Mosfet stack for higher breakdown voltage

[magpwr]

hi Jeg,

Nice work there but i think you have overlooked "the cost" of stacking many mosfet together including the other supporting components.

Maybe 2 600v mosfet stacked is still ok to get more than 1000volts.

Any higher mosfet count you would also need to take heatsink\cost of heatsink and space usage as well.
It is not a practical solution.sorry.

[d3x0r]

I saw a similar circuit, and considered doing one myself, but the gate for each one has to be relative to the divided voltage.  (if you have 10 500V, blocking 5000V, then each is 500V above the other....)


I saw a schematic for one, he used normal npn transistor to trigger a small transformer that was connected to that transistor on one side, and the source and gate of each mosfet down the chain, so the voltage difference on the base is relative to is's source voltage....


I can't find that circuit now....


https://pure.strath.ac.uk/portal/files/19755869/A_Magnetically_Isolated_Gate_Driver.pdf  (this shows the idea, it can be carried ad infinitim.... )

[fritz]

The old school way seems to fit to my needs! I will re design it today. Thanks Fritz!

There was some fear that you will jump on that train;-))
That works fine for gate capacities up to 1nF - which should be fine for high-voltage mosfets.
If you want fast, reliable, top notch switching - there is no way around integrated drivers.
Another issue is that there is a small but noticeable capacity between drain and gate.
If you turn off a coil, the fast rising drain voltage may lift your gate voltage depending on your driver output resistance.
So it´s possible that the gate-source voltage rises above absolute maximum rating and kills the mosfet.
Even an external zener diode (12v) will be too slow to prevent that. Yet another reason why integrated drivers are worth to think about.

About the ignition coil, i will follow your advice by feeding it through a mosfet!

My point was to use galvanic isolation (mains transformer, batteries, isolation transformer) to prevent electrocution.
There is appropriate power @ 40.000V available.
Pls. be careful.

Having an oscilloscope and HV probe (Tektronix P5100) or things like that:
http://www.ebay.com/itm/New-4KV-100MHz-High-Voltage-Oscilloscope-Probe-100-1-Ship-From-USA-/271189553044
would be useful to see whats going on.
A normal 1:10 probe(rated 250AC) will die after few 600V spikes(lessons learned).

rgds.

[fritz]

Nice primer on that topic from people who should know:

http://www.irf.com/technical-info/appnotes/an-937.pdf

[Jeg]

 Why always so many solutions!!! Couldn't it be just one for every problem??? hahaha

Magpwr, it's for experimentation and not for productivity. But, you are right. I should start considering the cost also! By the way, heat-sinks were taken from an aluminum door rail. see in pic.

d3x0r, firefox prevents me for opening this site. I'll try to locate it thanks

Frirz i thought that ign. coils don't go over 25KV but i'll give it a try. I think that i have two transformers 220V to 12V. I'll make an isolation with them and fire it. I have used ign. coil again driving it with an IRfP460. Never thought of driving it from the main supply.