How Ignition Systems Work

    Before you can figure out how to make something work better you first need to understand how it works. An ignition system builds up a charge then releases it at the right moment to ignite the air fuel mixture in the cylinder. There are two ways to improve an ignition system, increase the amount of energy that is thrown at the spark plug to increase the chances of lighting the mixture, and better time the spark delivery to get the most from the burn. The timing issue is covered on the Ignition Curve page. On this page I'll do my best to explain how ignition systems work. But first...

What the ignition system does

    The internal combustion engine is the most popular motivator of vehicles big and small. Internal combustion engines convert the energy of burning fuel into mechanical motion. An engine ingests fuel, along with air so the fuel will burn, compresses it then burns it. The burning mixture expands rapidly and pushes a piston, rotor, or turbine. Gasoline is the fuel of choice for most road going vehicles because it packs a lot of power per gallon (you can go farther with fewer and quicker fill-ups). The biggest problem with using gasoline is that it must be vaporized and mixed with the proper ratio of air or it will not burn. The flammability range is 1.4% to 7.6% (by mass not volume), so for every pound of air the motor pulls in you must mix in 0.224 to 1.216 grams of vaporized gasoline. Any more or less and it will not burn. The duty of the carburetor or fuel injection is to monitor the air input and meter the appropriate amount of gasoline. The engine then compresses the volatile air fuel mixture. The more its compresses the more power will be extracted from it when burned. Compressing the mixture heats it, if you compress it too much it will ignite on its own. That is how diesel engines work, however gasoline tends to explode rather than simply ignite which will tear an engine apart. The goal is to compress it as much as you can without it self igniting. Then apply more heat (from an external source) to a point in the cylinder to initiate combustion. Early engines used a glow plug. Sometimes that was as simple as a copper rod threaded into the head and heated with a torch. Eventually, they figured out the same could be done by creating a spark inside the cylinder. The first spark ignition systems made a constant spark so they functioned the same as a glow plug. The revelation that changed gasoline engines forever was the timed spark ignition. Instead of creating a constant spark that would light the mixture at any random time, a single spark is delivered with precise timing to most efficiently burn the fuel.
     There are several challenges with timed spark ignitions. The spark only lasts about a millisecond. If the conditions are not just right for that one millisecond then the fuel will not be ignited. The carburetor or fuel injection doesn't always add the right amount of fuel. Even if the exact amount of gas was mixed with the air it doesn't mean that every cubic centimeter of the cylinder has the perfect air fuel ratio. There will always be pockets of rich and lean. On top of that, not all the gasoline will be completely vaporized. Better fuel systems, intake manifolds, and heads are constantly being developed so that the mixture inside the combustion chamber stays as accurate and consistent as possible. But even with all the advancements and technology put into modern engines the ideal conditions for combustion are not always met. A good ignition system can not compensate for these bad conditions but it will increase your chances of lighting a less than perfect mixture. One trick is to increase the size of the spark gap. Making the spark travel through more air fuel mixture increases the odds that it will hit a pocket that is combustible. Another trick is to make the spark last longer. The air fuel mixture is swirling around in the chamber, the longer the spark is present the better the odds are that a good pocket of fuel will run into it and burn. To get a larger and/or longer spark you need to throw more energy at the spark plug. To do that you need to know...

How ignition systems work

    If you want to make a spark you need a spark gap. The electricity has to jump from somewhere to somewhere. Since the spark gap is subject to erosion and fouling it is made to be easily removed and replaced. There is a threaded hole in the head that leads to the combustion chamber. The spark gap threads into this hole and plugs it so the chamber is sealed. That's as good a reason as any to call it a spark plug, I guess. A spark plug has three main parts, the center electrode, the porcelain insulator, and the body. The center electrode is what carries the electricity into the combustion chamber. The porcelain insulator keeps the electricity in the center electrode from grounding out to the head before it has a chance to jump the gap. The body of the spark plug is what threads into the head and it also has a ground electrode connected to it which catches the spark from the center electrode and grounds it to the head.
     On many engines the spark plug is fed electricity through a spark plug wire. Spark plug wires need to contain very high voltage so the insulation is extremely thick. The conductor is often a resistive or inductive material to reduce electro magnetic interference which can interfere with electronic components. On older multi-cylinder engines the spark plug wires were connected to a distributor. A distributor is a rotary "switch" that is used to feed the spark to the appropriate cylinder. The distributor is fed a spark through a coil lead, which is the same thing as a spark plug wire. The other end of the coil lead is hooked to the coil. The coil is what produces the high voltage necessary to generate a spark. In an effort to simplify systems and increase service life, the coil lead was eliminated and the coil was mounted inside the distributor. The next step was to eliminate the distributor and instead hook several coils straight to the spark plug wires, often referred to as a Distributorless Ignition System or DIS. The next inevitable step was to eliminate the spark plug wires and mount the coils right to the end of the spark plugs, called Coil On Plug or COP.
    The ignition coil is what produces the several thousand volts needed to make a spark across the spark plug gap. Surprisingly the basic design of ignition coils hasn't changed at all in the last century. What we call a coil is actually two separate coils wrapped around a common metal core. In any other electrical system this would be called a transformer.
The coil you put electricity into is called the primary and the coil that sends the electricity out is the secondary. Transformers are used to step up or step down voltage. The voltage is changed according to the ratio of the turns. Say you have a transformer with 50 turns of wire on the primary and 100 turns on the secondary, a 1:2 ratio. If you put 10 volts into the primary you will get 20 volts out of the secondary. The trick to transformers is that to see anything from the secondary the current of the primary must be changing. If you put a constant DC current through the primary you will have no current flow through the secondary. Transformers are usually used in AC systems since the current is constantly changing. If you put 120 volts AC through a transformer with a turns ratio of 10:1 then you will get 12 volts AC out of the secondary. An ignition coil isn't fed with a constant AC signal. All we need is one spark so all we are going to put into the primary is one voltage spike. An ignition coil typically has a 1:100 turns ratio, so if 10,000 volts is required to generate a spark then we need to feed the primary a 100 volt spike.
    Notice I said if 10,000 volts is required. The required voltage changes constantly. Firing a spark plug outside the engine requires far less voltage than it does when firing an engine at wide open throttle. Pressure plays a major role, the higher the pressure in the chamber the more voltage is required. The size of the spark gap is the other big factor, the further the spark has to jump the more voltage is required. Firing a spark through an air fuel mixture requires more voltage than it would in pure air, especially with exotic fuels. There are many other small factors that determine how much voltage will be required to make a spark. You may see "high performance" coils advertises as 50,000 volt coils but all that is meaningless since under most conditions you will only need 5,000 - 10,000 volts. Now if you are running wide gap plugs in an alcohol burning, supercharged engine then you will likely need to step up to a system with a higher voltage potential.
    So all that's left is to explain where we get the large voltage spike for the primary. The ignition coil may not have changed in the last 100 years but the primary circuit sure has. There are two different types of ignition systems, inductive and CDI. Each system has its strengths and weaknesses. The ignition coil, spark plugs and such are the same between them. The difference is how they generate the primary voltage spike that drives the coil.

Inductive Ignition

induct.gif (1367 bytes)     Inductive ignition systems have been around almost as long as the internal combustion engine. It is a simple and rugged design which is why it is still being used on new cars today. With an inductive ignition all that's needed to make enough juice for a spark is the ignition coil, a switch and a low voltage power source. Wait a minute, how do you get a big voltage spike from a 12 volt battery with only a switch? The truth is, an inductive ignition doesn't actually throw a voltage spike at the primary. Instead, it makes the primary generate its own voltage spike. Oh, ok... wait... what? A coil of wire, also called an inductor, exhibits strange properties when you run electricity through it. You probably remember from elementary school what happens when you wrap a bunch of wire around a metal object then hook it to a battery. That's right, you get a magnet. The more current you run through a coil the greater the magnetic field. What you probably don't remember (unless you happened to be touching both ends of the wire when you disconnected the battery) is that the magnetic field is actually stored energy, and when the current flow through the coil stops the stored energy causes a voltage spike. Inductors resist current changes. Think of them like the electrical equivalent of a flywheel. When current is low and you try increasing it, the inductor will add resistance to try and keep the current low. When current is high and you try reducing it, the inductor will increase voltage to try and keep the current high. Since the current flow in the primary of our ignition coil goes from several amps to zero amps almost instantly, the voltage will rise very fast until it can get the current flowing again. The goal is to prevent any more current flow in the primary so the coil will start the current flowing again in the secondary, giving us a spark.
    Inductive ignitions are self adjusting. If a 100 volt spike is needed to initiate a spark the primary will only rise to 100 volts. The spark will then burn until all the energy in the coil is used up. The lower the voltage the longer the spark will last. Let's say you open up the spark plug gap and now 150 volts is required to fire the coil. When the current is interrupted on the primary, the magnetic field will cause the voltage to rise rapidly. The voltage will rise until it reaches 150 volts at which point the spark plug will fire. After the spark is started the voltage will actually drop since it takes less voltage to maintain a spark than it does to start one. The spark will continue to burn as long as there is energy left in the coil. Only now the spark duration is shorter since the spark fired at a higher voltage than before. Increasing the spark plug gaps without increasing the amount of energy stored in the coil can actually hurt performance. If most of the energy is used to initiate the spark then there will be little left to maintain it. It's even possible there won't be enough energy to make a spark in the first place. To get the benefits of wider spark gaps you need to put more energy into the coil. When you put more energy into the coil you will increase spark duration. When you have longer spark duration you can afford to lose some by opening up the spark gaps.
    Early inductive ignitions actually used a mechanical switch, commonly referred to as points, to physically break the primary circuit. When the points close it closes the primary circuit and current flows through the primary, when the points open the circuit is open and there is no current flow. In the seventies points were phased out in favor of electronic ignitions. Electronic ignitions use a transistor to switch the current on and off instead of a mechanical switch. There is no magic to electronic ignitions, they are just a switch the same as points. It's the switch opening (stopping current flow) that triggers the voltage spike that fires the coil.
    When the switch closes current flows through the primary, but it takes a while for the current to build up. That is the biggest issue with inductive ignitions. The inductance of the coil limits change in current. When the switch is open there is zero amps flowing through the primary. When the switch closes the current starts at zero and ramps up until it reaches its limit. It takes time for the current to build up. You can calculate how long it takes for a coil to charge with this formula, T = ( L * I ) / V where T is time in seconds, L is inductance in henries, I is amps, and V is volts. Say you have a 7 millihenry coil and want to charge it to 6 amps with 14 volts. T = ( .007 * 6 ) / 14 It will take 3 milliseconds (0.003 sec.) for the coil to reach six amps. That is not much time at all, so what's the problem? Believe it or not you don't always have that much time to charge the coil. For example, a V8 with a single coil running at 5000 rpm fires every 3 milliseconds. If you take away one millisecond for the spark that only leaves two milliseconds to charge the coil. In two milliseconds the coil will only reach four amps before it has to fire again. The energy stored in the coil is determined by the inductance and the current. The formula is, J = 0.5 * L * I * I where J is the energy in joules, L is henries and I is amps. Notice amps is squared which means small changes in amps will greatly effect the amount of energy stored. Our 7 millihenry coil at 6 amps holds 126 millijoules (0.126 joules), the same coil at 4 amps only has 56 millijoules of energy available to fire the spark. So more amps mean more energy but it also means more heat. The current needs to be limited or the coil will go up in smoke.
    One strategy for controlling current is to add resistance to the primary circuit. Ohm's law says that current equals volts divided by resistance, I = V / R. If you run 12 volts through a coil with a primary resistance of 3 ohms then the peak current will be 4 amps. Running a coil with a high primary resistance is the simplest way to limit current but far from the best. When voltage drops, like when the starter motor is engaged, the peak current drops and as a result you're spark output drops. A band aid fix is to use a lower resistance coil and add a ballast resistor. On many old cars you will see a 1.5 ohm coil and a 1.5 ohm ballast resistor, total series resistance is still 3 ohms so the peak current is the same. The trick is that the ballast resistor is bypassed when the starter motor is engaged so the circuit consists of only the 1.5 ohm coil. Now if the voltage drops to 9 volts when the starter is engaged your peak current will be 6 amps. That means you will have a more powerful spark when starting than you do when the engine is running which is good since it's harder to initiate combustion in a cold engine. The coil can handle the increased current because it is only done for a short period of time. If you bypassed the ballast resistor permanently you would likely burn up the coil. Early electronic ignitions still used a ballast resistor but they soon figured out the same task could be accomplished with the switching transistor. This makes the circuit simpler since there is no longer a ballast resistor or bypass circuit. But the big advantage is that it can automatically adjust resistance to keep the peak current consistent. If you have a 1.5 ohm coil it will add 1.5 ohms to keep peak current at 4 amps. If voltage drops to 10 volts it will only add 1 ohm so peak current will remain 4 amps. If you swap to a coil with a 0.5 ohm primary it will add 2.5 ohms. An automatic current limited works well but puts a greater load on the ignition module. A big heat sink is required to dissipate the heat from the switching transistor since it is still using resistance to limit current. With computer controlled dwell came the "ramp-and-fire" dwell strategy. Since the current ramps up gradually and at a known rate, the computer starts charging the coil late enough that there won't be enough time for it to go over current. If it takes three milliseconds for the coil to reach 7 amps then the computer will start charging the coil exactly three milliseconds before it has to fire. With this strategy you can run higher peak current without over heating the coil or module.
     That segues us nicely into the next critical element of inductive ignitions, coil charge time also known as dwell. Most guys when they hear the word dwell immediately think of old breaker point systems where you would set the dwell by adjusting point gap. Dwell readings were in degrees. Thirty degrees dwell meant that the distributor rotated 30 degrees between the time the points closed and when they opened again. When the points are closed the coil is charging. A V8 fires every 45 degrees of distributor rotation so the coil is charging for 30 degrees and is given 15 degrees to discharge. When the engine is running the coil is being charged 2/3 of the time regardless of engine speed. This is not a good dwell strategy. To see why, you have to look at dwell as time and not some arbitrary distance. At 1000 rpm it takes 15 milliseconds for the distributor to turn 45 degrees, so the coil will be charging for 10 milliseconds and given 5 milliseconds to discharge. It only takes 2 milliseconds for the coil to reach 4 amps where it stays for 8 milliseconds before the points open and the coil fires. The coil has reached its full power potential after 2 milliseconds, the additional 8 milliseconds does nothing but heat up the coil and waste power. The dwell strategy on early electronic ignitions wasn't much better than points (some were considerably worse) but as they improved the dwell times got closer and closer to the actual charge time of the coil. This means less power was being wasted. Coils were running cooler, so the peak current could be increased. More amps into the coil mean more energy out of the coil. As stated before, with more energy you will have longer spark duration and can run larger spark gaps. When you improve both of those you increase your chances of lighting the fuel mixture which ultimately means more power and better fuel economy.
     To sum it all up, inductive ignitions are simple and make a nice long duration spark. To do so they need as much current run through the primary as possible, which if not controlled can burn up the coil. It takes time to build up primary current and as rpm increases the time available to build up current decreases. When there is insufficient time to reach a high current, the energy available to fire the spark plugs drops significantly. This leads to misfiring at higher rpms.

CDI (Capacitive Discharge Ignition)

CDI.gif (1877 bytes)     CDI systems came out about the time inductive ignitions were becoming electronic. Conventional electronic ignitions simply replaced the points with a transistor but CDI completely reinvented the way the spark is generated. Instead of slowly charging the coil then relying on it to generate its own voltage spike, capacitive discharge ignitions charge a capacitor with high voltage which is discharged through the coil to make a spark. The capacitor can charge and discharge much faster than a coil so a CDI can operate at a much higher speed than an inductive ignition. The basic components of a CDI are a high voltage power supply, a capacitor, a switch, and a coil. The construction of a CDI system is a bit more complicated but the principle is quite simple. The capacitor is connected to the high voltage supply and charged. When it's time to fire, the capacitor is connected to the coil. The high voltage applied to the primary causes current to rise very rapidly and that's where the secondary get the power to make the spark. So the coil is basically just used to step up the voltage that the CDI module produces. That, in a nutshell, is the difference between a conventional electronic ignition module and a CDI module. A CDI module puts out power where a conventional ignition module simply switches the power to the coil on and off like a switch. Some guys refer to any electronic ignition as a CDI but it's only a CDI if spark is achieved by discharging a high voltage capacitor through the coil.
    It all starts with the high voltage power supply. There are two common approaches to generating the necessary high voltage. One is to have a voltage converter built into the module that converts the 12 volts from the vehicles electrical system into a high voltage. The converter consists of an oscillator to convert the 12 volt DC to 12 volt AC, a transformer to step up the AC voltage to several hundred volts, and a rectifier to make the AC back into DC. This arrangement is what makes CDI modules more complex and costly than a conventional electronic ignition. You will see voltage converters used mainly on cars and street bikes where there is a battery and charging system to power it. The alternative is to use the alternator to produce a high voltage AC signal. Then the CDI module only has to rectify it before it can be used to charge the capacitor. This design is generally used on dirt bikes and lawn equipment where there is no battery or charging system. It's also used on many quads and enduros, even those that have a battery and charging system. In a setup like that the alternator has two outputs, a high voltage output to power the ignition and a 12 volt output to run the lights and charge the battery.
     The high voltage supply is used to charge the capacitor. The characteristics of a capacitor charging are the opposite of a coil. When you put power to a coil the current starts at zero then ramps up at a linear rate as high as it's allowed, and you need to keep the current flowing through it to keep it charged. The amount of time the power is applied to the coil is critical. Too little time and the coil will not be charged. Too much time means that the current used to maintain the charge is being wasted heating up the coil. CDI eliminates these concerns. When you put power to a capacitor the current starts extremely high then drops exponentially as the capacitor charges. When the capacitor is charged the current flow is nearly zero amps. You can leave the capacitor hooked to the power and almost no energy is wasted. You can even disconnect the power and the capacitor will retain its charge. As you can see, with CDI there is no need for a current limiter or finely tuned dwell time. Generally, CDIs are more efficient since very little energy is wasted charging the capacitor.
    The energy stored in a capacitor is determined by the size of the capacitor and the stored voltage. J = .5 * C * V * V Inductive ignitions take a long time to charge because of the large coil required to hold a charge. A CDI can hold the same charge with a very small capacitor since it is being charged with several hundred volts. For example, an MSD 6 ignition charges a 1 microfarad (.000001 farad) capacitor to 500 volts. Do the math and you see that it stores 125 millijoules (0.125 joules). That's the same as the inductive ignition example in the previous segment. However, the inductive ignition took 3 milliseconds to charge where the CDI charges in only a few microseconds.
     OK, so the cap is charged. The only step left is to dump the charge into the coil. For that we need a switch. The schematic above is an overly simplified representation of a CDI circuit. An actual CDI circuit doesn't use a physical switch. Instead, an SCR is most commonly used as the switch. An SCR is basically a transistor that when turned on, stays on until current stops flowing through it. That eliminates the need to control the switch time. All you have to do is trigger the SCR, it will stay on as long as the capacitor is discharging and shuts off automatically when the cap is empty.
     There are a few variations to CDI circuits but the one shown is a very common scheme. It's a bit hard to see how it works when you first look at it. I animated the drawing to make it a little clearer. The process starts when the switch is open. The high voltage supply charges the capacitor. Notice how the current flows through the diode, bypassing the coil. [A diode is like an electric check valve, current can only flow through it one way.] The diode is not mandatory, many circuits don't have this diode, I only added it to make it easier to see when the cap is charging and when it's discharging through the coil. When the capacitor is fully charged the current flow stops. The fully charged capacitor sits waiting to be dumped. When the switch closes it connects the positive side of the capacitor to ground. This in effect bypasses the power supply and hooks the capacitor straight to the coil. Notice the positive side of the capacitor is hooked to the negative side of the coil. This is why, on many CDIs, you see a negative voltage when probing the positive coil lead. Once the capacitor is drained the switch turns itself off and the capacitor charges again.

    An inductive ignition charges a coil then allows the primary voltage to rise high enough to initiate a spark. CDIs don't wait for the coil to do the work. The high voltage surge from the capacitor causes the voltage on the primary to rise much faster and higher than it would on its own. This creates a very intense spark. [CDI is great for two strokes since the hotter spark can better fire an oily spark plug] Unfortunately to get the big hot spark you have to give up duration. A CDI spark may last only 50 microseconds (0.00005 seconds) where the spark from an inductive ignition typically lasts about 1 millisecond (.001 seconds). The short spark can hurt both driveability and gas mileage. A common approach to covering up this deficiency is to fire the CDI multiple times at lower rpm. While a few short sparks are better than one short spark, it's still not as effective as one long duration spark.

Which is better

    Each system has it's strengths and weaknesses. The question isn't 'Which is better' but rather 'Which is better for my application'. Inductive ignitions are rugged and simple. An inductive ignition with good current control and well timed dwell works excellent for most applications. However, inductive ignitions can't keep up on a high revving, multi-cylinder engine running a single coil. Modern motorcycle engines spin fast enough that even with a distributorless setup there is not enough time to fully charge the coils. That's why capacitive discharge ignitions are seen mostly on race cars and motorcycles.


If you have any questions or comments please e-mail me at