Ignition Timing Circuit for 80cc 2 stroke motor

Thread Starter

solmillin

Joined Aug 18, 2016
22
I've been trying to build a circuit based on a timing wheel with a hole that switches on a photo resistor by a LED shining through the timing hole and then producing a high tension spark through an ignition coil.

I've got two parts working.

1. The comparator LM393P that gives correct ON/OFF signal dependent on whether the timing hole is sensed.

2. Spark driven by a N3055 with safety diode across collector/emitter and dissapative snubber across collector and positive rail. The N3055 circuit works with a 4.7K pull up resistor on the base. Strangely enough it oscillates with a continuing spark which is not a bad thing as long as it will switch OFF reliably !!

The 'theory' is that when the LM393P is ON it has an 'open collector' which if connected to the base of the N3055 should earth it and switch off the N3055 and hence the spark.

The problem is when I connect the LM393P output to the N3055 base the two circuits become unstable.

Can anyone help with a circuit to connect the two parts correctly?

Perhaps I should try a simple npn low current transistor between the LM393P out with the transistor's collector connected to the base of the N3055. Is this the next step?
 

#12

Joined Nov 30, 2010
18,224
The next step is to post a drawing of the circuit so we don't have to guess which mistakes you don't know are mistakes.
 

#12

Joined Nov 30, 2010
18,224
I tried 2 sources and can't find the, "N3055"
Can you point to a datasheet?
The 'theory' is that when the LM393P is ON it has an 'open collector' which if connected to the base of the N3055 should earth it and switch off the N3055 and hence the spark.
The way I read this chip, when the positive input is higher than the inverting input, the open collector output does not pass current to ground. Are we on the same page here?
Does the photo-resistor (with no part number) output high or low when it sees the light?
Common LEDs should be run not higher than 20ma. You're using at least 33 ma. What are the characteristics of the LED (which has no part number)?

Now for a guess: You should run a separate power wire to the coil and snubber, starting clear over at the switch, and you should run a separate wire from the bottom of the transistor and coil secondary to the ground point so the brains don't share wires with the brawn. That means, Draw a vertical dotted line to the right of R4 and wire everything to the right separately. If that doesn't fix it, we'll work on the type of transistor and how to calm its input connection.

Maybe somebody else can help with my question: How does a 10k resistor form any kind of effective snubber compared to a coil which uses an amp?
 

#12

Joined Nov 30, 2010
18,224
You used the symbol for an n-channel mosfet to represent this 2N3055 NPN transistor. That, along with missing part of the part number gives too many wrong clues.
That might be part of the problem. The 2N3055 would struggle to get 1/4th of an amp going with a 4700 ohm base resistor.
 

crutschow

Joined Mar 14, 2008
34,285
The base current of a BJT (such as the 2N3055) should be no less that 1/10th of the maximum collector current for proper saturation when ON.
 

tcmtech

Joined Nov 4, 2013
2,867
Unfortunately a 2N3055 is a poor choice for a 12 volt ignition coil driver being it only has a ~60 volt rating and tends to have terrible shoot through characteristics with that low of voltage limit that causes considerable inductive spike dampening which drastically reduces your spark energy.

Also you should use a standard ignition condenser across the switching device just as is used in a mechanical points based system. It will greatly improve your spark voltage and intensity.

If it was me I would replace it with either a multi hundred volt 15 amp or higher rated Power Mosfet or IGBT device which of either could likely be driven directly off your photo resistor module assuming it has a clean high speed state switching output in a basic totem pole design.

If not then any old single channel 8 pin Op- Amp could work as the driver interface.
 

Thread Starter

solmillin

Joined Aug 18, 2016
22
Thanks guys, much appreciated. With your replies it would appear the instability is being caused by me using the N3055 ! (which circuit, by the way, I found on the internet).

I bought an IRF540N the other day just for this purpose, so your responses are like me going to see a doctor who has diagnosed the problem ! Woo hoo !

If I replace the N3055 with the IRF540N from what I understand I should put a capacitor across collector and emitter?

May I ask what value? And also, what value pull up resistor to the IRF540N base/gate?

The open collector of the LM393 in the ON state pulls the base/gate OFF and in it's OFF state the output is open circuit I believe, so the Mosfet goes on.

Hope I've got it right. Onwards ever onwards.

Very helpful. Thanks.
 

shortbus

Joined Sep 30, 2009
10,045
why not do yourself a big favor and use either a GM HEI module or Ford Dura-spark module as the coil driver? There are resources online giving the way to use them.
 

tcmtech

Joined Nov 4, 2013
2,867
If I replace the N3055 with the IRF540N from what I understand I should put a capacitor across collector and emitter?

May I ask what value? And also, what value pull up resistor to the IRF540N base/gate?
Depends on what value your particular coil is designed to work with. It could be anywhere from around .5 uF to several uF.
 

#12

Joined Nov 30, 2010
18,224
There are still some specs that would be helpful. The current of the transformer primary would help, as would the inductance of the transformer primary.
I know the worst case gate capacitance of the RF540N is 79 nanocoulombs but I don't know the max RPM of the motor, so time can not be calculated, the limits of R4 can not be calculated, and the capacitor across the mosfet can not be calculated.

If you want rule-of-thumb wisdom, that's what you're getting. If you want hard numbers, you have to give us hard measurements or spec sheet data.
 

tcmtech

Joined Nov 4, 2013
2,867
Electronicly switched ignition coils do not need a condencer like points ign systems.
I've built many and they are greatly improved by it. Just because something works doesn't mean it works as well as it could.

BTW I see on many well proven electronic ignition designs they do in fact include a capacitor across the main switching device and the ones that don't have it there put the capacitor across the ignition coil's primary which works well too.
 
Top