Ignition Coil Won't Spark from Mosfet but DOES spark By Hand? HELP?

[pha3z]

Ignition Coil Won't Spark from Mosfet but DOES spark By Hand? HELP!?

I'm trying to get an ignition coil to create sparks running it with a Mosfet powered by a 555 timer. The Mosfet is definitely switching. I checked it with resistor, ampmeter, and o-scope. Everything is behaving predictably but when I use it to power a car Ignition Coil, I don't get any sparks. I can tap the terminals of the Ignition coil by hand -- Even with it connected up to the mosfet circuit -- and get sparks. But when I try to let the mosfet drive it, there are no sparks. There is definitely current flowing. And if I change the frequency to super low (like a pulse every second), the ampmeter is registering the pulses. If I change the frequency to super high (like 100khz), I get averaged current flow on the ampmeter. But I don't get sparks! I also switched out mosfet with other kinds (different ratings). Didn't fix it.

Anyone else ever run into this problem?

Here's a schematic similar to my setup:
http://www3.telus.net/chemelec/Projects/Car-Coil/Car-Coil-1.png

Hope someone else has had this problem before and can help!

Thanks,
Jim

 

[jim hardy]

you said SIMILAR.. how close are you?

double check polarity on those 1N4148's
either or both reversed would mess you up.

And make sure the 1uf tantalum is there, in fact it could stand to be be larger. Try paralleling a substantial (100uf) aluminum electrolytic there - but keep the tantalum too.

i think max frequency for this will be somewhat less than 100 hz because it takes a while to build up current in your coil to a high enough level to make a spark.
A TV flyback transformer might give you more frequency range.

So get it working at 1 hz first..

 

[sophiecentaur]

To get sparks from 'electronic ignition' you need to have a fat capacitor with something like 400V on it and to discharge it through your coil - lots of current. If you look with an oscilloscope you may find you get some voltage magnification - but just not enough 'oomph' for a spark.

 

[jim hardy]

also - try adding a "condenser" (capacitor ~0.2 uf 600 volts) from coil's DIST terminal to common.

And - with what are you powering this? Hopefully a stout supply like a car battery.

 

[likephysics]

Is the Mosfer getting fully turned ON when you switch?
How fast is the Mosfet drain turning OFF.

I think you can get rid of the 1N4005 diode across the mosfet. The Mosfet body diode is sufficient to protect it.

This circuit is similar to flyback dcdc converter. May be a snubber ckt across the primary coil will help?

 

[berkeman]

I'm trying to get an ignition coil to create sparks running it with a Mosfet powered by a 555 timer. The Mosfet is definitely switching. I checked it with resistor, ampmeter, and o-scope. Everything is behaving predictably but when I use it to power a car Ignition Coil, I don't get any sparks. I can tap the terminals of the Ignition coil by hand -- Even with it connected up to the mosfet circuit -- and get sparks. But when I try to let the mosfet drive it, there are no sparks. There is definitely current flowing. And if I change the frequency to super low (like a pulse every second), the ampmeter is registering the pulses. If I change the frequency to super high (like 100khz), I get averaged current flow on the ampmeter. But I don't get sparks! I also switched out mosfet with other kinds (different ratings). Didn't fix it.

Anyone else ever run into this problem?

Here's a schematic similar to my setup:
http://www3.telus.net/chemelec/Projects/Car-Coil/Car-Coil-1.png

Hope someone else has had this problem before and can help!

Thanks,
Jim

What are you wanting to make sparks for?

 

[sophiecentaur]

How can you expect sparks when the transformer needs 400V and you only have 12V? You either need to use a capacitor discharge circuit with an inverter or a convention contact breaker and resonating capacitor as in old motor engines. Look up "car ignition systems".

[Edit: that's not quite right about "only 12V" but I am sure you need a capacitor (I seem to remember 0.01uF was right for resonance). They used to say it was to protect the contact points but it produces a much higher voltage spike than without, when used with a contact breaker. The Mosfet is only there to replace the conventional contact breaker. I do remember some DIY electronic systems which people made in the 60s, using a power transistor (2n3055? in aTO3 can) as the switch but they were never as good as the capacitor discharge systems.
See the wiki article on car ignition systems - they have a 0.47uF capacitor where I suggested.]

 

[Mike_In_Plano]

The basic Idea looks sound. The IRF640 has an avalanche voltage of 200 volts, which I suspect would get you a spark out the secondary.

I'm pretty sure that you can get away without the 1N4005's as the Mosfet will clamp on the negative going excursion in any case.

I propose you try the following to try to troubleshoot:

Disconnect the 555 and it's related stuff leaving the MOSFET and 1N4005 rectifiers.
Jmper across the MOSFET's gate and source, but otherwise leave it in circuit.
Now, tap-tap-tap a jumper between the source and drain of the MOSFET.
Does this cause sparks?

PS - I did this about 30 years ago with an audio output transistor and blocking oscillator. Made a mess of whining sparks, so it can be done :) Keep on it!

 

[sophiecentaur]

With no 'Condenser'? A new one nearly always sorted out poor sparks in car engines. The coil is supposed to resonate, isn't it?

 

[Mike_In_Plano]

Though I've heard arguments to the contrary, I believe the principle purpose of the capacitor is to act as a snubber, decreasing the dv/dt across the points and allowing the arc to break. Thus, extending the life of the points. An added benefit is that the peak voltage is rounded and the arc life extended.

This is not so important when starting out on a project like this. When I first did it, I was totally ignorant of the effects (about 12 years old). If you want to tweak the system for good performance, you can find the resonance of the primary with a .022uF cap or so (200 Volt film) and adjust the off time of the MOSFET for a little more than a 1/2 resonant cycle. This will give you a half sine wave when it switches off
Then again, the whole coil is suffering from leakage capacitance on the secondary which is reflected back to the primary to the square of the turns ratio. Messy.

 

[sophiecentaur]

"Snubber"? I'm not sure of this term.
When the contacts/ transistor turns off you should get many cycles of ringing if there is a parallel C. If the solid state switch is not open for both polarities of cycle then a lot of energy is wasted. Just one cycle at a high natural frequency (with no C) would probably involve more energy loss in the core. Your core of 30 years ago could perhaps have been different from this one.

 

[jim hardy]

Automotive analyzers show to expect spark duration of about 1 millisecond

i too am suspect that the coil's magbetic energy is going into the mosfet body diode and core eddy currents instead of spark.

Try an automotive condenser from DIST to COM.

If that doesn't help, well, we know one more thing it's not.