One wonders why you didn't simply use the exact circuit given as Figure 2 in the pdf you linked. It switches 1500 V, uses the dirt-cheap IRF840 mosfets, and common 2n3904 and 2n3906 transistors to drive. Just about any signal, like from your 555+comparator PWM, could feed the transistors, I suppose.

Hi Tinsel, tnks for the suggestion

My driver has an output of 12V peaks and that makes the input capacitance of IRFP460 as large as 41nF!. Perhaps I am wrong but i thought that i need Amber level peaks. I think i have some of them.. i'll try it tonight.

IRFP460 is my favorite one and i have many of them. It handles more current than IRF840 and this is very convenient to me.

I use the P460 a lot too; it is the mosfet I use in my little 4MHz SSTC, and in some other projects, but your system's current demand will be determined by the load. The 840 system in the schematic apparently provides the maximum current that a 50 ohm load can draw from a 1500 volt source, and the scopeshot shows an impressively short risetime for the pulse. I don't know what the max frequency could be with that schematic, but I'm already working on a solid state Tesla Coil based on it.

I'll be very interested to see what your final project looks like. I love HV, but recently I've been working with low-voltage high current mosfets like the IRF3205, which can handle 110 amps in a TO-220 package, with Rdss of 0.008 Ohm. Daisy-chaining a few of those to get to higher voltage handling might make a real "killer" SSTC primary driver!

Wow gyulasun..i was trying many days to find this article. Tnks for the link.

Tinsel i finally did it. From what i learn until now...

1. I doubt that schematic in fig.2 in the link i posted above works as it is. It needs lowering the 10M resistances. The reason is the leakage current between Drain and source. With resistances so large like the 10M, the voltage that mosfet's share is not equal. With pots in the place of these resistances, i found out that my IRFP460 stack, equalize the voltage at about 80Kohm

2. 12V for pulsing is not a good idea. 18V is better because Ciss and Crss are much much smaller!

3. The 3904 and 3906 idea is fine up to about 2KHz in my 460 stack. After 2Khz the switch starts to close. The reason is that it needs power. MC1406/7 puts out enough Amber to charge quickly the input capacitance. The above transistors in the mA range are not enough. In addition, this transistor driver doesn't work with voltages lower than 120V as in my 48V case. In case of using this driver, you have to change the 470pF if you have different than 1500V supply.

Does anybody have a suggestion for a dc to dc boost converter? I'd like to make 12V to 500V or higher, for making a bigger supply for the stack.

Tnks
Jeg
     

Tinsel do you have a proposal for making high voltage supply with mosfets? I don't like transformers for some reason!

In the attached picture you can see  a stack of 14 Mosfets capable for handling up to 7000V. At least in theory... cause until now i have burned twice the hole stack. (28MOsfets!!!) My electr. components supplier is making good business with me!

...
Does anybody have a suggestion for a dc to dc boost converter? I'd like to make 12V to 500V or higher, for making a bigger supply for the stack.

Hello Jeg,

You may find this link useful:  http://electronics.stackexchange.com/questions/31833/3v-to-500v-dc-converter

It refers to an EDN design idea article on a 1kV DC upconverter from 12V: http://www.edn.com/design/power-management/4334168/1-kV-power-supply-produces-a-continuous-arc

This page may also serve some food for thought: http://www.eleccircuit.com/dc-power-supply-300v-from-battery-12v/

Gyula

Hi Jeg,

Some folks do it with ignition coils;-)))

http://www.instructables.com/id/Variable-voltage-ignition-coil-power-supply/
http://www.instructables.com/id/High-voltage-ignition-coil-supply/


rgds.

Hi Jeg,

What concerns the ignition coil supply - I´d like to point out that you should operate that from batteries or isolation transformer. Be careful.

Mosfet Stack:
If you apply 48 Volts to a 127uH coil - you reach 0,4A within a microsecond, 4A within 10 microseconds.
If you triple the load capacity of the fet driver by stacking 3 fets - it takes probably triple the time to charge the gates.
So I would run for proper integrated mosfet driver for each fet.
For fast switching a driver capable of 1A peak.
I personally use FAN3100 from fairchild - thats a sot23-5 package - maybe to small for your flavour.
The "old school" way (for people using 555s) was to take 4049 or 4050 CMOS chips (inv or noninv.) and connect all gates in parallel,
or simply enhance current sourcing by using 2n2222 2n2907 complementary stage.

rgds.