Last month this column discussed the combustion part of the operation of internal combustion engines in general terms. This month I want to zero in on the combustion problems sometimes encountered with Oilhead BMW motorcycles in more specific terms. The Oilhead BMWs are the R1100 and R1150 series bikes. Most of the comments also apply to the R1200C cruiser motorcycles, and many also apply to the R1200 Hexhead motorcycles.

The Systems

The Oilhead motorcycles are equipped with fuel injection systems composed of a fuel pump, fuel pressure regulator, throttle bodies, fuel injectors, an Engine Control Unit (ECU), several sensors, and the associated wiring and plumbing that connects all the parts together as a system so as to deliver fuel to the cylinders.

            The Oilhead motorcycles are also equipped with a “computer controlled electronic ignition system” composed of a crankshaft position sensor (Hall sensor), the Engine Control Unit (the same ECU that controls injection), coils, spark plugs, and the wiring that connects all the parts together as a system so as to deliver ignition sparks to the cylinders.

            The key part of both the fuel injection system and the ignition system—the part that ties it all together to deliver the proper amount of fuel and ignition sparks at the proper times is the Engine Control Unit (ECU). The ECU units used by BMW on the Oilhead series of motorcycles are Bosch Motronic units. The Motronic ECU uses a computer program to determine when, and for how long the fuel injectors squirt. It likewise determines when the spark plugs fire.  As noted, it uses the inputs from various sensors to do this. The two primary inputs to the ECU which have an effect on fuel injection are the crankshaft position sensor and the throttle position sensor. For ignition, the primary input comes from the crankshaft position sensor. One critical part of the computer program in the ECU is the “map” which can be viewed as a three dimensional set of numerical values. The inputs from key sensors are matched to the pre-programmed values which then determines when and how much fuel is injected.

            While not exactly a system, the combustion chamber itself is a critical part of the attainment of proper combustion in an engine. In an Oilhead engine combustion occurs inside the cylinder and the pressure created by the combustion drives the pistons which in turn drive the connecting rods which turn the crankshaft and all of the connected parts, including the rear wheel which drives you down the road. For the purposes of this discussion the “combustion chamber” is that portion of the cylinder and head assembly in which the air fuel mixture is contained at the moment of ignition.

            Conventional engine terminology includes the terms “top” and “bottom” to denote things closest to the head(s) and furthest from the head closest to the crankshaft respectively. This terminology arose when engines were typically built so the cylinders were upright and had a top and bottom.  But it has been rather uniformly applied to V8 engines, radial aircraft engines where the cylinders point every which way, and horizontal engines like Continental aircraft engines, Subaru engines and Volkswagen Beetle boxer engines. While I do know that Oilhead engines have horizontal cylinders, it is too confusing to try to talk about inside and outside, head side and crank side, or other such terms. I will use the conventional terminology of top and bottom and above and below even though we all know it means out there by the valve covers and in there by the crankshaft.

            BMW uses a semi-hemispheric combustion chamber in the heads on Oilhead engines. A true hemisphere is half of a round ball (sphere). The semi-hemispheric combustion chamber should be viewed as shaped like a piece sliced off the edge of a round ball rather than a full half of that ball. For the purposes of this discussion the combustion chamber also includes a small portion of the top of the cylinder above the piston, and the irregularities in shape created by the valves and spark plug(s).

Combustion

When you start the engine lots of things happen at once. The crankshaft turns and the pistons move up and down in the cylinders. The valves open and close. The fuel injectors squirt. The spark plugs spark. When operating properly it is a mechanical symphony.  The fuel injector squirts fuel into the intake tract behind the intake valves, which then open allowing the fuel and fresh air to be drawn into the cylinder. The piston then travels toward the top of the cylinder compressing the air. Slightly before the piston reaches top dead center (all the way to the top) the spark plug(s) fire igniting the air-fuel mixture. As the mixture burns, pressure builds, pushing the piston to the bottom of its stroke. As the piston again travels toward the head the exhaust valve is open and the burnt gasses are pushed out of the cylinder. It starts all over again. The complete cycle takes two complete revolutions of the crankshaft. On the BMW Oilhead engines the cylinders alternate firing—one or the other cylinder firing each revolution.

            If everything is working properly the engine runs well. However, if any of a number of things are not quite right—off just a little bit—the engine won’t run as well. If some things are not quite right, the engine won’t run at all.

Oilhead Combustion Issues:

The earliest and most notable combustion issue which reared its ugly head on the Oilhead bikes was surging. Surging is hard to define precisely but you know it when you see it. At its most benign it is a vague feeling of the engine slightly speeding up and slowing down. At its worst it results in a bucking sensation in the motorcycle almost as if the engine us being turned on and off very rapidly. Surging exhibited itself at constant throttle settings and low to moderate speeds and loads. Except for the possibility of some strange cases, the surging exhibited by Oilhead motorcycles is “lean surging” caused by the fuel air mixture being too lean and not burning properly. To witness a typical lean surge all you need to do is go out and run your lawn mower out of gas. You will notice the engine speed up and slow down several times before it finally dies from fuel starvation.

            Somewhat like final drive failures are an issue today, surging was the big issue with the early Oilhead bikes. On some bikes the surging was minor but irritating. On others it could become major and almost disabling. My late friend Rob Lentini did a lot of research and experimentation regarding surging on these bikes. His research and writings included changes to the throttle position sensor settings, and substitutions for the CAT code plug which is an electrical jumper block that the ECU uses to tell what model bike and what market (U.S.A., Europe, etc) it is and thus which of several pre-programmed “maps” to use.  Rob also experimented with spark plugs, and the air tubes running from the air box to the throttle bodies.

            On many of the early bikes it was possible to eliminate the surging altogether, or to minimize it to the point it was barely noticeable by very carefully tuning the engine. This included being certain the valves were precisely adjusted and the throttle bodies were precisely synchronized. Sometimes substituting a single-electrode spark plug for the double and triple electrode plugs originally supplied OEM helped too. Many owners and even many dealership technicians could not believe or understand how such subtle changes in engine tuning could possibly make such a huge difference in engine performance. A careful step by step analysis makes it pretty clear.

            The key sensor which enables the ECU to determine how much fuel to inject each revolution is the throttle position sensor (TPS). BMW Oilhead engines do not use a mass air sensor to directly measure air flow into the engine. Rather, the ECU computes how much air is entering the engine based on the position of the throttle plate in the left throttle body. The TPS in mounted on the left throttle body and as the throttle plate opens and its shaft turns, the shaft also rotates the internal wiper in the TPS, altering the voltage returned to the ECU from the TPS. As the throttle opens more air is allowed into the engine, and it is the TPS that tells the ECU how wide open the throttle is.

            The TPS mounts to the left throttle body with two screws located in slotted holes. Its adjustment is accomplished by loosening the screws and rotating the entire TPS slightly. On the earlier versions of the Oilhead bikes equipped with the Motronic Version 2.2 the shop computer tool, Moditech, read TPS voltage and displayed dashes and a 0 to indicate the correct setting.  The Moditech read a range of .370 to .400 volts at idle as acceptable and would give the display for the in-range setting. Even within this acceptable range there are mixture differences between a TPS idle setting of .400 volts and .370 volts. Even which way the technician rotates the TPS while adjusting it makes a difference. Since the Moditech uses a range of voltage as acceptable, rotating the TPS clockwise until out of range and then back until the Moditech display flops in, leaves the voltage near .370 volts. Rotating the TPS counter clockwise until out of range, and then back until the display flops in, leaves the voltage near .400 volts, thus richer.

            The later Motronic Version 2.4 control units are more sophisticated and memorize the throttle position. When the key is on and the throttle is rotated from closed to fully open and back several times before starting the engine, the Motronic memorizes the range and values of the TPS voltage.

            Note—the ECU only receives an input from the left throttle body. It has absolutely no way of knowing how wide open the right throttle body is. By design, the ECU assumes that the right throttle body and left throttle body are open exactly the same amount in all positions at all times. So adjusting the throttle bodies so that they are synchronized (open the same) is critical to having the proper mixture in both cylinders.

            If one throttle plate is open further than the other then that cylinder will get more air. But the injectors are supposed to squirt the same amount of fuel. So that cylinder will be leaner than the other. Enter again the Motronic ECU and the Oxygen (O2) sensor inserted into the exhaust stream. Based on the voltage of the signal from the O2 sensor the ECU will adjust the duration of the fuel injection pulses to enrichen or lean out the mixture. Once this adjustment happens, if the throttle plates are not synchronized and open the same amount, one cylinder will be rich and the other lean, even though the ECU and O2 sensor are happy with the blended exhaust from both cylinders.

            If the two throttle bodies are not synchronized engine vibration will be noticeably increased. The engine will have a rough idle. Power may seem somewhat sluggish. In the extreme, heat or pre-ignition damage can occur to the exhaust valve or piston on the lean side.

            So synchronization of the throttle bodies is critical to combustion mixture. Throttle plate position and synchronization at idle is set on a flow bench at the factory. The idle stop screws are then “locked” with blue paint and are not supposed to be disturbed. Fine adjustments are made by adjusting the brass air bleed screws in the throttle body and not by disturbing the idle stop screws.

            OK—good—the throttle bodies are perfectly synchronized at idle. You don’t ride around much, if any, at idle. Those throttle bodies are located about two feet apart, one out on each side, connected only by the throttle cables. So, for the throttle plate positions to be synchronized anywhere except at idle, the cables need to be precisely adjusted.

            When you think about it and do it methodically the process of synchronizing the throttle bodies is pretty simple. To do it you need a dual manometer (vacuum measuring device) to directly compare manifold vacuum in the two cylinders, a wrench to fit the locknut on the cable adjuster (10mm usually), and a small screw driver.

            Step one is to make sure that the idle speed lever (wrongfully called the choke) is fully off and not holding the throttle plate linkages off the idle stops. Step two is to make sure there is some slack (1 to 2 mm) in both throttle cables. Step 3 is to attach the hoses from your dual manometer (Carb Stix, Twinmax, or equivalent) to the vacuum take off stub pipes on the bottom of each throttle body. You need to pull the hoses to the charcoal canister off the stub pipes to attach the hoses from the manometer. Step 4 is to place a good fan in front of and blowing on the engine. Step 5 is to start the engine.

            Step 6 is to adjust the brass idle air screws so that the vacuum readings for the two cylinders are as close to identical as possible at idle.

            Step 7 is to then synchronize the cable adjustment. There are two distinct schools of thought as to at what engine speed you should do this—just off idle, or at a higher RPM such as 3,000. To explain why, I have to get technical. You are trying to equalize air flow through the throttle bodies. You are adjusting the angle of the throttle plates, and thus the cross sectional area of the open spaces between the throttle plates and the walls of the throttle bodies. You are not measuring air flow directly. You are using manifold vacuum as a surrogate measure for air flow.

            If you synchronize the throttle plates just off of, but very close to idle then a very slight difference in throttle plate angle will result in the greatest percentage difference in open area, actual air flow, and vacuum. This is the most precise way to synchronize the throttle plate position.

            Some folks have noticed that when the synchronization is done just off idle the throttle bodies are not well synchronized at higher RPMs where they normally ride. This is sometimes true. Other factors sometimes interfere with synchronized air flow. For example, intake valves not adjusted equally, or carbon deposits on valve stems or the back of valve heads can alter air flow to one cylinder or the other despite the opening of the throttle body; so can air leaks. Sometimes small pebbles or grit located in a throttle cable pulley alter its radius and thus the position of the throttle plate.

            I personally synchronize the cables just off, but as close to idle as I can. Then I check synchronization by observing it at higher RPMs. If it drifts off I look for the reason so I can correct it.

            Now back to actually synchronizing the cable adjustments. First I make sure that I have at least one millimeter of free play on the left cable. If I do I leave it alone. If I don’t I adjust that cable so that I do and lock the adjuster by lightly tightening the locknut. Then I go to the right side and adjust the right cable so that I have equal manometer readings with the throttle held just off idle—say 1,200 RPM. Tighten the locknut. Recheck. Done!

            Several things might interfere with this process going smoothly and they do affect combustion. If turning the idle air screws doesn’t seem to change things much you might have a clogged air passage in the throttle body. Remove the screw. Clean the carbon and gums off its tip. A few squirts of good carburetor cleaner followed by a few bursts of compressed air into the hole where you removed the screw usually fixes this problem.

            As noted, if you get good synchronization just off idle but it drifts off at higher RPMs then something else is wrong. You really should adjust the valves before synchronizing the throttle bodies. Cables stretch faster than valves wear. Checking synchronization more frequently than actually adjusting valves is common. If synchronization is not fairly stable throughout the RPM range then you should check the valve adjustment.

            If the valve adjustment is perfect and the problem persists, then you may have valve deposits, or you might have an air leak. Or you might have a little pebble stuck in a throttle body pulley. Check this first because it is easiest. The most likely locations for air leaks are where the throttle plate manifold bolts to the head; where the throttle body attaches to the rubber manifold; or at a throttle plate shaft seal. To check for air leaks I either spray carburetor cleaner to see if it gets sucked in making the engine speed up, or I use WD40 while looking for puffs of smoke in the exhaust.

            If you don’t find a leak you probably have valve deposits which are interfering slightly with air flow in the intake tract. To combat or remove such deposits a dose of Chevron Techron Fuel System Cleaner is useful. Read the label. Reading comprehension counts.  Run an appropriate amount in one or two tanks of gas. Since it is supposed to remove carbon, gums, and other deposits, changing the oil afterward to get rid of the junk is recommended.

            So much for the delivery of air into the combustion chambers. Next we need to consider the delivery of fuel. Too much or too little air effects mixture. So does too much or too little fuel. Start with the fundamentals. The amount of fuel squirted by a fuel injector depends on three things:  how big the opening in the injector is; fuel pressure; and how long the injector stays open.

            A fuel injector is simply a little magnetic valve. When a pulse of current is applied an electromagnet pulls the needle which is blocking the nozzle out of the way, allowing the nozzle to squirt. It stays open as long as the pulse of electric current remains applied. The opening is relatively fixed—the hole is only so big. Two things can alter the size of the hole. Eventually, at high mileage an injector can wear—the hole gets bigger. But more commonly, little bits of gummed fuel, varnish, or grit can accumulate making the hole smaller. Re-read what I said above about Chevron Techron Fuel System Cleaner (or other quality fuel injector cleaner) to solve this problem. I generally use one dose in a tank of gas just prior to an oil change with the major service at 12,000 mile intervals.

            The length of time the injector stays open is determined by the length of time the ECU applies electrical current each pulse. This length of time, or pulse width, is determined from that “map” in the program in the ECU based on the inputs from several sensors. The TPS adjustment is critical. But the temperature sensor and other sensors come into play too. If an Oilhead runs too rich it is almost always the TPS adjustment or a faulty temperature sensor. It could be worn injectors but that doesn’t happen all that often. Certainly it seldom happens on low mileage bikes.

            The final variable that affects fuel delivery is fuel pressure. The most frequent cause for pressure variations is a clogging or clogged fuel filter. If your Oilhead runs OK at idle and at modest throttle openings and loads, but stutters and bogs down at higher throttle setting, suspect your fuel filter.  If this happens to you on the road be very careful. Run it only where it is happy, even if it means a very slow trip into town, or a tow. Forcing it to run where it is very lean due to a fuel constriction can burn valves or the pistons very quickly. That loud bang can be very expensive. A tow is a lot cheaper.

            I can’t omit mentioning water in the fuel at this point. Water droplets sometimes actually clog injectors due to surface tension. Or what gets squirted is a little gas and a lot of water. In either case the engine is not getting sufficient fuel.  It will run poorly, or not at all. If you encounter these symptoms shortly following a fuel stop, suspect water. Add some Heet brand or other good alcohol based “gas line antifreeze” or fuel dryer. I actually carry a bottle in the bike at all times while traveling. Heet brand “IsoHeet” (isopropyl alcohol) in the red bottle is my favorite but the stuff in the yellow bottle (methanol) works, too.

            Sometimes a hose at the pump or filter inside the fuel tank will pop off causing no fuel pressure at all but this one is easy to spot. It is possible for a faulty fuel pressure regulator to cause pressure changes or fluctuations but this is not a common problem. If you are having a lean or rich condition and other causes such as a clogged filter have been eliminated as possibilities, then checking, or having the fuel pressure checked is a good idea.

            Finally, there is at least a remote possibility that a faulty fuel pump might be the problem. I say remote possibility because due to the design of the system, fuel pump problems are typically an all-or-nothing situation. Generally the pump runs or it doesn’t. To understand why look at the system. The fuel pump runs. It forces fuel to the pressure regulator. Fuel at the regulated pressure of about 40 psi is delivered to the injectors while the excess fuel is circulated back to the fuel tank through the return line from the regulator.

            Unregulated, the fuel pump can deliver fuel at a pressure in excess of 100 psi. I once discovered this the hard way, as I saw the needle on my pressure gauge peg at 125 psi just before the gauge shot out of the end of the hose, striking the wall about 20 feet away, and breaking. So a fuel pump can be fairly well worn before its pressure drops so low as to affect the regulated pressure of the fuel delivered to the injectors.

            Nothing is more critical to combustion in an internal combustion engine than the air-fuel mixture actually being ignited so it can combust—which brings us to the ignition system. In my experience, the ignition system on Oilhead motorcycles is a very robust and trouble free system —with one notable exception. The system operates as follows.

            The crankshaft position is reported to the ECY by the crankshaft position sensor —usually called the “Hall sensor” named for Mr. Hall, its inventor. A slot in the pulley on the end of the crankshaft that drives the alternator belt passes between a magnet and a detector. When the body of the pulley is between the magnet and the detector nothing happens but as soon as the slot passes—so there is no metal in between them—the sensor sends a pulse to the ECU. There are two sensors—one at top dead center and one at bottom dead center. The ECU uses these pulses to determine crankshaft position and to compute RPM, fuel injection, and ignition timing—when to spark the spark plugs.

            Current flows through the primary windings in the coils, controlled by the ECU. As current flows a magnetic field builds up in the coils. This is most of the time. Based on the signal from the Hall sensor, at the correct time the ECU turns the current to the coils off. As soon as the current is turned off the magnetic field collapses, inducing current in the secondary windings in the coils. The secondary windings are connected to the spark plug wires so a high voltage surge of current is conducted to the spark plugs. The spark jumps the gap from the main electrode to a ground electrode. The spark ignites the fuel-air mixture and we have combustion. Meanwhile the current through the primary windings in the coils has been turned back on to build up the magnetic field in time for the next spark needed.

            Spark plug wires do age and can deteriorate to where they leak electricity. Coils can go bad from heat or mechanical shock. Generally these are trouble free components.

            BMW specifies that spark plugs should be replaced at 12,000 mile intervals. But often when inspected at that interval the plugs look perfectly fine. So owners get casual about replacing spark plugs and eventually the electrodes get worn, the gaps increase, and the spark becomes erratic.

            More interestingly, in experimenting with how to reduce surging, Rob Lentini found that using a single electrode spark plug—specifically Autolite 3923 spark plugs—reduced surging. At the time he disclosed those results some folks refused to believe that changing a spark plug could reduce surging. As an old gear head, I believed it. Good combustion requires complete combustion and the spark plug can make a difference. Back in my gear head high school days I experimented with “indexing” spark plugs in car engines. That’s what the hot rod magazines said we should do. Indexing is accomplished by installing the spark plug so that the ground electrode is located opposite and away from the intake valve(s) so that the spark is completely exposed to the incoming fuel charge. The rotary position of the spark plug can be controlled by varying the thickness of the crush washer on the plug.

            So in the case of the BMW Oilheads, using a single ground electrode plug instead of the two or three ground electrode plugs originally specified and supplied, minimized the likelihood that the ground electrode position masked the spark from the fuel charge. However, unless you are experiencing combustion issues there is no need to consider changing spark plugs. If you are it is worth a try.

            BMW itself addressed one spark plug related combustion issue when they went to the heads with two spark plugs per cylinder. The Oilhead combustion chamber is semi-hemispheric in shape. The single spark plug models have the spark plug located near one edge of the combustion chamber. When the air-fuel mixture is ignited by the single spark plug the flame front needs to burn from its ignition point all the way across the combustion chamber to the far side. In the dual spark plug twin spark engines the fuel-air mixture burns from the two ignition points till the flame fronts meet near the center of the combustion chamber. This results in quicker and more complete combustion.

            The one element of the ignition system that has proven to be problematic on Oilheads—which also affects fuel delivery—is the crankshaft position (Hall effect) sensor assembly. In most cases the problem is not the sensors themselves but rather the problem is the wiring bundle from the sensors to the connection to the ECU. The wires attached directly to the sensors are only about two inches long. They seem to be high quality high temperature wire. The wires are crimp connected onto four longer wires that run inside a woven shield and outer bundle cover to the connection to the ECU, about two feet long or so.

            The insulation on these four wires, inside the bundle, has been found to deteriorate. The insulation dries, cracks, and shrinks. This exposes bare copper wire at various points along the bundle. The damage is most pronounced where the bundle runs beneath the engine front cover and is exposed to engine heat. The damage is less often found where the bundle is located out from behind the engine front cover.

            When the insulation deteriorates engine operation can be very erratic. Shorting occurs from wire to wire in the bundle. This can result in the engine not starting or operating at all. Or, it can result in the engine running, but doing so very poorly. If the shorting is from the hot wire to the signal wire from the top sensor back to the ECU, every spark tells the computer that the crankshaft is at TDC. The ECU uses these erratic signals to determine when to fire the injectors and spark plugs. Thus the injectors can squirt and the spark plugs can fire at random intervals whether the engine crankshaft is turning or not. As in many cases electrical, these conditions can appear when moisture has entered the wiring bundle and might not be present when the interior of the bundle is totally dry.

            The cure for this defect is to either replace the entire crankshaft position sensor assembly (plate, sensors, wiring harness, and connector) or to strip the bundle apart and replace the deteriorated wiring with good high-temperature wire. If you are on the road, riding in rain, and symptoms occur, you may be able to cure it long enough to get home by making an opening in the outer insulation of the whole bundle near the top, under the tank, and spraying an alcohol based electrical or electronic cleaner inside the bundle. Water conducts electricity. Alcohol doesn’t. Alcohol absorbs water so the alcohol can dry it out as far as the electrons are concerned. You can accomplish substantially the same result by flooding the bundle with WD40. Once the symptoms occur you need to replace or repair the damaged assembly.

            While on the topic of ignition, I need to mention pre-ignition and detonation. These are names for conditions where the fuel-air mixture is ignited by factors other than the sparking of the spark plug(s). The good physical condition of the engine depends on ignition and combustion occurring at the proper time. Several factors can lead to pre-ignition or detonation, but the two primary causes are the use of fuel with too low an octane rating and the build up of carbon and creation of glowing hot spots within the combustion chamber.

            You should always use fuel with at least as high an octane rating as specified for your motorcycle by BMW. Generally using anything lower can be harmful and using anything higher isn’t necessary. If you have significant carbon deposits inside the combustion chamber they can raise the compression in the engine. In these cases the proper procedure is to physically remove the carbon deposits but using a higher octane fuel in the short term can prevent engine damage. Do not assume that if you can’t hear pinging it isn’t happening. It might be. And pinging in an engine is akin to hitting internal parts with a ball peen hammer. Valves and pistons don’t like it at all and sometimes break.

            This pretty well covers the common combustion problems in Oilhead engines. There are others, like low compression from burned valves or worn rings. These are far less common than those discussed here. If you pay attention to how your motorcycle starts, runs, and performs, you will usually detect most common combustion problems.

Good Wrenching!