Re: [TCML] flyback driver sucess.

[bturner@xxxxxxxx]

Scott -

A few things - what you need is LOTS of gate current on the FET. I am
assuming it's a "HexFet" or similar - these FET's need lots of silicon
surface area to handle the drain current, which means lots of gate surface
area and subsequently lots of gate capacitance. The usual way of driving
the gate is via a totem-pole bipolar transistor 'front end'. In my
original design (Do a web search on Brent Turner and Solid-State Tesla
Coil - it was printed in Radio Electronics and also Electronics
Experimentor in the early 90's...) What I did was simply use a low-value
series gate resistor and an NPN transistor switching the gate to ground
with a pull-up resistor on the collector.

What you need to do is switch the FET off very quickly, thus collapsing
the magnetic field in the flyback primary winding. You ALSO need to snub
or clamp the resulting back-EMF which can reach 2-3x the FET's maximum Vds
rating. I simply tinkered and found a good capacitace value to act as a
'primary resonance capacitor'... Though in this circuit topology, the
primary 'tank' really doesn't oscillate.

And yes, use a good heatsink and thermal goo on the FET.

I used a Samsung FET, 800V, 15A if I remember right. Used a 555 timer
oscillator and the whole thing ran on 12 volts. I was able to get a good
1/2-3/4" flaming discharge to ground off the SMALL flyback transformer.
Drove a Plasma lamp to incredible brilliance.

- b



> It was an optocoupled FET driver, 2.5A peak output, and it still works
> surprisingly...  I'll work on a schematic, it was pretty plane Jane
> however.  After thinking about it, I'm suspecting it overheated, the
> heatsink was very undersized (a 2*3 inch * 1/8 aluminum plate) and the
> transistor was not fastened tightly.  It might have been voltage killed,
> it
> was not running zero current or zero voltage switching at all, and I had
> no
> snubbing of any kind, but it was a 400V FET if memory serves, and it was
> only running 25V...  I'll try again using a proper heat sink with thermal
> compound and see how it goes...  Thanks for the quick reply!
>
> Scott Bogard.
>
>
>
>> Do you mean an optocoupled FET driver, or just a plain phototransistor
>> optocoupler?  If possible give us a schematic.
>>
>> >  Also I believe it was pumping out 18V on the output, is that too much
>> > voltage to drive a gate, I believe they are 5V gates...
>>
>> It's possible, but most power FETs can handle gate voltages of 20V at
>> least,
>> even if their threshold is a lot lower.  You'd have to check your
>> datasheet
>> to be sure.
>>
>> Another possibility is that you exceed the drain voltage the FET can
>> handle.  You'll probably want a clamp on the primary side, and maybe
>> some
>> kind of snubber as well.  And you could have just given it more current
>> than
>> it could handle.  Anythings feasible, but it's hard to be sure without
>> some
>> hard numbers to go on.
>>
>> -Mike
>> _______________________________________________
>> Tesla mailing list
>> Tesla@xxxxxxxxxx
>> http://www.pupman.com/mailman/listinfo/tesla
>>
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla
>
>


_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla