mosfet shorted after i connect them in parallel ? did i doing it wrong ?

[tonny]

one of my mosfet source-drain shorted after i connect them in parallel ?

so i was making an ARC speaker yesterday,i use 1 mosfet ( irf540) ,and it gets really hot,but it doesn't blow ,so i decided to put another mosfet connected in parallel ,and before i connect them i have tested both mosfet and both of them work normally , but when i connect them in parallel ,it shutdown the immediately due to overloading ,and i find that on of my mosfet was source-drain shorted thats why the power supply shut down to protect itself

how do you connect the mosfet ?
just normal parallel circuit gate to gate,drain to drain ,source to source ,connected with a piece of wire

pwm controller ?
TL494CN ,powered by a battery,completely isolated from the mosfet side

mosfets ?
IRF540 from IR


heres the full setup

as you can see from the video there are 2 mosfet on the heat sink but the lead was disconnected because it was shorted
oh btw this video is not specially made to show the setup so it would be more focused on the flyback

 

[Bluejets]

There is a summary at the end of this particular pdf of things to watch when paralleling mosfets.

www.irf.com/technical-info/appnotes/an-941.pdf

 

[(*steve*)]

It doesn't appear that you have separate gate resistors and your wiring could not be described as "low inductance".

I would suggest the problem is caused by the lack of gate resistors causing one mosfet to be held in its linear region while the other switched off. This results in a lot of heat and a dead mosfet.

Any more specific advice would require we know more about the load and (especially) the gate drive circuit.

 

[GPG]

What BJ said. And
http://www.microsemi.com/document-p...arasitic-oscillation-between-parallel-mosfets
And your gate drive current is shared leading to higher switch times. The TL494 may be a bit lacking. What battery?

 

[tonny]

heres the full schematic and the pwm board

http://s1072.photobucket.com/user/T...151028_161700_zps0bubbxio.jpg.html?sort=3&o=0

http://s1072.photobucket.com/user/T...[user]=143082917&filters[recent]=1&sort=1&o=4

 

[GPG]

I think you should raise the drive current. Perhaps something like
http://www.ti.com/lit/ds/symlink/ucc27322-ep.pdf
There are others. You may find that the single fet is cooler.
There is also the possibility of uncontrolled flyback voltages on the drain. Got a scope? with HV probe?