It is not generally understood how many different kinds of ignition systems have been used on powersports (motocycles, ATVs, scooters, personal watercraft, utility vehicles, side-by-sides, etc.) vehicles. Presently there are only two found on today's road going vehicles, and one more on off-road vehicles such as utility vehicles. But altogether there have been about six notable ignition system divisions since say WWII, with a few more splits that could be made within those ranks, variations if you will. Before examining them let's lay some technology ground rules. All ignitions are defined (not just by me but by engineers too) by two things: their source, and their secondary discharge methodology, that is, whether they fire their spark plugs by rising the primary or collapsing it. The importance of the first is obvious. I'll address the importance of the second later. Let's jump right in then, starting with magneto ignition.
When talking about magneto ignition folks tend to think that anything without a battery is necessarily magneto, and this is not so. There are no fewer than three different ignition systems that work without a battery, and magneto is just one. Magneto ignition, as most know, is self-powered. There is no need for a battery. So this takes care of our first all-important ignition characteristic, source. More important however is the second characteristic: magneto ignition is collapsing field. This means that its primary coil is activated before the ignition event, then deactivated, turned off that is, and then the spark plug fires. Why is this so important? Because it, as they say, goes right to the bottom line where what we expect from an ignition system is concerned. Collapsing field ignition systems fire their plugs only after a two-stage process of energizing then deenergizing, and this spells sluggish spark generation. A human whose timeframe is his beating heart and his eye movements finds it hard to comprehend how a difference measured in milliseconds (thousandths of a second) can mean so much, but it does. To the spark plug, the difference from one ignition type to another in the area of trigger to fire time, which is nearly twice as long in collapsing field systems as it is in rising field systems, means everything. The slower buildup of voltage at the spark plug means the electricity gathering there has time to bleed off sideways on carbon deposits, which it inevitably does, resulting in reduced firing snap at the plug. Magneto is somewhat famous for being highly intolerant of "dirty" plugs that many other ignitions have no trouble with. And what did we have on magneto equipped bikes but poorly carbureted, poorly maintained, dirty air filter engines? Now back to the first criteria, power. Magneto has another weakness, and that is its self-powered nature, which in some ways is a an advantage of course, depending. It's a liability though in the sense that it means its power source is a magnet-driven wire coil, which of course can only produce AC. Magneto is AC driven, which means that its source voltage rythymically goes through a zero point. Happen to have that zero point, through misadjustment or lack of maintenance, cooincide with the system's trigger point, and you have the reason folks trucked their magneto ignition equipped bikes as often as they rode them. Though magneto for many years was points operated, versions have reappeared in recent years on stationary engines with transistors instead of points. These are still magneto systems however.
Energy Transfer Ignition
Energy transfer is that system that is often mistaken for magneto, but it is so different as to be in a different category altogether. Yup, you guessed it, ET as it is often called is, unlike magneto, *not* collapsing field, but rising field. When attempting to determine whether your bike has magneto or the more common ET, you can look at the manual schematic if you have it, or at the parts themselves under their covers. What you will find is ET is different from magneto in that it has one more major part, a third coil. Magneto has just two, the primary and the secondary, because the rotating magnets use the primary coil as the source. It does double duty. ET on the other hand has a separate source coil in addition to the primary and secondary. Then there is that second criteria again, collapsing/rising field. Where magneto works on collapse of its constantly built-up primary when the points open, ET rises the field in the prmary at the moment when its points open. Very different systems. Because, remember, the rising and collapsing split is everything in ignition. The holy grail. Rising field systems are automatically better than collapsing field systems because of that time element. An ignition system that works on the rising field principle can get its energy to the spark plug so much faster (faster in its world, remember, not ours, there is no difference in ignition timing relative to TDC) that the voltage has no time to bleed off, and more of it stays put until there is enough to jump the gap, resulting in a much higher tolerance of dirty plugs and more efficiency power wise. That is, rising field systems do more with less. ET is a good system, better in many ways than magneto, but it does share magneto's source type, so those drawbacks related to source are present. For example, because of that whole AC sinewave synch thing, a misadjusted set of points can make ET not even work, just as in magneto.
Capacitor Discharge Ignition
Capacitor discharge ignition (CDI for short) is at the top of the heap, technologically. Invented at a about the same time as ET (ET soon came to be known as mechanical CDI in fact), it shares many characteristics, and one of them the all-important one, it is a rising field system, meaning it fires it plug in just one stage from trigger to spark, just like ET. CDI is different from all others though in its source configuration (remember, source and firing method are the two important metrics). Instead of relying on a feeble, oscillating magnet alone for power, CDI takes the source coil and feeds its output into a capacitor, storing and building it from a nominal 7-10 volts to as much as 300 volts (systems vary, some peak at only 150-200). Then, when this is stored up to its maximum, the system dumps this high voltage into the ignition coil, all at once, and the result is predictable. High input means high energy downstream at the plug, and rising field triggering means only one stage, not the wasteful two stages of magneto. CDI ignition coils work with so much higher triggering voltage that they can be and often are made very small, one fourth the size of conventional coils (scooters often have coils the size of oldschool camera film containers). CDI was actually purpose-developed for snowmobiles. With such dramaticallly faster plug voltage buildup time, it didn't matter what deposits were on those plugs, those babies would fire! Folks got the notion it was CDI's higher plug voltage that was the boon. Howevr, although CDI technically can be higher in some applications, in production vehicles it is only marginally so. It was the ultra fast (40 percent faster) rise time that did the magic, not the voltage. As with many things, this fast rise time, though it blessed snowmobile riders with much more safety and convenience, promised to be a liablity for street bike riders. Here is why. The utra-fast rise time was accompanied by a commensurately quick sign-off. That is, the ignition event started earlier and finished earlier. The spark duration shrunk to about half its usual time. This was of no concern to off-road riders running generously carbureted engines, but on emissions spec road machines having lean carburetion it just didn't work. So for many years after its invention CDI was a no-show on road bikes. Finally, though, an enterprising aftermarket company developed a CDI that repeated its spark four or five times in the space of time that other ignitions fied just once, and CDI suddenly could be prime time, but for its added cost. You have heard of MSD, multiple spark discharge? They were the company. Street bikes really didn't struggle with ignition problems like snowmobiles and other off-road vehicles though, so CDI's cost still kept it out if mainstream use. Honda however popularized CDI on their 400 Hawk and CX500 in 1978. Later, manufacturers such as Honda would equip their sport bikes with a battery-powered version of multi-spark CDI, and still do today, giving the industry the best mass-produced ignition available.
Charles Kettering, of the famous Sloan-Kettering Foundation, developed (but did not invent) the electric starter, was the manager responsible for the developement of tetra ethyl lead and freon (both of which ironically cast him in an ignoble light today, unfortunately), and most famously of all invented battery-point ignition. More correctly known as Kettering ignition, his 1908 invention (the date on the patent sketch, anyway), would serve as the longest successful ignition system in the history of motoring. Honda stopped using Kettering in its 1979-1980 roadbikes, but not because of emissions. Not directly at least. That is, not because Kettering wasn't up to the task of supporting good combustion, but because emissions standards as they were applied to motorcycles for the first time beginning in 1978, included something called an emissions warranty. The 5 year emissions warranty's provision specifies that the ignition must not have to have attention paid it during all that time. And of course, no points based system was up to that task. Kettering ignition is stupendous. More elegant a system while at the same time powerful, it is hard to imagine. Everyone knows how it works. Points closed, the ignition primary winding saturates. Points open, the primary field collapses, inducing the transformer-like secondary winding to build, firing the plug. A great system, easy to understand, easy to maintain, and adequate for all normally used road going engines. Kettering's one great fault is its collapsing design. It is like magneto in that sense. That is, Kettering fires its plug only after going through two stages, first buildup of its primary coil, then collapsing that coil, and as we have already explored, this is a pretty signifcant liability, at least in the spark plug's world. But in most cases, none of us are aware of this, and certainly few were in 1908, the year California became a state, the year of the first passenger carrying airplane, the $825 Ford model T's innaugural year, the year General Motors was founded, and four years before the sinking of the Titanic! Most motorcycles made before 1980 use the Kettering system. Although the battery source is an advantage over magneto and ET's AC source, Kettering is not without some issues, power wise. For one thing, it is a power hog. As engine rpm increase, so does its need for power. On many vehicles, this was countered by the inclusion of a so-called ballast resistor. Honda Goldwings had one, and so did your dad's 57 Chevy. The ballast resistor is supposed to store energy at low rpm to compensate for the shortage that occurs at high rpm. A voltage conserving tactic the technician can use, and one that I have frequently used, is to gap Kettering at the lower end of its gap range, thus maximizing coil dwell time. This works well, too. Kettering's appetite for voltage shows also in Kettering's positive response to attempts to lower the commonly observed 2-3 volt drop between battery and ignition coil on most bikes of the 70s. Another drawback of Kettering is that all the coil's primary energy must pass through the system's points. A system was devised by someone that added to Kettering a transistor layer between the points and the coil, with the result the transistors handled all the switching and the points merely timed the whole shebang. This makes the points last longer, as it reduces the arcing that errodes the points' faces. This is a very good system, and a fella on the SOHC4 forum (Mark Paris) is producing and selling these for their bikes and I believe others as well. The condensor in Kettering serves an interesting job. Many regard the condensor as merely a buffer, an electrical shock absorber, if you will, between the points and ignition coil. It is that, for sure. As the points open, a spark tries to jump the points instead of the spark plug, and the condensor largely prevents this, or at least discourages it, and that's its job, for sure. What many don't realize though is that the prevention of this arc also boosts ignition coil output, by making the points' break, and thus the collapse of the primary, clean and sharp, not to mention sudden, and this in turn is helpful to a sharp and sudden secondary buildup, with the result a brighter spark. So the condensor is a points protector as well as a coil booster. By the way, Kettering equipped bikes will run without the condensor, but not very well. At about 1/4 throttle the point arcing is so severe that the coils can't make enough spark for any more rpm. Sort of a crude rpm limiter, and not incidentally a giveaway symptom for condensor failure (but also a symptom that could be fuel system and other problems, unfortunately). One more thing before we leave Kettering. There are a handful of standard modifications that have been developed to make Kettering work better. Maximizing dwell, already mentioned, is one of the more classic ones. Double-springing is another. Substituting the ballast resistor is another yet. And there are more, such as Mark Paris' transistorizing kit.
Speaking of transistors, 1980 and later bikes lost their points because of the aforementioned federal meddling. What's interesting is transistorized ignition got in return two parts that replaced the points. Two parts were necessary because Kettering's points in fact did two jobs. They timed and they switched. It is a vivid testament to Kettering's elegant simplicity and efficiency that two parts were needed in transistorized ignition to take the place of just one part in Kettering. And make no mistake, other than this one change, there is absolutely no difference between transistorized ignition and Kettering, or at least there wasn't until much later when rev limiters and other electronics were added. Let me say it again. Honda's TPI (transistorized pointless ignition) is identical to Kettering except the points are replaced by pulsers (for timing) and spark units (transistor boxes, the switches). This is significant because it means all of Kettering's characteristics, good and bad, strrengths and weaknesses, are present in TPI. Kettering's energy-hogging nature, its resulting sensitivity to voltage drop, and its collapsing field design, all are the same in transistorized ignition. Even the Kettering ignition coil's tendancy to overheat is the same in TPI. Transistorized ignition offers no gain in engine power, no better starting, none of that. The only advantage is maintenance. And this is true of most of the aftermarket systems too, even Dyna. Yes, that's right. Even the famous Dyna S is nothing but an aftermarket version of Honda's TPI, and thus in reality simply a modified Kettering. Transistorized ignition would remain strictly identical to Kettering until about 1985 when manufacturers began adding things to the spark box, such as electronic advance, rpm limiters, and by the 1990s computer circuitry controlling virtually everything the engine does. But even today, Kawaski's awesome 8.5 second 2013 ZX-14R is powered by basically the same ignition in a Ninja 900, if you don't count the computers doing all the added chores. It's still Kettering in other words, and thus has all the earmarks of its 1908 origin.
High Performance Ignition Coils
Resistor Plugs, Wires, and Related Issues
U-Gap and Splitfire Spark Plugs
Optimizing the Honda SOHC Ignition
Honda DOHC, CBX Ignition System Troubleshooting
Reading Spark Plugs
GL1000 Ignition Tech