan device was consuming around 100watt at 24volts 4.5...5Amps)

The blocking diode 25Amps "600volts" rated is connected to 12-0-12 turns center tap via mini toroid inductor(5 turns).

Next stage to lower duty cycle to 18% or less after replacing IGBT and later in the day i will also be testing out with 1KW 220volts halogen bulb.

The key is to achieve similar brightness at below 20% duty cycle using this attached wiring method.

-----------------------------------------
Latest update -
I've managed to lower duty cycle to 20.4% this is the lowest setting base on the capacitor value used in my circuit and also added a fan to cool the IGBTs.
The bulb was lit at 210volts dc before the IGBT died again."This was the same brightness at 23% duty which i thought the bulb was lit to the max."

Final plan for the night to replace the IGBT and capacitor for my DIY 74HC132 dual PWM generator so that i can get lower duty %.

Time for my cat nap to recharge.