Tests of CDI Ignition Modules

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
It's been awhile since I've posted here or even tried to keep up with what has been posted. But here is another thing that has been left out on this research but, it's stuff I have researched already myself and would like to share.

One thing that happens in a combustion engine when the sparkplug fires (creates an arc) is that it turns a stoichiometric air/fuel mixture at 14.7:1 into carbon which helps to short out the spark and it is a near dead short. It is very hard to ignite a lean mixture or a rich mixture but, for opposite reasons which I'll explain in a bit. Don is doing exactly what I did years ago but, I did it for automotive street engines and racing engines. I measured E & I with an open sparkplug gap with a current sense resistor on the lead and I also measured E & I at across a current sense resistor replacing the plug gap. But things didn't seem to add up. It wasn't until I started measuring E & I on my race engine under load (transbrake) that things stared to make sense because I totally forgot about burning a fuel makes carbon which conducts electricity. Playing around with the jetting on the carburetor from 11.5:1 and 18:1 fuel ratio (O2 Sensor in the header collector) this is what I found.

At 18:1 I was getting lean misfires with the plugs showing a whiteish powder. At 11.5:1 I was getting rich misfires with the plugs looking black with carbon but, they were also wet. So being lean meant there was not enough carbon being created and rich meant that the fuel which is nonconductive diluted the carbon. On the lean mix voltages were getting as high as 55Kv and the rich mix was down to around 24Kv. Because I was using my old AC/Delco engine analyzer I didn't/couldn't measure ignition current on the car. It is my belief that as the fuel is being converted into carbon that the carbon actually helps to create something like a carbon arc light. If you haven't seen one they have a very intense light and burn hot as hell.

Now to replicate an actual spark/cylinder combustion on a workbench that one can see to me would be very dangerous, I won't do it unless it's in an engine. But I'm trying to make a test rig using the gas discharge devices I used in the military that I mentioned. I know that when the voltage on them reaches their ionization level they are almost a dead short just like in a actual combustion chamber. I'm going to put a bunch of them together in series, 7,500 volt ones to start with and neon lamps tapped off them and add a current sense resistor. If it works the way I hope, cascading down the row, the last neon light to blink is the voltage output. I'll report back once it's built and tested, in about a month.

Interesting observations. A lot going on during the 1ms or so duration of a spark. I don't have what it takes to understand the chemistry going on in a plasma. I would like to hear whatever you learn from your tests.
Interesting observations. A lot going on during the 1ms or so duration of a spark. I don't have what it takes to understand the chemistry going on in a plasma. I would like to hear whatever you learn from your tests.
Well Don I've been working on engine combustion now for almost 40 years off and on, so there is a lot I have learned and even more I'd like to try. The biggest thing I have found is that one ignition I made and trying I'm trying to upgrade with newer electronic devices combined the best features of a CDI and purely inductive ignitions. The ignition I made would not be useful here on model engines because when charging the coil current would hit 48 amps. This ignition put out a @1msec spark event (msd). By event I mean it would multi spark each time between 3 to 6 sparks where each individual spark was stronger than the previous one. The first spark would be around 50Kv and the last could be as high as 120Kv and the current across a 1 ohm resistor as high as 1 amp. My biggest problem was containing this spark event in the plug wires and distributor cap. I had to limit the power output to @75% of max power.

Seat-of-the pants stuff:
When I did have it on my race car we found that we could drop the ignition timing from 36 degrees Adv. down to 28 degrees and still have the same power output which made it easier on the bearings. This indicated that the deflagration event (combustion process) was a lot shorter. We also found that the exhaust emissions were way down. Instead of your eyes and throat burning from the exhaust it was down to not bad with almost no unburnt fuel coming out (radical cam timing).

I guess the biggest Seat-of-the pants stuff was when I loaned my brother-in-law my pickup truck to use for a weekend. My truck was a 76 Ford F100 camper special with a 1968 302 that was in dire need of a rebuild. Because the valve guides and seals were shot, it would smoke quite a bit in the morning when first started. When my brother-in-law brought the truck back he said to me "you got your ignition on the truck don't you", I said yes why do you ask? He said because when he started it up in the morning it wasn't smoking. Proof that the combustion process was very efficient to the point it was burning the excess oil.

One of the bad things I learned is that with such a strong spark event it will light a lean mixture without miss-fires. This may sound great but the pistons and valves wouldn't like it, high combustion temperatures. One of the good things is that if the engine went rich it would still burn the mixture and not having fuel wash down the cylinders. This made tuning the engine easier also when the air changed.

I've only been working on small ignitions for about the last 10 years and not as much as I would like to. I know from my R/C experience that everyone (the gas engine people) wants a battery CDI ignition because they are easier to start. They also want the smallest and lightest ignition with or without a battery. But if one is going to use a battery I still think a low voltage TCI ignition is the way to go, with a good ignition coil of coarse.

If I were doing your tests I would use just one CDI and one inductive coil for all the ignition box tests and just one ignition box for all the coil tests. Use a CDI coil for CDI boxes and standard coils for the inductive ignitions. Performance wise I use to search for ignition coils with the lowest primary resistance. I found that coils with the lowest primary resistance and that still had high inductance made the best ignition coils. These coils usually had thinner wire and more windings, this also meant they could not handle as much current before over heating or opening up, sometimes bursting. Next thing to look at was step-up ratio, usually done by specs or inductance readings, higher the better. Core material and design also mattered. Multi-plate cores were better than solid. Also do not use any resistance anywhere if one can, it just sucks the life out of the spark. If it is making noise on your radio then use resistance. Also as cylinder or test pressure goes up the weaker the spark, compressed air has more resistance so one needs more output voltage. And without the carbon the lower the current will be. As you have shown, matching up an ignition box with a coil to get the best spark isn't that easy. For the racing engines that I deal with when it comes to an ignition system my first question is "what's the compression ratio?" and the it is cam, RPM, and power adders. No sense in putting a $1,000 mini-arc welder ignition on a 8.5:1 street engine.

There is a ton more we can get into Don but, that's it for now.

I was 'surfing' youtube and this channel came up with this about high voltage coil drivers;

I hope this is of interest to this thread.
I had viewed this some time back and he does a nice presentation of his attempts to get to the large sparks that are shown.

His explained in very good detail with schematics and reasoning why he chose which components he used for each failure or success. What he doesn't tell you on his final version is what supply voltage was used, it is simply shown as Vcc rather than showing previous versions stating 12v. Earlier in his presentation he made mention that is power supply went up to 30v or 2-1/2 times that of 12v which is why some of the components needed upgrading. He's smart, if you want to put on a good show, don't tell the audience what's behind the curtain.

Look at his video starting around the 5 minute mark, his schematic and it shows Vcc is 12-30 volts! Sorry should have looked it up earlier.
Last edited:
I was 'surfing' youtube and this channel came up with this about high voltage coil drivers;

I hope this is of interest to this thread.

Really interesting video. Thanks. A lot of raw meat to digest; more questions than answers. I'm going to try to get back to this when I get some more time. Right now, I'm still trying to get my engine running. I don't think it likes me very much. There seems to be more going on there than just ignition problems.
I want to give everyone a heads up that Dan Williams is now building my CDI to my original specs and quality. I have been working with Dan since the CF Expo and he is now up to speed to provide the CDI's. I will continue working with him to supply parts for the Snow Engine and Lee's radial engines. You can check out his introduction on my main page at https://www.cncengines.com/
With Dan now building CDI's I should have some spare time to work on casting kits and plan sets and stay active in the engine community.

Thanks to everyone for your support,
Welcome to retirement Roy. Well deserved. I'm very glad you were able to find another builder for the CDI's. There would be a HUGE hole in the model community if they were to disappear.
And Thanks Dan for picking it up.
We all appreciate it.

Latest posts