DIY: How to make your own Electronic Ignition for contact point engines

[kinyo]

*Kinyo - Where did you get the CDI layout sir? Does it have the multi-spark feature?

the CDI i built did not have multi-spark feature ... it was a very simple circuit consisting of a bjt (2 x 2N3055 plus transformer) multivibrator circuit to charge the capacitor to almost 400 volts and an scr to discharge onto the primary of ignition coil.

[Chef Noob]

the CDI i built did not have multi-spark feature ... it was a very simple circuit consisting of a bjt (2 x 2N3055 plus transformer) multivibrator circuit to charge the capacitor to almost 400 volts and an scr to discharge onto the primary of ignition coil.

Nice design! How were you able to overcome the hard starting symptom of the CDI?

I always wanted to build one myself. Unfortunately, I'm still looking into the "safety" factor - for myself & the car's sake. Getting an electric shock from a 400v CDI module just doesn't sound that appealing to me.

[Ginnova]

hope you can post where to have the pcb's fabricated.

[kinyo]

Nice design! How were you able to overcome the hard starting symptom of the CDI?

I always wanted to build one myself. Unfortunately, I'm still looking into the "safety" factor - for myself & the car's sake. Getting an electric shock from a 400v CDI module just doesn't sound that appealing to me.

Not sure if i understand the question, because i did not experience "hard starting" with the CDI. It consistently starts the car quickly even on cold starts.

Yes 400V could be a bit shocking, but it is still considered low voltage by electrical standards.

[Chef Noob]

Not sure if i understand the question, because i did not experience "hard starting" with the CDI. It consistently starts the car quickly even on cold starts.

Based on the research I done regarding CDI, it increases the voltage produced by the spark plug but shortens the spark duration considerably. This in some way, creates situation for certain engines to have a hard time starting...

Although I havent witnessed it personally.

[duskylim]

Hello Chef Noob:

Your work reminded me of when I opened up and copied a Nippondenso igniter (the electronic ignition of Japanese cars in the 80's) using locally sourced components from Raon.

I too had my share of problems with it, a straight copy of the design was less than satisfactory because at first I didn't insulate it well enough.

The igniter design uses a PNP to NPN overdriven switch triggered by the original contact points where the turn on current (base-to-collector) is multiple of the hfe gain.

Basically you looked up the transistor's specs - particularly the hfe and rated current it could switch, and multiplied the hfe by a factor of 5 or so.

When you do this, the transistor ceases to behave as a linear amplifier and at that point and simply becomes a high-speed switch - like the contact points themselves.

Although the design gives up a lot of gain (amplification) the excess turn-on current makes a bipolar junction transistor's internal resistance drop dramatically - typically to 0.5 to 0.1 ohm - which is the key benefit of providing excess turn-on current (aka overdrive)!

This means that when switched on and carrying the full load of the coil primary - the transistor itself has the least resistance of all the ignition circuit's components.

In this design the stock contact points only had to switch 1/150th of the original current (from carrying 10 amps to just 0.067 amps) so that they no longer pitted, only very slowly (over a year or more!) got light grey oxidation.

A smaller PNP transistor was switched by the coil and it's output drove the larger, main NPN transistor.

The 1st problem was that if you didn't insulate the main NPN power transistor (2SD458) well enough, during the switching, the coil's primary winding voltage would temporarily go to 400-600 volts at the NPN transistor's collector and this would produce shorts into the aluminum heat sink frame.

You had to use mica insulation, nylon bushings an lots of silicone grease to insulate it properly.

Still even after all that there were still problems.

Basically the main failure was due to punch-through - the transistor would break down from the extremely high collector-to-emitter voltage.

The original Nippondenso design addressed this via a voltage divider (resistors) parallel the collector-to-base-to-emitter circuit, limiting the speed at which the transistor turned on and thus reducing the induced voltage at the coil primary - reducing the tendency of the voltage to punch through.

This also reduced both the voltage and power output to the spark - the slower switching rise meant lower induced primary voltage - hence lower induced secondary voltage - and lower energy transfer from coil primary to coil secondary to boot!

After loosing several igniters to this punch-through problem, rather than copy the original Nippondenso solution (i.e. - surrender), I decided to mimic the solution of the original ignition circuit.

Here's what I did:

In the original contact point circuit the points themselves to switch the coil primary.

However the points are not the only component in that circuit there is one more thing there that if you remove will prevent the ignition circuit from working correctly.

That's the capacitor which is wired in parallel to the points.

When the points close, current flows across the coil primary, and builds up a magnetic field, storing energy that becomes the spark in the coil secondary.

When the points open, the coil produces an induced current that tries to keep flowing - the faster the points open the higher the voltage of the induced current.

If there were only the points in the coil primary circuit - the induced voltage would jump the gap and spark and burn the points.

The capacitor is there to stop that from happening.

In fact, the 1st ignition circuits (before 1920s?) put the contact points INSIDE the combustion chamber - using that induced spark to create ignition, the ignition coils of that era had no secondaries only primaries!

When the points open the capacitor stores the coil's induced voltage spike instead of allowing it to be borne completely by contact points.

This allows them to separate without too much sparking and allow the transfer of energy from coil primary to secondary.

Using this as my model I computed (very roughly) the energy stored in the coil primary and added a capacitor in parallel to the NPN circuit - effectively storing the energy and preventing punch-through.

Its been a long time since I did that.... thanks for the memories.

Best Regards,

Dusky Lim

[Chef Noob]

Thank you for the very well explained theories. I can tell by your post that you also enjoyed the time you spent doing this.

I was amazed that you were able to create primary voltage up to 600v using only transistors. Even up to now, only CDI modules are able to produce those levels of voltage. Though it may prove very beneficial for the power gains in the engine, considering the cost & wear on your parts(coil, plugs, etc.) wont be cost efficient. Even if you were able to extend the life of your power transistor, I think the plugs were not meant to handle those voltage & currents. I myself also seen a few designs where they attached the capacitor in parallel to the transistor.


Honestly, the main reason we are discussing here right now is because of this capacitor aka "condensor".
In a perfect world, if it were able to work perfectly...I think most production cars right now will still be using the points system.
Theoretically, the capacitors are able to absorb the "back EMF" created by the induced current.
In the standard points system, when the points open, the capacitor absorbs the extra spike of current thus eliminating the dreaded "gap spark". However, that power is still stored within the capacitor. The problem lies from the closing of the points. Since the charge from the capacitor have no where to go, it "grounds" itself as the points closes, creating a "mini" spark as this happens. This spark may not be that much but after doing this for atleast 1000 times a minute, that tiny spark will do damage to the point surface. (based on the laws of electrolysis)
That is why in the real world, you can still see burnt points even though the capacitor still works perfectly.

The option I'm dwelling on right now is to induce a stronger spark...not by reducing the resistance of the transistor to induce more current...but to increase the output on the coil - at the same time prevent it from over heating.

As you stated in your post above, transistors are more fragile than we think. It can easily fry when subjected to extreme currents.
However, if you increase the current on the ignition coil, you can still attain good spark from it. One problem is: What if the coil overheats? What I'm thinking of is putting a cooling system to reduce the temp of the coil. A finned heatsink perhaps? Or even water-cooled? :-)

[Chef Noob]

*Ginnova I was trying to PM you sir but it says that your Inbox is full.

[duskylim]

Dear Chef Noob:

Actually I couldn't directly measure the peak primary voltage, (I don't have an oscilloscope with a high-voltage probe) but when the igniter failed, I would always check to find the cause of failure.

Consistently, I found that the main NPN power transistor had shorted out at the collector-to-emitter junction - classic failure by punch-through.

Now according to the Motorola transistor manual, the 2SD458 has a breakdown voltage of 600 volts (DC) at the collector-to-emitter junction, hence my assumption that the voltage was peaking past the transistor's limit, burning through and shorting it out.

That voltage peak is of course a transient peak voltage, created by the rapid switching of the transistor itself, while CDI ignitions can create sustained peaks of 600 volts or more - it is the duration that the voltage is sustained is the key difference between the 2 types of ignition systems, transistorized and capacitive.

Remember this equation?

V = (-) L*di/dt

<the induced voltage V is of opposite polarity and equal to the coil primary inductance L multiplied by the time rate of change di/dt of the switching current>

To reduce the heating of the coil (which is actually worse during low-rpms rather than high-rpms) you keep the ceramic ballast resistor in place - don't by-pass it!

At low-rpms the turn-on time is much longer hence there is more time and current for the coil to overheat, at high-rpms the short turn-on times effectively reduce the total current and the coil cools down rather than heats up.

The ballast resistor's resistance increases with increasing temperature, reducing the current to the coil, thus keeping it cooler at low rpms, but that resistance decreases with decreasing current at higher rpms, allowing more current to flow - it effectively behave like a current control (a little slow and crude - true) for the coil and ignition circuit.

The best choice I found was to get (2nd hand) very large coils that came with some high performance Japanese engines.

The way to test for a high power, high performance ignition coil is:

1) measure the coil primary resistance - a high power coil has very LOW primary resistance and thus will pass a lot of current and store a lot of energy

2) measure the primary-to-secondary resistance - a high power coil will have very HIGH primary-to-secondary resistance allowing it to withstand very high induced voltages

The Bosch red coil is a good example, but the huge 2nd hand Japanese coils we used back then had the lowest primary resistance I have ever measured - only 1 to 2 ohms!

We used them with ballast resistors taken from the Bosch red coils.

You are quite correct in describing how the capacitor works in a conventional points system.

We used to go a little further than this in modifying them.

We would look at the points and determine which face (moving or static) lost material and which face gained material - alin yung may uka at alin yung may bulutong.

Then you buy a capacitor with a different micro-farad rating until the transfer stopped.

All that became academic with the use of the transistorized ignitionl - point buring was no longer an issue.

However point bounce and flutter still remained - hence the move to different triggering systems.

As to the issue of coil overheating, well the Bosch red coil does tend to overheat, especially if you remove the ballast resistor - which will eventually burn it out.

The large Japanese 2nd-hand coils however, didn't overheat - they lasted as long as we had our cars.... go figure.

If you wish to cool your coil down - my vote is for an aluminum heat sink - keep water and fluids well away from your ignition system.

The ultimate in ignition systems is of course the magneto - its the only system where the spark power INCREASES with rpms. he he he.

Best Regards,

Dusky Lim

[Chef Noob]

The ballast resistor's resistance increases with increasing temperature, reducing the current to the coil, thus keeping it cooler at low rpms, but that resistance decreases with decreasing current at higher rpms, allowing more current to flow - it effectively behave like a current control (a little slow and crude - true) for the coil and ignition circuit.

Dusky Lim

I thought the ballast resistor installed in our cars are of the fixed type and not the self-variable ones?

I always thought that the coil heating up is the cause of repetitive induced voltage created by the opening of the points...See, we learn new things everyday. :-)