Ignition Coil problems

[DaveC]

I have a circuit based on a 556 chip to produce a 100hz square wave pwm 40%
on. But the Mosfet, a MTP3055E, keeps burning it self out when driving the
coil. I think the coil is quite standard. Here is what I'm doing

Car batery + is connected to the coil +. coil - is connected to the Mosfet
Drain. The Mosfet Source is connected to the car batery -. A bolt sits in
the secondary socket of the coil.. A stick with a wire is held close to the
bolt and this wire is connected to the batery -.

The circuit it self is seperately powered by a 12v dc wall wart.. 556 timer
generates the pulse that drives a TC4420 (A Microchip, mosfet driver) that
in turn drives the Gate of the Mosfet.

I can just hear the tone when I connect a little speaker up in place of the
coil, but when I add the coil the Mosfet burns out, very quickly.

Any help, please.

DaveC

 

[gagir]

Check your Mosfet rating against the coil peak current.

Also some coils need a ballast resistor (in series with the coil primary)
not to overheat.

gagir

 

[Jim Yanik]

I have a circuit based on a 556 chip to produce a 100hz square wave
pwm 40% on. But the Mosfet, a MTP3055E, keeps burning it self out when
driving the coil. I think the coil is quite standard. Here is what I'm
doing

Car batery + is connected to the coil +. coil - is connected to the
Mosfet Drain. The Mosfet Source is connected to the car batery -. A
bolt sits in the secondary socket of the coil.. A stick with a wire is
held close to the bolt and this wire is connected to the batery -.

The circuit it self is seperately powered by a 12v dc wall wart.. 556
timer generates the pulse that drives a TC4420 (A Microchip, mosfet
driver) that in turn drives the Gate of the Mosfet.

I can just hear the tone when I connect a little speaker up in place
of the coil, but when I add the coil the Mosfet burns out, very
quickly.

Any help, please.

DaveC

Inductive spike is killing the MOSFET (after the FET turns off).When the
FET turns off the magnetic field collapses and creates a very large reverse
voltage on the primary,breaking down the MOSFET.You probably need a bypass
diode or snubber circuit.

 

[Kevin McMurtrie]

I have a circuit based on a 556 chip to produce a 100hz square wave pwm 40%
on. But the Mosfet, a MTP3055E, keeps burning it self out when driving the
coil. I think the coil is quite standard. Here is what I'm doing

Car batery + is connected to the coil +. coil - is connected to the Mosfet
Drain. The Mosfet Source is connected to the car batery -. A bolt sits in
the secondary socket of the coil.. A stick with a wire is held close to the
bolt and this wire is connected to the batery -.

The circuit it self is seperately powered by a 12v dc wall wart.. 556 timer
generates the pulse that drives a TC4420 (A Microchip, mosfet driver) that
in turn drives the Gate of the Mosfet.

I can just hear the tone when I connect a little speaker up in place of the
coil, but when I add the coil the Mosfet burns out, very quickly.

Any help, please.

DaveC

The MTP3055E is rated for 60V. That's 1/10 of what the voltage rating
needs to be.

An even better idea might be to completely abandon fast switching. An
ignition coil has no use for high frequencies. Connect the MOSFET
directly to the timer chip. Add a fast 1A diode between the gate and
source, with reverse polarity, to block the current surge that comes
through the gate-drain capacitance when the coil voltage falls. Now
that you're switching slower you can put a small HV capacitor across the
source and drain. This will absorb some HV power from the coil so your
MOSFET lasts longer.

 

[[email protected]]

I have a circuit based on a 556 chip to produce a 100hz square wave pwm 40%
on. But the Mosfet, a MTP3055E, keeps burning it self out when driving the
coil. I think the coil is quite standard. Here is what I'm doing

DaveC

Maybe the capacitor ("condenser") is missing.

If so, when the MOFET "opens", the primary voltage will be excessive
as the coil tries to maintain the current (in vain) but if the
capacitor is present then that need will be satisfied, as the energy
swirls to and fro between primary winding and capacior and the primary
voltage will be "controlled".


Robin

 

[Stepan Novotill]

Use a high voltage mosfet out of a (computer?) switching power supply.
s

 

[Kevin McMurtrie]

Thanks for your feed back every one.. I did some maths, it sounds crazzy
but true.. The primary is 3.5 ohms at 12v = 3.4 amps just to drive the
coil.

I added a power resistor to limit the current to the coil and am able to
run it off a 500ma wall wart.. of course the energy on the secondry is smal
l but at least I can run it for an hour or so no problems.

Thanks again

DaveC

You still need a MUCH higher voltage MOSFET. See:
http://www.pixelmemory.us/Photos/Work/Women.com/JarJar-Q/Page1.html

Th IRF740 MOSFETs have been replaced with NTE2385, which is rated for a
minimum of 500V and avalanche tolerant. It's now even stronger.

 

[Stepan Novotill]

this is a very limiting approach.

you want to cut off the drive before the core hits saturation.
increase your drive frequency, and make the pulse narrower.
ideally, scope it and adjust the width and frequency so that it keeps
ringing, but in the main, cut off the pulse quicker, so the inductance is
still charging. if you go past that point, then you're just making it hot.

a current mode switcher control chip would work well also.

Of couse if you let your pulse width to get too long, then the flyback
will exceed the drain voltage limit of your FET and it will let some
smoke out. I think you still need to canibalize an old PC power supply
for a better FET than the woosie 60 volt one. The computer shops give
used ones away.

 

[Eric Y. Chang]

Stepan Novotill ([email protected]) wrote:
....
: smoke out. I think you still need to canibalize an old PC power supply
: for a better FET than the woosie 60 volt one. The computer shops give
: used ones away.

Yes, I got a burned out supply. It contained two power bipolar transistors
in a half bridge configuration, three dual diodes in the secondary, and
an SCR for a crowbar. Oh yeah, it did have a MOSFET. A small 4N50 for
the pass regulator for startup. This is a bit too small for salvaging
for power applications. It is better to try finding one of these in
a surplus shop unless you like to trace out junk PC power supplies and
maybe find a decent MOSFET. Oh, and by the way, this PC power supply
was marked "bad". Guess what was bad...it was the pass regulator for
startup. Maybe it was cooked by an unrelated fault, or maybe it died
due to a shorted MOSFET pass transistor. Anyway, it is probably better
to find your switching power MOSFET's elsewhere if you can.