Ignition Systems; historical & modern types, with an emphasis on BMW motorcycles.
Vastly more information is available on the Author’s website, https://bmwmotorcycletech.info/index.html …..there are numerous articles on BMW Airhead electrical systems and on ignition systems for single spark plug per cylinder, dual spark plug per cylinder, timing units, points ignitions, and electronics ignitions.
Particularly see these two articles:
https://bmwmotorcycletech.info/ignition.htm.
https://bmwmotorcycletech.info/ignitiontheory.htm (Expands upon the information in the below article.)
The following article is a SIMPLIFIED explanation of ignition systems, although lengthy.
A common type of ignition in very early cars typically consisted of small wooden boxes containing various items: Vibrating electrical contacts sent battery energy into a coil of a modest number of turns of thin solid copper wire wound on an iron core of some sort. There was another coil of wire wound over that first coil of wire, but with thousands of turns of a much thinner wire, and these turns ‘transformed’ the lower voltage of the ‘primary winding’ fed by the battery on/off power (those vibrating contacts) to a few thousand volts, and this was applied to the spark plugs. In these earliest engines the spark was continuous due to the vibrating contacts, and thus was applied continuously to the spark plugs. Somewhat later a rotating switch was added. It had a contact or close by contact, one for each cylinder, and this item was called a ‘distributor’. The only problem with the vibrating contact system is that the spark output can not be set to ‘fire’ the spark plug at a very specific piston position, although the distributor method helped some. With the very low compression ratio and very low power output of these engines per cylinder volume, together with the very long stroke and large diameter pistons, and in particular the very low maximum RPM, these ignition methods worked OK. The ignition was rather ‘iffy’ when the engine had many cylinders. There were many variations of these internal combustion engines, including types with very large heavy flywheels that were arranged to only fire the mixture after a number of revolutions…these were typically used by farmers for pumping water.
Magneto’s, which were popular on both antique cars and early motorcycles…even into fairly recent motorcycles (particularly single cylinder types), are simply a permanent magnet method of using mechanical rotational energy (instead of a battery) to produce the high voltage needed for a spark plug. The spark is simply a hot means of igniting the air-gasoline mixture. An engineer would think about this differently, and might describe the ionization/plasma ….as the electricity was about to cause the spark (magneto or regular coil ignition). Except for the earliest magneto’s which produced an untimed spark, something like the vibrating points wooden box type driven by a battery, magneto’s were equipped with a set of points (contact points plate), which are nothing more than a mechanically coupled ON-OFF switch…and the energy transfer from the small number of turns coil in that contact circuit is done at the moment the points open, which is set via a cam located in the magneto which drives the points, so as to precisely coincide with a particular high energy alignment of the iron core with the wire in relationship to the magnet. The relatively small number of turns in the ‘primary winding’ coil acts in somewhat the same way as in the vibrating points method and the later points-condenser-coil method….there is a ‘transforming action’. This SAME SORT of transforming action is used in all coil ignitions. The basic method is called the Kettering Ignition, usually meant to be a coil, points, and condenser (capacitor) across the points….and used with a battery; but the magneto type works somewhat the same. More on all this later herein.
BMW motorcycles used magneto’s in the early models, prior to late 1969 when the /5 series was introduced. Magneto’s are difficult to keep down in size when you need quite high energy sparks. Later lean-burning engines (and, especially, higher rpm engines) need high energy sparks. Magneto’s have certain advantages, often the biggest such is that no battery is needed. Magnetos produce more spark voltage at higher and higher rotational speeds (opposite to coil ignitions), which is another reason why magnetos were popular, particularly on some race engines a long time ago….besides the weight of a battery not being needed (race engines were not generally started by ‘starter motors’ and did not need lights way back then). Unfortunately, magnetos need to be fairly large to produce a good spark at cranking rotation speeds. Magnetos were, and still are, used on some ‘stationary’ engines; and versions of magneto ignition are still used on some small motorcycles, lawnmowers, snowblowers, etc. Many small engine magnetos are designed so that the magnets are part of the engine flywheel, and are located near the outer edge, so the speed of the magnets going by a stationary coil, is higher during cranking, which helps create a reasonably decent spark.
From the the time of introduction of the BMW /5 series in December of 1969, until 1981 when BMW introduced the Bosch electronic ignition, BMW’s had coil ignition with the points (‘contacts’) being driven by the engine camshaft, that same camshaft that operates the valves. The contacts, nothing more than a switch, have a capacitor, often called a condenser, electrically connected to them. This system of contact points, capacitor, and coil, all being driven by a battery, was invented by an engineer from GM, and even now, this system is still called the ‘Kettering’ ignition. In our Airheads, as in all common 4-stroke engines, the camshaft rotates at precisely half the crankshaft speed. The Kettering ignition is vastly more efficient at producing a strong spark; and while the only part seemingly added is the condenser, the function of the circuit is quite different in some respects. In a Kettering ignition, the points can be replaced by a moving vane and transistor circuit, and by other means. In quite modern vehicles, the ignition is triggered by a computer. Some ignitions have one ignition coil for each spark plug, and the ignition coil unit might mount directly onto the spark plug. Modern high performance engines require very carefully controlled ignition timing, the ignition is also very powerful, even possibly deadly if one comes into poor contact with it.
In the Airhead, at the forward nose area of the camshaft, is a bob-weight type of simple mechanism called the Automatic Advance Unit (ATU), whose purpose is to change the ignition timing in relation to the piston position in the cylinder, most all of its operation is in the slower rpm range. This is needed for an engine that must be started at a low rpm; this applies to just about every engine. This ATU mechanism, up through 1980, has its own ‘two-bump’ cam, and it is THAT cam that directly operates the ignition points. The automatic advance is so arranged that the ignition timing ‘spark advance’ finishes all advancing by approximately 3000 rpm, in all models (except some very early /5 models, when it was at approximately 2000 rpm). In 1979 BMW put the points and advance mechanism into a metal canister, and made some improvements in the stability of the mechanical drive. The engine valves cam nose became a flat slot and an offset coupling was used called an Oldham Drive, with the male portion being part of the canister mechanism. The next and final major BMW design change was incorporated in 1981, and it was in a similar canister, but eliminated the points in favor of an electronic (magnetic) triggering device, called a Hall element transistor, which triggers a complex electronic amplifier and coil driving device we call ‘the module’. The module is located under the fuel tank on a heat sink. The Module has electronics with certain electrical characteristics added that the points system did not, and the module and those characteristics were updated a few times, as were the coil(s) used. The 1981+ system, if not abused, has a quite high reliability, …..and exceptionally good performance. BMW used both two each 6 volt coils and one only 12 volt coil, depending on model and years. The coil primary windings had to match the voltage, and the current flow at the points (when points were used, up to 1980). Points will not directly handle the primary coil winding current drain of the electronic ignition coil(s).
How…and when… does a coil make the spark? Except for the very antique vibrating contact coils of engines generally made before the 1930’s, coils consists of an iron core of multiple thin iron rod lengths, that has wound around all of them (as one bunch), TWO separate windings; one of thicker wire and much fewer turns called the primary winding, and the other with many more turns of a thinner wire, called the secondary winding. The primary winding is connected to the vehicle electric system via two coil terminals, typically a male spade. The secondary winding is either connected to one tower terminal, or, two tower terminals. For the one-tower terminal type of coil, the remaining secondary winding wire connects to one of the primary terminals, internally.
BMW used both types of coils on its Airhead models, but only the single tower types (and hence TWO of those coils) on the earliest models, with the R80ST and R80GS having the first of the dual-tower one coil models….which eventually became standard. Although there are some technical differences, for practical purposes both ignitions perform mostly the same.
If the system has mechanical points and a capacitor (also called a condenser), here is how the system operates in an Airhead, that has the most common battery system, that is, the battery negative terminal connects to the engine casting and chassis:
The battery supplies electricity, through the ignition switch, to one spade terminal on the coil, which is one end of the mentioned PRIMARY winding. On some coils, that terminal is numbered, and on some it has a + symbol, and on some it has both number and + markings. The other spade terminal on the coil, the minus (-) end, connects to the non-grounded insulated points contact (or, to another coil primary winding + terminal, and the minus (-) side of IT goes to the points contact). A capacitor (condenser) connects to the points ….and to chassis ground via its metal case. Conventional wiring systems used by the Germans use a standardized numbering system, and terminal #1 is negative (-) and #15 is positive (+). Since the power into the coil from the battery is applied to terminal #15, it can be marked +. Since the switched end of the coil (switched to ground by the points or module) is in the chassis connection end of the coil, that coil terminal is marked #1 or -, or both. It is relatively important for this type of dual coils system that the coils be properly connected for + and – polarity, otherwise the electricity going to the spark plug cap terminal will be of wrong polarity and possibly have a more difficult time for the spark to occur at the spark plug, under some circumstances. For most twin-tower coils, there is no marking for 15, 1, nor – nor +. That means that one plug will always have a possibility of a more difficult time in firing the spark. This is not excessive, however.
One of the points contacts is mechanically and electrically connected to the metal plate of the points contact assembly, and thereby completes the circuit to the battery negative terminal, through the engine casing, etc., when the points are closed, that is, touching each other. The other points contact is insulated from the chassis and goes to the coil primary, as noted.
When the points are closed, which they are for a MUCH longer time of engine camshaft rotation compared to the time they are open (this is what is meant by dwell angle, as a portion of a single rotation); current from the battery flows through the coil, and back to the battery via the points. During the time that the points are CLOSED, the electricity from the battery, causes the iron core in the coil(s) to build up (charging magnetically) a strong magnetic field, hopefully up to the maximum the iron core will accept. If the engine is rotating relatively slowly, the design is such that this maximum always does occur. The greater the magnetic field charge, the higher the spark energy will be. If the points are closed, ignition is ON, and engine is not rotated, the coil(s) will charge to as strong a magnetic field as the coil design allows, given whatever the battery voltage is. When the points first CLOSED, the current (to charge the coil magnetically) began at a low level, just exactly WHY is not going to be discussed here… then current increased until it reached a maximum allowed by the resistance and other characteristics of the primary winding. This magnetic charging ‘curve’ (an exponential function) and the reason behind why it is not instantaneous, is a matter of electrical theory for engineering nerds.
For a BMW Airheads, a two-cylinder engine whose pistons go in and out at the same time, but whose cylinders are NOT on the same stroke, when the desired position of the particular piston (on its compression stroke) is reached, the points are ‘timed’ so they will JUST BARELY begin to separate. The maximum opening gap adjustment and points condition, and its cam lubrication is checked every 5000 miles or so on points ignition models. The points contacts, and rubbing block (which is lubricated by a felt against the cam on non-canister models) do wear, and the points usually need replacing at 15-20,000 miles or so; unless an aftermarket ‘amplifier’ has been installed which can extend the points life (if the cam is also faintly lubricated so the rubbing block insulator does not excessively wear). At the exact instant the points begin to open, the circuit is broken (opened), from which came the name “breaker points”), which means disconnected…..and the coil Primary Winding cannot accept any more current from the battery as there is no longer a completely connected circuit.
Once the coil is fully magnetized, there is no electrical reason to have the battery draining further into the coil. The small cam that drives the points is arranged to have a ‘dwell’ that is adequate, for any range of rpm the engine will work at…plus an additional safety amount. IF the coil(s) were operated on an engine that was capable of enough rpm (and/or, additional number of cylinders), the coil output would drop off. Another way of saying this is that as the number of spark outputs a coil produces …in a given time …rises above a certain number, …the coil can not be fully magnetically charged, as there is not enough dwell TIME between spark events. If the situation is bad enough, there will be INsufficient spark plug voltage/strength and misfiring or other ills will occur. These things do NOT happen with our Airheads (assuming the engine and ignition is in good condition), as rpm and number of cylinders is rather low, compared to coil capability, and the dwell is more than long enough.
Dwell used in our Airheads has varied over the years, and was reduced in the /5 era, when some Authorities (Police) models with metal shielded coils overheated. Still, the dwell is more than adequate, for a coil(s) on a two cylinder engine at Airhead capable rpm.
At the exact instant the points open, the coil is ‘loaded’ with magnetic energy, which needs to go someplace. I will get into that a bit later herein.
Some have asked about making measurements at the points to ‘static timing’ the engine (engine not running). This is acceptable, and BMW even provides an “S” mark for that, to begin the timing process, but the correct final method is by “timing the engine” at the maximum specified advance rpm. We actually use a slightly higher rpm to be SURE that the advancing HAS stopped increasing. But, yes, you can use an ohmmeter or a voltmeter, or a piece of the very thin ‘cigarette’ rolling paper at the points. If you use an ohmmeter, leave the ignition off or you might burn out the meter. Either an ohmmeter or a voltmeter is connected across the points which in some cases is easier to do by connecting the meter at the appropriate terminal on a coil. As the engine is VERY slowly manually rotated as it approaches the timing point, the ohmmeter indication will increase (more ohms). If using a voltmeter, turn on the ignition, and the timing is when the voltage rises suddenly from zero (points closed) to approximately battery voltage.
If using cigarette rolling paper, which is extremely thin, the timing point is when the cigarette paper can just barely be pulled through without the points holding it.
You must rotate the engine in the forward (normal) direction VERY SLOWLY when finding this exact timing point. Some do that with the spark plugs out, some with them still in place; some do it with an allen wrench into the alternator bolt, which I am not in favor of, as it can overtighten the bolt. I prefer bumping the rear wheel in top gear, spark plugs NOT removed, ignition OFF (or spark plug caps securely grounded!). I prefer spark plugs not removed, as every removal adds to threads wear in the head.
Remember that capacitor/condenser I have mentioned? During the longish dwell time the points were closed and the coil was charging up in magnetic energy, that capacitor is electrically SHORTED by the points. At the instant the points separate the tiniest amount, …that mentioned timing point, ….the capacitor now suddenly receives the voltage created by the magnetic field in the coil via the PRIMARY winding. I will restate that slightly….at that instant, the coil ‘tries’ to charge the capacitor with its energy, which it changed from magnetic to electrical. Since it takes TIME for a capacitor to ‘charge up’, and before it is charged-up the capacitor acts like a near short circuit…. the coil dumps energy into the secondary winding via transformer action. This is a somewhat wrong explanation if you are an electrical engineer…but it simplifies the understanding here. The secondary winding now develops the high voltage to enable firing the spark plug(s) by transformer action. The primary winding voltage is increased in the secondary, by the magnetic coupling, and the increase is approximately the RATIO of primary winding turns to secondary winding turns. NOW you know why there are MANY more turns of secondary winding wires…and why the secondary winding must be of very small gauge wire, in order to fit inside the coil case dimensions. The points contact winding (the Primary) also sees an increase in voltage, since transformation is not perfect, nor is the condenser function, although much lower in value. The bottom line here is that the approx. 12 volts of the battery is now transformed by the ratio of the number of turns between the two windings….and we get a quite large voltage to fire the spark plug. The voltage developed at the points can be high enough to cause some points erosion by arcing, even with the condenser. Most folks think that the reason for having the condenser at all is to prevent or lower the points contacts wear from their sparking (yes, there is a tiny spark at the points during use). This is SORT-OF true, but hardly all of the story, as you can see from the above explanation. …..you should understand that the shorted condenser is not shorted when the points open, so the coil primary winding does some fancy things to cause the secondary winding to have a lot of sudden high voltage. The effect of the primary winding is to reverse its voltage polarity and dump into the condenser, in a decaying oscillatory waveform. Since a transformer needs AC, not DC, that decaying oscillatory waveform is excellent for doing ‘transforming’. Now, don’t you feel smarter?
OK, a bit more. For complex reasons during the above described process, the condenser, which has a capacitance value selected to approximately match the coil(s) inductance value (an technical electrical term), reduces (but does not eliminate totally) the electrical arc (sparking) at the contacts, but it does give them a longer life. Nerdy thing: Since the points always have negative voltage at the grounded contact point, and mostly positive at the insulated contact point, electrical spark wear is such that one point gains a tit, the other gains a depression of the metal….which makes the tit. Is your head spinning?
More Nerdy-ness: In a system as described, the high voltage output of the coil has a rather complex type of waveform, but it is definitely polarized (positive and negative). That is why the two spade terminals on the coil must be connected to the correct wires on the coils with one high voltage tower output. While that coil WILL fire a spark plug if the coil primary terminals are connected backwards, the spark is effectively reduced, as a quite hot spark plug tip does its spark jumping job easier if it receives negative voltage. At least, that is the commonly accepted theory. In reality, it is more complex, with the effect actually being the gas/air ionization (or, call it plasma) from the high voltage.
In BMW Airhead motorcycles the NON-firing cylinder (a two-tower coil is assumed for this discussion) has vastly less air-fuel pressure since it is not on the compression stroke, and so the spark jumps that plug much easier, leaving plenty of energy left to fire the other cylinder which is on the compression stroke. The coil output is also generally designed to be higher than a two coils, one tower per coil setup that BMW also used on most models before 1981 (and, many into 1984), and electronic ignition was used (from 1981) to ensure an even more powerful amount of energy to the spark. A more complex interpretation will also explain other functions such as the speed of the module’s actions, and that the MODULE does not need the condenser, and the high voltage is therefore may build more rapidly; which for complex reasons fires the spark plugs easier.
High voltage is always potentially dangerous to humans. Voltages over maybe 40 or 50,000 can do more than just create pain. The high voltage (over 20,000 is typical on every BMW) available at the spark plug of modern engines is dangerous to some folks in particular…..and the high over-all energy in the high voltage in the electronic ignition models can be VERY dangerous. Be careful!
For those motorcycle models with TWO single tower coils, the coils primaries are wired so that both coils get + battery energy at the terminal marked + or 15. For that system, both spark plugs get the proper polarity for their particular cylinder on compression/firing stroke. For BMW Airhead motorcycles with dual plug ignition conversions TWO dual output coils are wired with the primary windings in series; the high voltage secondary outputs have a positive and negative output at any one particular coil’s terminals, the same as the above single coil with two outputs. The operation is very similar. Because the output of the coils is always + at one terminal tower with respect to the other, it is best to wire a twin tower coil used on a dual-plugged bike, such that EACH coil fires a spark plug on BOTH cylinders. For those who are extra-nerdy, I recommend, on dual-plugged conversions, that the BOTTOM spark plug, which tends to carbon-up more than the top plug, get the negative outputs (and, perhaps the spark plug will be one step hotter in heat range). One can discuss and debate all this.
In our BMW electronic ignition models (1981+), the points are replaced by some semiconductor parts (Hall elements/devices) that produce a triggering signal by the proximity of a specially shaped rotating metal plate. The electrical signal from these devices is very small, and the signal is applied to the electronic current amplifying circuit inside the module under the gas tank. The triggering device, called a Hall Element or Hall transistor, is VERY sensitive to stray electric fields. That is why the spark plug wires must NOT be disconnected if the engine is running, and why 5000 ohm spark plug caps are a must. This system is very reliable and requires no regular maintenance, with the exception of checking the timing every 10K or so, and reapplying heat conducting grease to the module under the tank every couple of years on those modules that are fastened by screws, not rivets. The timing on the 1981+ models changes VERY slowly over VAST mileages, mostly due to wear on the timing chain/sprockets.
No matter what the method, contacts or electronics, some means must be available to change the timing of the spark in relation to the piston stroke, as the rpm increases. The reason is MAINLY that as rpm increases, there is less and less time for the ignited mixture flame to completely burn, so the spark must occur EARLIER in the process. BMW has used a mechanical advance device, called an ATU (Automatic Timing Unit) in all its Airhead models. This mechanical device is simply a pair of movable weights of a certain shape, used with springs of specified strength. The weights move from the at-rest position to farther and farther outward as rpm increases, up to the weights preset mechanical stops. Rotational centrifugal (centrepital) force causes the change…just like a weight on a string that you could rotate at speed over your head. By carefully designing the weight shape, mechanical stops, and spring tension, the factory is able to give a reasonably optimum…..’advance curve’…so that the proper timing occurs at the rpm found best by factory dynamometer and road tests. No device such as the automotive type of vacuum advance/retard is added…but, it is quite good enough. On a practical basis, it works out PERFECTLY OK. Earliest /5 models used about 2000 rpm for the maximum point of advance, & later models used about 3000. Use of the early advance units (2000 rpm) is often impractical, causing pinging with today’s low octane fuels on some models at lower RPM, but can give modified engines on premium high octane fuels slightly more low rpm acceleration…although in an area of rpm not good for the engine to be operating in for longer than a very few seconds. As previously mentioned, there were changes in this advance unit over the years, including the weights, the cam profile (changed dwell) and springs tension and even the mechanical stop limits. I can identify your advance and its stock characteristics if you supply me the Bosch number stamped on the outer part, but have no good way to identify the spring force for you; although the advancing versus RPM is a strong indication. I have an article that translates the numbers for both BMW and Bosch (who made the ATU’s).
There are other types of ignition systems for engines, one popular high performance type is called a ‘capacitor discharge ignition’. If carefully designed, that type of spark can occur so fast that it will often even fire fouled spark plugs. They are NOT needed or desirable, on our BMW Airheads. There are various types of aftermarket electronics conversions for the BMW Airheads. At least one from Germany is very expensive and very fancy, and uses triggering right off the crankshaft, eliminating any instability or aging of the chain and sprockets. There are some American or UK ignitions that do that too. One type eliminates the points (Dyna), although they have other models; including an ‘amplifier’ designed to be used with the points that reduces points wear (but does NOT reduce points rubbing block wear). The Boyer ignition has a built-in advance curve, and does not needing the BMW mechanical advance. The Boyer advance curve is adequate, although it does not match the Airhead’s exact requirements.
Since this article was originally written, additional electronics ignition units are available from such as http://euromotoelectrics.com and http://motoelekt.com
02-13-2021
Snowbum
HTTPS://BMWMOTORCYCLETECH.INFO/INDEX.HTML
