what things.. cause mosfet gates to die? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

what things.. cause mosfet gates to die?

1 2 next

Move Thread

LAN_403

Killa-X

Sun Apr 12 2009, 05:07AM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

Today, I finally got a demo tesla coil to run. with about 2-3 inch streamers around the topload. Sadly, I could do better. my leyden jars consist of a thin crystal light container, and foil. Knowing this, all 3 leyden jars had a TON of 1cm tiny arcs all around them.

to the point. It randomly, shut off. My 555 box, the pot was warmed because the 3W resistor under it was hot. I checked the mosfets emitter -- source, Open. No short. This was weird. I put a speaker on it, faint high pitch ring, I could tune it. I turn the 20V on, EXTREMELY LOUD ringing, and well, smoked the speaker to death. Put it back on the flyback, NOTHING. Said wtf, it just worked! Put a new mosfet on, flyback works.

What causes a gate to fail like this?! The stupid 555 only puts out 8-10V with a 16V input (18V 555) and i got a 3W 39ohm resistor on it...It worked all this time, for many days of HV music. and on its first tesla test.. it.. died... My last mosfet died and when I tested the source--emitter, it was a short. Expected. But just the gate?!

Anymore info, please ask.

Arcstarter

Sun Apr 12 2009, 05:15AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

As i said, the 555 will put out more than what is measurable. Put a fast diode on the output and have it charge a small capacitor, then measure :D.

The gate was overvoltage, i suppose. I should have said you should use zeners on the gate. Just take two 12 volt zeners, put them 'back to back' and, then put them from the gate to the drain (middle pin and first pin).

Killa-X

Sun Apr 12 2009, 05:20AM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

Arcstarter wrote ...

As i said, the 555 will put out more than what is measurable. Put a fast diode on the output and have it charge a small capacitor, then measure :D.

The gate was overvoltage, i suppose. I should have said you should use zeners on the gate. Just take two 12 volt zeners, put them 'back to back' and, then put them from the gate to the drain (middle pin and first pin).

well, I did so with a 4007 diode and a 400V470 uf cap and got 16V at 0.192A

Maybe until then I can just use a 9V..Unless i have a 12V regulator around for temporary use..

Proud Mary

Sun Apr 12 2009, 06:23AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Inductive kick back spikes can appear almost anywhere in a poorly decoupled high voltage driver in the presence of a large electrostatic field. Harden up the circuit with liberal use of avalanche and TVS diodes, safeguard the gate with a suitable bi-directional zener, put the whole driver in an earthed die-cast metal box, and filter the mains input. This will probably do the trick, and ought to be done anyway, in good practice.

hvguy

Sun Apr 12 2009, 06:36AM

Registered Member #289 Joined: Mon Mar 06 2006, 10:45AM
Location: Conroe, TX
Posts: 154

What FET where you using? They normally have a 20V gate, and they will take much more, so I doubt your gate voltage was an issue.

Dr. Dark Current

Sun Apr 12 2009, 07:25AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Harry wrote ...

Inductive kick back spikes can appear almost anywhere in a poorly decoupled high voltage driver in the presence of a large electrostatic field. Harden up the circuit with liberal use of avalanche and TVS diodes, safeguard the gate with a suitable bi-directional zener, put the whole driver in an earthed die-cast metal box, and filter the mains input.

I'm not really sure, have you heard about avalanche rating? Most (if not ALL) today's MOSFETs are avalanche rated and will usually "clamp" a current of the same value as their Ids rating and will clamp this current as long as the junction temperature does not exceed maximum allowable. In other words, if they don't overheat, they can't be destroyed by flyback spikes.

To the OP: Do I get it right that you power a Tesla coil with a flyback transformer? Where do you have the 39ohm resistor?

Killa-X

Sun Apr 12 2009, 07:36AM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

IRFP460 is the mosfet, 500V 50A and 20+ max rated gate.

Yes Dr. Kilovolt, With the snubber circuit I have 0.22nf capacitor on the primary, and I can get near 3 inch THICK white (very similar to ZVS) arcs at 77V in.

with only 28V its still ZVS like but a max of 1-2 inches. Teslacoil was able to preform on it. I used a Rotary sparkgap to prevent the caps from charging and just giving me a solid arc. Running a mini 1X8 inch tesla coil with about 34 awg wire from a wallwart.

The 39ohm resistor is from 555 pin 3, to the mosfets gate. arcstarter told me to try a 10ohm resistor for lest 'ringing' on the mosfet. So now it's a 10ohm 10W resistor between 555 pin 3 and the mosfet gate.

And more questions, please ask.

is all I can give you about my flybacks sadly.

teslacoolguy

Sun Apr 12 2009, 01:09PM

Registered Member #1107 Joined: Thu Nov 08 2007, 10:09PM
Location:
Posts: 792

I suspect your gate resistor is way to high, it should be around 5.1 ohms in your application. I would also suggest using some kind of mosfet driver such as a ucc37xxx or at least a npn/pnp pair. The 555 alone may not be able to supply enough power to fully turn the gate on and that could be causing your problems.

Antonio

Sun Apr 12 2009, 02:53PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

Schematic diagram? Without it everything is guesswork.
Arcs without current limitation cause the flyback to operate in continuous mode, what causes huge currents and very large kickback pulses in the driver transistor. Without adequate protection it will be unavoidably destroyed.
Try something like this:

The transistor with the collector at pin 5 protects the driver against excessive current. C3 and D1 (most mosfets already have it) protect against excessive voltage and recycle unused energy.
And always connect the circuit ground to ground, or high voltage appears everywhere.

Arcstarter

Sun Apr 12 2009, 06:10PM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Arcs without current limitation cause the flyback to operate in continuous mode, what causes huge currents and very large kickback pulses in the driver transistor.

He is using a homewound primary on the opposite leg of the core in relation to the secondary. This provides leakage inductance which *should* limit to something low enough.

I suspect your gate resistor is way to high, it should be around 5.1 ohms in your application.

Well, extra resistance will do nothing really other than some extra dead-time. This is a single mosfet, so there is no shoot-through. It might lower the gate voltage some though, but only a tiny bit if anything.

What FET where you using? They normally have a 20V gate, and they will take much more, so I doubt your gate voltage was an issue.

This is a fact... So i don't understand what else could have done it. The gate must have shorted. What i think is the tesla coil caused some crazy oscillations in the gate, which where high frequency. This could have made the gate die from overheating, or other crazy stuff.

And always connect the circuit ground to ground, or high voltage appears everywhere.

This would be a great idea XD. Just ground the common negative of your circuit to the mains ground. As i have been saying, the ground you made will already have a bunch of RF interference and junk in it. Just use a separate ground, AKA the mains earth ground. Also to this ground hook the shield.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob