Transistor controlled ignition coil driver design-need suggestion pls.

[Chriss]

Hi!
A friend of mine asked me to make some sort of transistor controlled ignition coil driver for his moped.
It is a 150ccm engine with one ignition coil and one spark plug but with some sort of mechanical triggered
ignition coil mechanics.

Here is a simple design I come up with, theoretically it should work but I'm not sure practically does it will last and do the job.

A short description of the circuit:

The whole system is working on a U_max=14.7V = battery charging voltage.
S1 is the mechanical circuit breaker which should for now trigger the Q2 transistor and no more the ignition coil.
The LED1 represents the ignition coil in the circuit simulator software for now and it will be replaced with the ign. coil and the R2 will be also removed.
The C1, R3 and D1A are for protecting the Q2 from HV spikes which would maybe produced by the ignition coil when it
is de-energised.

It is actually a PNP transistor in switching mode circuit, but actually that is what we search for actually.
What you think about the parts and the circuit?

Thank you for any suggestion.

[Dundarave]

As I recall from my motorcycle repair days ('73 Triumph Trident, among others) the cam actuated "points" were normally closed, opening only at the exact time the spark is called for.  This meant that the coil was actually energised most of the time, through what was known as a "ballast resistor" to limit the current to something reasonable.

There was a small capacitor across the points to reduce the sparking.  The coil's magnetic field collapse was what triggered the spark.

You might want to check the circuit schematic of the moped to see if they still do it that way, as you might need to modify your circuit accordingly.

Assuming the moped uses the old points/coil model that I speak of, I don't understand what exactly you would be changing/eliminating with your circuit.  Back in the day, electronic ignitions were used to eliminate the points entirely, and instead use a circuit triggered by a Hall sensor and magnet mounted on the timing cam.  From what I can tell, if you still use the mechanical contacts, you've not really gained anything.

But perhaps basic moped ignitions have changed drastically in 40 years, in which case, what do I know, lol.

[joeqsmith]

How are you adjusting the timing?  Mechanical?

[duak]

i am assuming the ignition circuit on a moped is like that of larger engines.  Some values might be different because there is only one cylinder.  The peak voltage on the primary winding of the coil is quite high, 100 to 200 V, and is very important as it allows the secondary winding to generate the high voltage needed to ignite the mixture.  You will not be able to use a clamp diode to protect the transistor and so require a transistor with a BVCEO of 300 to 400 V.  You will also need a capacitor with a value of 100 to 220 nF in parallel with the transistor that will resonate with the inductance of the coil to generate the correct shape of the primary waveform.

There is a description and waveforms of a simple ignition circuit in this note: https://www.st.com/content/ccc/resource/technical/document/application_note/8d/46/36/a7/b6/f2/45/10/CD00003911.pdf/files/CD00003911.pdf/jcr:content/translations/en.CD00003911.pdf  (I hope this link works - it seems to have repeated info in it)

[wilfred]

I remembered this circuit in a magazine project. This link is to a PDF for the instruction from a kit supplier.  Maybe it will give you some info/ideas. It's 7 pages.

http://www.radiomanual.info/schemi/ACC_powersupply/Dick_Smith_K-3301_user.pdf

[Chriss]

Dundarave:

...if you still use the mechanical contacts, you've not really gained anything.

I'm absolutely with you, I think you are right, but I don't really know why my friend wish to substitute the original design with some of these circuit.
I have to talk to him today and ask him and I will check the moped circuit too.


joeqsmith:

How are you adjusting the timing?  Mechanical?

Yes, mechanically I think, the same way as without the circuit.
I think the breaker can be adjusted with some bolt or so, I 'm not familiar with that for now.
Have to check on the moped.

duak:

You will not be able to use a clamp diode to protect the transistor

I'm not sure why should a schottky diode of 20A and 100V not protect the transistor?
The transistor I would use is a power transistor TIP36C type with a VCEO=100V, total power dissipation of 125W.
But I can switch to an IGBT maybe, but I saw several circuit on the net with a 2N3055 transistors and so...
The link you shared is really valuable to me because I never done such of task like this circuit.
Thank you very much for this.

wilfred:
Thank you very much for this pdf, I really appreciate the content.

Thanks guys for such of nice infos/suggestions.
Today I will go to my friend and see how is the moped constructed and why he actually wish to make such of circuit.

Thank you.
My best regards.

[magic]

Your schematic is upside down Please draw high voltage up, ground down in the future.

I think the only advantage to be gained from using such device is elimination of sparking and reduced wear on the mechanical switch, are those things a problem for that guy? Perhaps a slightly stronger spark by elimination of the snubbing capacitor, but I'm not sure.

As for your design, it applies (almost) full coil voltage to the switch through the BE junction of the tranny, albeit somewhat current limited by the resistor. Not sure how it will behave sparking-wise with a real coil instead of an LED and whether you gain anything.
As it stands, it frankly isn't going to work at all, because it needs to get to a point where several amps flow though the tranny and its output voltage is barely above zero. Go calculate how much base current is needed for that and whether a 10k resistor will do.

No, really, switch it to an NPN and add another inverter stage in front of it. Its output will reach closer to zero, less heat will be generated and base current will be independent of coil and ballast resistor voltage drop.
And actually, N MOSFET might be better still. Even less voltage drop if you get a good one, no heatsink needed, doesn't need overvoltage protection if you get an avalanche rated one.
I've actually built a spark generator using automotive ignition coil once and used the IRFP4768 to drive it. A bit overkill in retrospect, a cheaper part will probably suffice with slightly more heat generation.

Power dissipation can be calculated, just assume that the tranny is conducting full coil current continously and do the math with RDS(on).

A few nF capacitor may be needed from source to drain to avoid exceeding dV/dt rating of the transistor when it turns off. I calculated that 1nF sufficed for my use, IIRC.

I've heard mechanical contacts can oxidize when switched without load. So for sensing if the switch is closed or open, pass some generous current through it. Maybe 10mA at least? I'm not sure, perhaps someone will know.

[mikerj]

As I recall from my motorcycle repair days ('73 Triumph Trident, among others) the cam actuated "points" were normally closed, opening only at the exact time the spark is called for.  This meant that the coil was actually energised most of the time, through what was known as a "ballast resistor" to limit the current to something reasonable.

There was a small capacitor across the points to reduce the sparking.  The coil's magnetic field collapse was what triggered the spark.

Many small engines (two and four stroke) from that era used a slightly different scheme.  Rather than have the coil energised via the battery through the points, there was a separate winding in the generator that was connected to the ignition coil, and the points shunted the output of the winding until they opened.  This means they can run without any battery connected.

[xavier60]

If it is originally magneto ignition, the cam's dwell angle is likely to be too much for battery powered ignition. If the original coil is used, it will likely need a ballast resistor also.
If I had a need to do something like this, I would time the turn on so that the current ramps to the correct current at the time the spark is required.

[Benta]

If it's a magneto ignition, there's no reason for doing this.
If it's a points/coil ignition it will bring your friend no joy.
The points need a certain current, usually in the 1...2 A range, otherwise they won't be self cleaning and foul quickly, leaving your friend with a non-functioning ignition.

[LukeW]

There are IGBTs such as the ISL9V3040 that are specifically designed to stand up to the task.

[soldar]

From what I can tell, if you still use the mechanical contacts, you've not really gained anything.

You're missing the points

The entire point of a transistor-assisted circuit was that not all the current of the coil went through the points. That high current and the spark caused by the inductance is what wore and burnt the points. Once that current was being handled by the transistor the points could last virtually forever.

[vk6zgo]

From what I can tell, if you still use the mechanical contacts, you've not really gained anything.

You're missing the points

The entire point of a transistor-assisted circuit was that not all the current of the coil went through the points. That high current and the spark caused by the inductance is what wore and burnt the points. Once that current was being handled by the transistor the points could last virtually forever.

Exactly!
Transistor assisted ignitions which retained the points were much more the rule than the exception, back in the '60s & '70s.
Just about every Elecronics magazine produced a number of such designs in those years.
They did give a better spark under test conditions, & the mags usually gave figures pointing this out.

In practice, though, those improvements didn't flow through to improved performance, so the long points life was the major benefit, & motive for builders.

Until factory made magnetic or optical replacements for the points became available, either intended for "homebrew" projects, or bought as parts from later vehicles, only the most dedicated experimenters went down the path of points elimination.

"Electronics Australia" did design & build an optical one, & some people built them, including my old Shift Boss, who was an inveterate tinkerer.