Sometime in 2004 I found my first e28. This is a four door sedan made by BMW from 1982 through 1988. The common models available in the US had either a 2.7, 3.2, or 3.4 liter inline six cylinder engine. This one was listed for sale on the Roadfly message board in the classified section. This was a nationwide forum, but this car happened to be located only a few miles away from me. It was a 1985 528e in Alpine White, with a pearl-beige interior and a 5-speed manual transmission. The owner listed it in non-running condition, so upon paying $1300 for it we towed it home. From a storage unit off of Highway 528, we pulled it with my Dad’s Nissan pickup and a tow strap all the way back to the house. With some basic troubleshooting it was soon determined that something had gone wrong in cylinder number 5, as the spark plug showed substantial physical damage that was indicative of ingestion of some sort of foreign debris. This required a nearly full disassembly of the engine to inspect the cylinder head and bottom end. The head showed only superficial damage, so with not much more than the bare engine block left in the car, we began reassembly. One piston was replaced, the bottom end was buttoned up, the cylinder head was reinstalled as well as the manifolds, wiring and fuel system connections.
Upon initial start-up, the oil pump did not prime and the engine ran for nearly a minute with no oil pressure. Once we realized that this situation was not going to remedy itself, the oil pan was removed, the oil pump inspected, and the pan reinstalled. The second time was the charm, as the oil pump began to pump oil and the engine ran once again. The car had somewhere in the vicinity of a quarter million miles on it, though most of them were done on the highway. Overall, it was in good condition cosmetically but it was one of the slowest cars on the road. This model was never intended to generate excitement, it was mildly tuned from the factory and produced a very modest 121 horsepower at the crank, at sea level. At my altitude, around 5500 feet above sea level, it was producing about 102 horsepower and perhaps 85 of those were making it to the rear wheels.
For the sake of comparison, I currently own a 2002 Yamaha FZ1, a 1000cc sport touring motorcycle that makes 130 horsepower at the rear wheel, and probably about 107 at my altitude. Comparing these two in terms of their power to weight ratio, the e28 carries 35 pounds for each one of its horsepower and the FZ1 has less than five. On the road, this means one of these vehicles will run a standing start quarter mile in 10.62 seconds at 130 miles and hour, and the other clears the quarter in about 19 seconds with a trap speed of about 80 miles per hour. Zero to sixty times are another common benchmark for performance, and here again the FZ1 gets the job done in under three seconds while the 1985 four door sedan will struggle to get to 60 in less than ten. In the simplest terms, this car was a slug.
As I imagine is typical, the ordinary young man has a strong interest in making his car go faster and I was no exception. I had been riding bicycles since I was very young and loved going fast, riding wheelies, and hitting jumps. I began riding motorcycles when I was about ten years old, and those went even faster still. I rode motocross at a competitive level into my high school years and was accustomed to high performance machines. The car did not fit in with my sensation seeking nature, and I very quickly began looking into what could be done to improve its performance. People on the internet suggested a performance chip for the computer and a lower final drive ratio, giving the engine more mechanical advantage over the weight of the vehicle. These modifications were done without delay, but the car was still dreadfully slow. One of the worst things I have ever done to a car followed shortly thereafter, which was the removal of the original exhaust system and the installation of glass-pack mufflers. I never liked the sound, which was loud and awful to an extent beyond my ability to describe, but I had hoped to increase performance with a minimally restrictive exhaust system.
The results were not outstanding. If the car was any faster it was only by an imperceptible amount. The noise was horrendous, which reverberated and droned inside the cabin in a way that may have sounded even worse than it did to those outside the car. Not to be deterred, my attempts at finding horsepower continued. I installed a larger air flow meter, larger fuel injectors, a larger throttle body. None of this made any difference, other than to make the car less fuel efficient and less reliable. Eventually, I decided it was time to do some more substantial engine work. I found a camshaft from an M20B25 and a cylinder head from a 323i, both being different variations of this engine in a more performance-oriented state of tune. This camshaft and cylinder head were installed and the engine actually responded with an increase in power. Still, not a fast car by any means but I was moving in the right direction.
The next step forward was significant but quite difficult. The original engine management system was not calibrated to work with anything other than the standard M20B27 engine. Upon installation of the 2.5 liter cam, the engine dynamics had been changed significantly and the computer was now a significant obstacle to progress. I made a feeble attempt to explore EPROM tuning, which I had no chance of accomplishing anything with. What I needed to do was install the 2.5 liter engine management system, which meant re-wiring a harness from a later 3 series and splicing it into a 5 series engine compartment and the existing wiring around it’s periphery. Many of the sensors were different, so these needed to be changed over as well.
Not to be deterred, I began scouring through the BMW ETM. This is the electrical troubleshooting manual which contains comprehensive wiring diagrams for both the donor car and my 5 series. The main obstacle was the C101 junction at the fuse box, through which many of the critical connections were carried. About 16 wires needed to be spliced, and since the connector was a different type on the 3 series harness the e28 connector had to be carefully grafted on. Various other changes were required, mostly wiring extensions, but the work was soon completed. With the homemade wiring harness and all of the correct sensors and switches installed, the car did not start. I was disappointed, but not too surprised.
Looking back over the ETM, I checked all of my work and verified all of the connections. Everything I had done was correct, but the car did not start.
I considered this situation over the following days. Looking back over the C101 connections, I realized that there was nothing here that connects the DME to the ignition switch. This seemed like it would be necessary. How would the ignition key turn the engine off if there were no connection between the two? I looked at that section of the ETM and quickly discovered the culprit, a few wires going through a separate and not terribly important looking connector in the vicinity of the glovebox. This C104 junction carried the very important start/run signal to the computer, as well as the tachometer signal and a fuel consumption information to the instrument cluster.
A few quick splices here and the engine started immediately. The new engine management system was a vast improvement over the previous iteration. The improved idle control system was immediately apparent, and it also had advanced adaptive capabilities to adjust fueling for maximum efficiency. Best of all, the fuel and ignition mapping were suited to work with the camshaft and intake manifold I had previously installed. The car was running quite a bit better by this point, although still unacceptably loud.
The next few steps were more of the same. I installed some low quality “shorty” headers. I ordered a reground camshaft with more lift and duration. I sourced an intake manifold from a 323i on German eBay and had it shipped in. I found a chip from an Alpina C2 2.7, which was a German builder who specialized in tuning these cars. There couldn’t have been more than a few hundred of these made originally, but I found a chip out of one of those and plugged it into my DME. It provided more advanced ignition timing and a fuel map suited for a more aggressive camshaft profile.
At one point during the summer of 2006 while I was working at BMW, one of my co-workers informed me that his dental hygienist mentioned that she had an old BMW she was looking to get rid of. Vinny knew this was right up my alley, so he passed her information along and I got in touch. The car was located in a barn in Tome, NM and had been sitting for a very long time. The owner indicated that there was a problem with the alternator, but overall the car was a viable candidate. Being a 1988, it came with a number of desirable parts, including a factory installed version of the fuel injection system I had just painstakingly wired into my ’85. The price was $500, and I was shortly on my way back down to get it. The plan was to drive down in my ’85, put the good battery from my car into the ’88, the dead battery from the ’88 into the ’85, start the ’88 with the good battery and dead alternator, jump start the ’85 with the good alternator and dead battery, and drive both cars home about 45 miles before the battery in the ’88 ran down. The drive was about an hour, 45 miles distance. My buddy Brain soon discovered that the brakes didn’t work very well and nearly rolled out in front of a passing truck on highway 47. After that though, we had no issues. Both cars made it home and my second e28 project had begun.
Fortunately for myself, the neighbors, and the rest of Rio Rancho, this ‘88 had a factory dual exhaust system which was suitable for the level of horsepower I was making. Some welding was required to repair the center resonator but the system in general was solid. The glasspacks were unceremoniously shit-canned and the complete, stock ‘88 exhaust system was installed on my ‘85. The car seemed to run better than ever, while being quiet and pleasant to drive. Meanwhile, the ’88 was in the process of being comprehensively stripped down. The interior was ruined with rodent activity, so it was gutted and new carpet, seats, and dashboard parts were ordered. The engine and transmission were removed for a rebuild. This car was originally an automatic which I had no plans on retaining, so the process of converting the driveline to a manual transmission was started. This was quite easy with the engine and everything else out of the way.
At this point with the original Alpine White ’85, I had essentially reached the end of what was achievable within the limitations of this engine’s displacement and design characteristics. There were no stones left unturned, no other parts left to try. The only ways to make more power were to increase the size of the engine or spin it at higher RPM. There are practical limitations to both approaches. For the ’88, a new plan was devised using forced induction to break out of the confines of the 2.7 liter, single overhead cam architecture. Forced induction utilizes a compressor, either a belt driven supercharger or exhaust gas driven turbocharger, to supply more densely compressed air in high volume to the engine. This pushes more air mass into the existing engine and correspondingly more oxygen into the combustion chambers. Atmospheric pressure at my altitude is about 12.2 pounds per square inch. If I were to add 12.2 pounds of boost with a turbocharger, it should in theory double the amount of power the engine produces.
In practice, it doesn’t quite work out that well. The turbocharger presents a restriction in the path of exhaust gases trying to egress from the combustion chambers. This suppresses volumetric efficiency by leaving excessive residual pressure behind after the piston attempts to “blow down” the cylinder on the exhaust stroke. This residual gas remains behind in the cylinder and occupies a certain volume that cannot be filled by fresh fuel and air on the following intake stroke. If a belt driven supercharger is used, the power to drive the compressor is subtracted directly from the crankshaft. Neither approach is perfectly efficient, but both will make power.
I decided a turbocharger was the way to go. I spent countless hours reading everything available on the subject. I pored over technical information available from turbocharger manufacturers. Honeywell Garrett had an excellent website with comprehensive information regarding compressor characteristics and equations that could be used to determine what kind of air mass would be required by any given engine at any given boost level. My calculations determined that a GT2560R compressor would likely be suitable for my engine at my elevation, so I bought one, part number 466541-4. Originally, I didn’t plan on high boost levels or outrageous performance, so this turbocharger being capable of about 300 horsepower was the ideal choice.
The first problem was, how to install this device on a BMW M20 engine? This engine was never designed for turbocharging and all I have available are the various exhaust manifolds BMW manufactured for this engine. A turbocharger is generally installed close to the engine, though it can be installed practically anywhere you can fit it. They have been installed in place of rear mufflers, sticking out of hoods, inside passenger compartments, and in spare tire wells just to name a few options. I considered a rear mounted system for the sake of room, but I didn’t want to deal with feeding and returning pressurized oil to and from the back of the car. This Garrett turbo was water cooled as well, so it would also require a pair of coolant hoses running to and from where it was located. The oil return would require an electric pump, which is a potential point of failure along with the many feet of plumbing to carry these vital fluids.
After some research and consideration, I decided a conventional installation in the engine compartment was the way to go. BMW had made a diesel version of the engine with a turbocharger, and the manifold could be modified to fit the gas engine. I ordered a manifold and began work on modifying it. The manifold itself was close enough to work, only requiring a few hours of drilling, grinding and notching to fit the stud pattern on the gas engine. I found the cross-sectional area of the collector to be smaller than I liked, so I ground this out until it was significantly larger than the turbine that it fed into.
The ’85 528 was still on daily driver duty while the ’88 project progressed. By this time, it had accrued more than 314 thousand miles, the most recent of which under my ownership had seen a lifetime worth of what some might consider abuse. It was modified and making more power than ever, and being driven in a way that I’m certain it had not been accustomed to. Eventually, this would come to an end. On my way to work one morning, it began to emit a rapid, metallic, cyclical tapping noise at moderate RPM under light load. This was the sound of rod knock, which occurs when a connecting rod bearing can bear no more. The bearing had “spun” inside of the connecting rod and had opened up too great a clearance for the hydrodynamic film to carry and cushion the reciprocating load. The engine was finished. I made it to work, then back home again as carefully as possible, and the car never drove again after that.
This situation was problematic, I still had to get to work. I did have some vacation time so I used that to put in some work on the ’88. The interior had been refurbished and only the driveline remained to be addressed. I already had that engine stripped down, cleaned up, and I had parts on hand to rebuild it from a bare block. Starting with the main bearing shells, then the crankshaft, then the main bearing caps, then the rods and pistons. The front timing cover, rear main seal carrier, the oil pump and oil pan. The short block was done in an afternoon. The cylinder head had been assembled previously and was ready to install. The ’88 had a different cylinder head with larger valves and improved combustion chambers, suited to fit the hemispherical dish-in-a-dome pistons. A significant improvement over the ’85 engine, even before forced induction.
Over the course of that week, I finished assembly of the engine, installed the engine in the car, installed the transmission, driveshaft, wiring harness, reinstalled the ’88 exhaust from the deceased ’85 and got the car running before the weekend. This was a car built from a gutted, empty shell. It had its own sort of new car smell. It had a new interior, with seats I had upholstered, new carpet and new trim. The scent of the simple green I used to clean the rodent urine from the sheet metal stampings of the floor may have been faintly present. One of the last things I needed to do was bleed the brakes, which were still full of original 1988 vintage DOT 3 fluid. It was semi congealed into a lumpy, slimy broth. No wonder the brakes didn’t work.
At this point, it was on the road but not yet turbocharged. The timeline of the project had been disrupted by the untimely mechanical failure of the ’85. In any case, it was on the road and it ran beautifully. The engine was new, cylinder and valve leakage was near zero. The valvetrain was new, the camshaft was new, everything that mileage and use subjects to wear was fresh and in perfect condition. Power balance across all six cylinders was optimal. It delivered strong, smooth power over a broad range. This was a satisfactory result for the substantial work that had gone into the ’88 up to this point. With my transportation needs met, work on the turbo project continued.
I had begun to accumulate a sizable collection of spare engine parts, including a cylinder head I used to test fitment and mockup the turbo and manifold. The GT2560R used a larger compressor than BMW had fitted on the diesel engine, which would not clear the valve cover when installed. A simple but inelegant solution to this was a stack of at least seven or eight T25 flanges between the manifold and turbocharger, raising the turbo high enough to clear the cylinder head and bolt everything together. I ran this setup for probably a year and a half, having to periodically tighten the very long bolts used to hold the large stack of flanges together. Exhaust leaks were constant but with the volume of flow through the engine it had little effect on the operation of the turbocharger.
Next up was plumbing. Turbochargers draw atmosphere through a short inlet system to filter undesirable particles and deliver this under pressure through a length of plumbing. This reaches from where the turbo was located all the way to where the throttle body admits air into the intake manifold. Any self-respecting builder will certainly include an intercooler to get rid of the heat generated by the compression of the air through the turbocharger, which lengthens and further complicates the plumbing. I started with a small front mounted air to air intercooler, located below the air conditioning condenser under the front bumper. I was still considering trying to retain the functionality of the air conditioner at this point, so a certain trade off was required at the expense of intercooling.
The ’88 was now the daily driver, but I was able to begin to fabricate and assemble the plumbing in a piecemeal manner. The intercooler was installed on improvised brackets and the plumbing worked outward in both directions. The compressor outlet was on the passenger side of the engine compartment and the plumbing was constructed to feed downward and curve through the bodywork in a horseshoe shape around each side, where it connected to the intercooler in front. The throttle body was on the driver’s side, where the plumbing came up just inside the frame rail and fed into the air flow meter, which the computer uses to determine the load on the engine.
This was probably the most significant obstacle to turbocharging this engine. The turbo can supply a massive amount of air mass to the engine, however the stoichiometry of combustion requires a correspondingly large amount of fuel to be delivered to maintain an appropriate air/fuel ratio. The fuel injection system was never designed to work with anything other than atmosphere, so this required some engineering. At the time in 2006, options were limited. The best solution at the time was a rising rate fuel pressure regulator. This simple device references fuel pressure to boost pressure, and scales up fuel pressure at an adjustable rate of gain to deliver enrichment under boost. The computer does not see the boost or any change in load and the engine still displaces 2.7 liters. The RRFPR simply overrides the standard fuel pressure regulator when boost pressure is generated. The computer fires the fuel injectors for the exact same duration that it would ordinarily, but with the increased fuel pressure additional fuel is delivered.
This completely resolves the enrichment issue, but doesn’t completely address the obstacle posed by the fuel injection system. The other factor must be addressed is ignition timing, which is advanced at lower loads and pulled back at higher loads. The original computer handles this assuming maximum load is at atmospheric pressure. With forced induction, maximum load is determined by the maximum boost pressure and ignition timing should be scaled downward further as boost increases. If the air fuel ratio or ignition timing is not carefully adjusted, abnormal combustion occurs known as “detonation” which is a spontaneous, powerful and rapid ignition of mixture inside the combustion chamber. This phenomenon is very destructive and is the most common cause of engine damage in high performance engines, especially those using forced induction.
Fortunately in a way, and unbeknownst to me at the time, my intercooler was significantly undersized. It was unable to control charge air temperatures effectively, so the air flow meter was seeing significantly elevated air intake temperatures. The original computer responds to this by pulling back ignition timing to avoid detonation, which is exactly what was needed to save my bacon. I’m not sure to what extent this was beneficial overall, as these same elevated air intake temperatures contribute to detonation as this heat enters the combustion chamber and exacerbates the adiabatic heating that occurs during the compression stroke.
In any case, with the fuel enrichment handled and a certain amount of indifference toward my inability to address ignition timing, I moved on with the project. The turbocharger and exhaust manifold were ready, I had flanges, bends and tubing to fabricate a new exhaust system. I had the intake plumbing essentially competed. I had routed coolant from the original throttle body heater circuit to the turbocharger. I tapped into the clean, pressurized supply of oil available at the oil cooler adapter and installed a fitting here to supply the turbocharger. The oil return was plugged haphazardly into the oil pan and sealed with the finest quality high strength, steel reinforced epoxy.
In spring 2007, the ’88 was ready for boost. The installation did not require vacation time as all of the loose ends and potential hang-ups had been sorted out in advance. The manifold had been test fitted, the plumbing had been fabricated and was already almost completely installed before I installed the business end of the system. I would still need to sort out the rate of gain on the RRFPR to dial in the fueling, but this was a simple device with few adjustments. Cautiously at first, the original turbo 528e hit the road. The tuning was simple enough, I had previously spent many hours adjusting carburetors on the dirt bikes to fine tune fuel mixtures. This experience translated well to the turbo car. I started with excessively rich, or fuel heavy air fuel mixtures. As I leaned the mixture out, the car stopped breaking up and misfiring and began making what felt like a ridiculous amount of horsepower.
At that boost level, I would estimate the car was making about 200 rear wheel horsepower. At the crank, close to 240. This car was quick even at 7 psi. For this altitude, anything without a turbo on it is paying a roughly 20% penalty with the reduced air density. A stock LS1 in a Camaro SS or Trans AM WS6 was making 300 horsepower at the crank, at sea level. In Rio Rancho at the rear wheels? Maybe 205. I recall a specific instance where I lined up with one of those LS-equipped Pontiacs, and we both opened it up when the light turned green. Though both cars were accelerating rapidly, we were completely stationary relative to one another. Despite the vastly different powertrains and 20+ year difference between these cars, the different gear ratios and torque curves, we were accelerating at exactly the same rate. I think we let off at about 100, which was fortunate as there was a police officer sitting in a Target parking lot off to the side, perhaps another quarter mile down the road. We would have been going more than triple the posted limit by that point, and doing so while street racing. That could have been a bad night, but was the first of many such close calls.
I would have loved to talk to the guy in the Pontiac. I couldn’t believe I hung with a car like that, and I’m sure they were just as surprised as I was. This was the most awesome thing about that e28. Dragged out of a barn, covered in rodent excrement. Resurrected, at least mechanically, and modified to produce enough power to stand toe to toe with the latest iteration of GM’s high performance 5.7 liter V8. Though the interior was nice, the paint and body on the ’88 was not on the same level as the ’85. Even when new, it was painted a butt-ugly metallic brown. Nerzbraun, if I remember correctly, which translates to “Mink Brown”. The paint was faded and the clear coat had begun to peel, so it wasn’t much to look at in 2007. Worst of all, the car had been hit and damaged in a rear end accident shortly after it was returned to the road. The damage had affected the right rear quarter panel and broken the rear bumper in half. This pissed me off severely but it didn’t look too bad once I had replaced the taillight. The car had been paid out as a total loss, so I had more money for turbo parts. This disguise added to the shock and dismay of it’s victims. You would never see it coming, and it was a serious contender.
Shortly after the system was completed, the ’88 went to the dragstrip for a sort of performance evaluation. This provides a substantive basis for comparison, the sum of all factors. Some aspect of this is left to the ability of the driver but for the most part, the quarter mile run provides a very good test. Launching this car was difficult, even with relatively mild amounts of horsepower. This was the furthest thing from this vehicle’s intended purpose, and the power output of the original engine had been more than doubled. Despite uniformly poor results in the first 60 feet, during the period between not moving and moving, the car covered the distance in very close to 14 seconds flat. Better launches and a lower 60 foot time would have provided better ET’s, but a 14 flat is nothing to sneeze at. Trap speeds were around 100 miles an hour, which roughly corresponds to the elapsed time.
The thrill and euphoria of these things is unfortunately short lived. Given a few weeks of excitement, the newfound performance became familiar and increasingly ordinary. Still a young man, always looking for more, I began exploring the possibility of turning the boost up beyond the original .5 bar allowed by the wastegate actuator. The system was already marginal in many ways, and I bumped against these limitations as soon as I tried to push more boost. The original computer knew nothing of the hot rod shenanigans that it was involved with, and was increasingly unsuitable for boost as it crept higher. I found a custom chip with moderated ignition timing through Todd at TCD, a supplier who had recently brought legitimate turbo systems to market for e28’s. This got me to 11 psi, which made the car go faster still. I had to resort to zip ties and duct tape between the air flow meter and throttle body though, as the original rubber boot was never designed to hold even a strong vacuum, let alone 11 pounds of boost. Fuel pressure was reaching levels more than double what the system was intended to run, which was soon going to require more fundamental changes.
This was about when the car went to the dyno for the first time. Quoting myself from forum posts of the period, I was running 11 psi and the car made 230 rear wheel horsepower and 266 foot-pounds of torque. This would perform comparably to a contemporary BMW M3, whose S54B32 engine produced 333 horsepower. At my elevation, 333 crank horsepower translates to 227 at the wheels. And my e28 was lighter. My only gripe at the time was the relative lack of top end power, the dyno curve began to drop excessively as revs passed 5500 RPM. Peak horsepower was lower than would be expected for this engine configuration, so something was suppressing the last third of the power curve. In hindsight, I think this can be attributed to the insufficient intercooling and rapidly climbing intake air temps. Not only does this elevate the probability of detonation, the hot boost is less dense than it could be, as the overall air mass delivered for a given volume of air is reduced.
I think the highest I ever went with the original system was 14 psi, and this was more than the original clutch could handle. I had to turn it back down immediately and start shopping for a stronger clutch. Todd at TCD had a solution for me for a few hundred dollars. This was an organic disc with a special pressure plate designed to generate lots of clamping force. Designed to hold almost 400 foot-pounds of torque. That seemed sufficient for any prospective boost levels, so I ordered one up.
The next step for power was a bigger turbo and standalone fuel injection. I was not at all concerned about a bigger turbocharger, this is a bolt on affair. My hypothesis was that the GT2560R was restrictive on the turbine side, which presents a choke point for the exhaust gases on their way out from the combustion chamber. Every pound of air that goes through the engine has to squeeze through this opening in the turbine housing that was approximately the size of a large grape. 230 rear wheel horsepower requires about 23 pounds of air per minute, which is a lot of air. A larger turbine would trade off some of the response and raise the boost threshold, which is the RPM which the turbo fully spools and makes significant boost. Regardless, this was the next step, so the turbo was ordered, a Garrett GT2871R .86, part number 472560-15. Part of the appeal of the GT2560R was that it shared almost all of it’s design features with a family of other turbochargers. Garrett described these as “outline interchangeable”, meaning I could bolt on a larger unit without having to weld on new flanges or change plumbing.
The real battle was shaping up for fuel injection. My ability to work around the limitations of Bosch Motronic was at its end. I had done all I could but it was simply no longer possible to make any more power without an unacceptable risk of engine damage. The solution was a product called Megasquirt, a standalone computer that replaces the original engine management computer and offers full control over every parameter of the engine’s operation. Tuning can be performed while you drive with a laptop computer. This sounds great, but it also means that you are responsible for configuring and tuning every single parameter of the engine’s operation. I wanted more power though, so I found a guy in the e28 community who was building these MS2 kits into the original Bosch Motronic boxes. This allows a plug and play functionality, which retained the ability to plug the original computer back in if need arose and I had to get the car home.
At the time, this was the most significant challenge of the project. This was far above and beyond fabrication and designing a layout. I had read a great deal about tuning, and I had a lot of experience with carbs on two stroke engines. I had no experience building three dimensional maps for fuel and ignition timing though, let alone tuning acceleration enrichments, cranking pulsewidth, configuring ignition settings or any of the other 100+ variables that I was now trying to sort out. It could be compared to teaching someone how to put on shoes, stand up, walk, run, jump, then compete at an Olympic level. This is a process that stretches into the hundreds of hours.
The good news in that the car did eventually get going on standalone, and the new turbo did make more power. Standalone injection offered many advantages once the initial period of configuration hell had been successfully navigated. I now had the ability to record datalogs containing a record of every sensor input and every output by the standalone computer. I had shortly discovered how severely undersized my intercooler was by observing the intake air temperature under boost. It climbed as rapidly as RPM, reaching levels that would rapidly defrost a breakfast sandwich. I immediately ordered another intercooler to install in series with the original. The internet said it couldn’t be done, that pressure drop would be too great and it would be horribly inefficient. I did it anyway. Pressure drop was not significantly greater than it was with the single front mounted intercooler and the intake air temps were within twenty degrees of ambient. A significant improvement in both performance and reliability.
The next trip to the dragstrip was later that summer. On this night, with more boost provided by the larger turbocharger and supported by the standalone fuel injection, the car was running 13 second quarter mile times at 108 miles an hour. Launching was still not great, and this was the only thing holding the car back from breaking into the 12 second category. I had to settle for a best ET of 13.01 at 108 miles an hour. The car got off the line better on the street. The surface at the dragstrip was a nearly smooth type of pavement with a coating of rubber laid down on top. With the right type of tire, this generates enough traction to destroy driveline parts and tear subframe and differential mounts right out of the unibody. With the type of tires I was running, I could only produce wheelspin at best. I rarely drove at the dragstrip and drag radials would almost certainly require a substantial reinforcement of the rear suspension, so this was fine with me.
The next trip to the dyno was in the fall of 2007, and the car made 300 rear wheel horsepower and 360 foot-pounds of torque. Would compare well on the streets of Albuquerque to a non-turbo engine that makes 450 crank horsepower at sea level. It did in fact compare well to a stock BMW 335i, the new twin turbo 3 series introduced in 2007. Another member of a local car forum had one of those and was left behind by bus-lengths while going head to head with the e28. Video of this evening still exists in the hallowed halls of YouTube. That was a fun night, and I engaged in this sort of high risk activity on a frequent basis.
One instance in which I wasn’t so lucky occurred on a lunch break, likely on my way to Arby’s for a three for five deal. I had come down Second Street and gotten on eastbound Paseo Del Norte, and seeing a clear road ahead for at least a half mile I put the hammer down. The turbo spooled up in third gear and the traffic that was once close behind diminished to distant specks in the rearview mirror. I had not checked my mirrors closely enough though, as there was a Bernalillo County Sherriff among that group of cars. Third gear would hit 120 and I’m pretty sure it did before I began coasting down toward Jefferson. About the time I got off the power, I observed the red and blue lights catching up from behind, and I pulled over on Jefferson for what was certain to be an unpleasant experience.
Got bored at work, wrote up my life story
Re: Got bored at work, wrote up my life story
Continued:
The officer was polite and professional. I provided my documents. He returned to my window and had me step out of the vehicle. He asked about the car. I briefly described what had been done to it, eager to spill the beans but averse to digging myself an even deeper hole. The officer seemed perhaps somewhat interested in high performance vehicles and even intrigued to some extent. He said that the fastest he was able to pace me was 77 miles an hour, so he wrote the ticket for that. We both knew I was going quite a bit faster and I certainly could have been in a much worse predicament. This was the only time I was ever caught having too much fun, which is truly incredible considering the frequency with which I indulged in this pleasure.
As 2007 neared its end, the summer of reckless good fun and pushing boundaries had eventually exacted some consequences. The head gasket I installed in late 2006 had been through hell, periods of severe detonation were inevitable during this process. My learning curve was steep and the margins of safety were diminished. I believe it was October when the head gasket had seen enough. Combustion gases had breached the sealing rings around the cylinders and begun to displace the liquid circulated through the cylinder head to wick away the wasted heat of combustion. This is not a failure of the same magnitude experienced by the ’85 about a year prior, but it required repair nonetheless. The top end of the engine was disassembled, the head gasket was replaced, and I was operating under the assumption that I was destined to be back on the road after the weekend.
Upon the initial test drive, problems arose. The coolant level was not stable, but no leaks were apparent. This is when I learned something very specific about this M20 engine. Coolant mixing with the oil in the crankcase is very common among head gasket failures in a wide variety of engines. The M20, on the other hand, does not mix these fluids unless it is through a crack in the cylinder head. Removal of the valve cover revealed a crack under the camshaft in the area of cylinder number 5. Time to remove the cylinder head again, not to replace the head gasket but the entire cylinder head. The original head that came with the car, from the barn in Tome, with close to 250 thousand miles on it, had made its last horsepower.
This was mostly just an inconvenience. It would have been nice to discover this crack before building the engine with a doomed cylinder head. A valuable lesson had been learned though. Next time I see coolant in the oil, just throw the whole head away and get another one. One would hope this was the last time I destroyed a cylinder head, but one would be disappointed in doing so.
By 2008, I was still tinkering. With the obstacle of fuel injection comprehensively addressed, I was only inhibited by the integrity of the engine and my ability to calibrate the tune to safely support any given boost level. The Megasquirt ECU could read up to 255 kPa of manifold pressure, which would be 172 kPa of boost above atmospheric. So, I could run up to about 1.7 bar of boost before the computer would reach the limit provided by the range of its sensors. This is 25 pounds of boost, which is very nearly as much as the pressure in most car tires. A hell of a lot to run through an engine designed to run on ordinary atmospheric pressure. I didn’t get there right away, it was always a methodical, incremental process.
The next major change was a different turbocharger and manifold from Todd at TCD. This was a manifold designed with a more performance-oriented application in mind. The manifold was still cast iron but used larger internal cross-sectional area and a larger overall volume. It also utilized an external wastegate, which offered improved control and easy adjustability of boost pressures. Todd had a turbo and manifold laying around that weren’t part of one of his complete kits, so I decided to give this system a try. The modified turbo diesel manifold was never ideal and it had the potential to be contributing to the head gasket destroying detonation I had learned to fear.
This system required some alteration to fit. A new exhaust system was fabricated with a larger diameter all the way back. The intake and charge plumbing were revised to connect to the new location of the compressor. The rest of the system was unaffected, so this was another weekend job. In conjunction with this kit, I decided to install a multi-layered steel head gasket. The stock gasket was susceptible to damage and this reinforcement seemed like a very good idea.
This system did offer more power but the tradeoff in response was substantial. The previous Garrett turbochargers used advanced, ball bearing cartridges and were remarkably responsive. The new turbo system used a journal bearing cartridge and could only be described as sleepy below the boost threshold. I could no longer put my foot into third gear at 3000 RPM and expect anything to happen. Once I got used to it, this was not a big deal. The car did go fast once it got going. Some tuning was required to suit the change in the engine’s compression ratio resulting from the new head gasket. Not only was it a multi-layer steel gasket, it was 1.4mm thicker than the original composite gasket. This lowers the compression ratio, which reduces the amount of heating that occurs during compression and improving the engine’s resistance to detonation.
This process should have been more meticulous than it was. I recall it was a Thursday evening and I had gone to a car meet. The car was sluggish and I had no patience for that, so I dialed up the ignition advance until it started making power. Throwing caution to the wind produced results, and there was a rematch between the previously mentioned 335i and the ’88 528. My buddy had modified his 335i to generate additional boost pressure and was eager to see how it compared to the turbo shitbox. I don’t recall to what extent he was aware that the e28 had been turned up, but I think he at least had some idea. The second matchup went down the same as the first, though the 335i made a better showing this time around. Trey’s hopes weren’t the only thing that got smoked that night, my high-performance clutch had met it’s match as well. It got the job done but it was slipping under power and could no longer be trusted.
Todd at TCD expressed some surprise that this unit had failed. I don’t know exactly how much torque it was being subjected to but it was operating close to its rated capacity. Upon disassembly I closely inspected the flywheel, which comprises one half of the clamping surface that the friction disc uses to transmit power. Here I discovered part of the issue, the face of the flywheel had been worn in a conical manner, and it tapered inward toward the center of the flywheel. Imagine a very shallow ice cream cone. The new clutch was a very exotic unit consisting of a segmented Kevlar disc. It would only be appropriate to install this clutch onto a brand-new flywheel assured to be machined to the optimal dimensions.
The new clutch was a strong unit, and I never again experienced slippage. This new turbo system was better at maintaining full boost pressure across the full RPM range. The car was going faster than ever. The next trip to the dragstrip produced an ET of 13 seconds flat, again, but with a trap speed of 114 miles per hour. This was not as fast as it should have been, and I was up against the limits of fuel delivery at the time with the size of fuel injector I was running. On this day, I must have been in a certain mood. I was going to turn it up until something happened, good or bad. Something did indeed happen. I don’t know how much boost it was making, the manual boost controllers are very touchy. Minor changes make big results. The clutch held, but the engine experienced a quarter mile of the most severe detonation I’ve ever heard. I did not lift but instead held the throttle open and ran it for what it was worth. The ET was not the best of the day, and the coolant light illuminating informed me that I had, at very least, blown that steel head gasket.
Sometimes these things happen. This should have been avoided, but it is what it is. The car was hurt but I was able to drive it back home from the dragstrip. It only ran slightly hot on the gauge as I neared my destination. Perhaps it wasn’t too bad. Upon my arrival at home, which was near Constitution and San Pedro at that time, I found signs of what I already knew was a severe and damaging event. There was again coolant in the oil. Head was cracked. Removing the intake manifold revealed a green/blue crust on the intake valves near the back of the engine. The head gasket was compromised to the extent that coolant was blowing into the combustion chamber and being drawn up into the intake tract by manifold vacuum. This was remarkable.
The tear down was completed and more parts were ordered. I was perhaps motivated to run this thing to destruction by a sense of dissatisfaction with its performance. With larger fuel injectors it might have safely made more power. I never really bothered to properly explore this avenue. Maybe I was bored and felt the need to buy an even larger turbo. That’s what I ended up doing. I didn’t really know where I was relative to the GT2560R I started with, or the GT2871R I had used prior to this current system. I know it was running a T04E 50 trim compressor which is certainly large enough. The turbine side was a bit of a mystery. The decision was made at that time to order a turbo that was, without a doubt, too large for this engine. In the future I was already planning to replace the M20B27 with something like an M52B28, a 24 valve engine from the family that replaced the older 12 valve design I was current working with. The Garrett GT3582R was more suitable for this M52, and it was again a ball bearing turbo. I longed for the responsiveness of the ball bearing units I had run previously.
This required a new manifold, this GT35R would not fit the bottom mount cast iron unit I was running with the T04E. At the time there were a large quantity of very cheaply made tubular units available on eBay, but they were supposed to fit the M20 engine and they did use a T3 turbine flange. The manifold was undoubtedly as crappy as expected when it arrived and required substantial reinforcement over the coming months as it would repeatedly crack, be repaired, and crack again. This manifold also had no provisions for an external wastegate, which was required as the GT35R had no internal wastegate. I added an appendage to the manifold, that could only be described as intestinal in it's appearance, which laid the external wastegate down under the turbine outlet. This was a rather slick matter of packaging, though it left little room for a wastegate outlet pipe to be connected. For now, this was satisfactory.
The thoroughly abused engine was repaired with a new cylinder head. This time I opted to use the original head gasket. My experience with the MLS gasket was not outstanding. Nothing is det proof, and the MLS was challenging to successfully seal up after installation. Overall I was not impressed. I did utilize stronger hardware, some quality head studs manufactured by Raceware. These were robust and reusable. They also required less torque to generate tremendous clamping force, on account of the finer thread pitch. These studs and the stock head gasket proved sufficient for all future purposes.
This system had a certain appeal for me. I had gotten back to an entirely homemade recipe. The TCD kit was nice but it wasn’t mine. I doubt I had achieved its full potential but I was ready for the next step either way. I also had a large turbocharger with known technical specifications. The turbine side of the GT35R was twice as large as the original GT2560R. I had come to understand over the course of this project that this had a substantial effect on power output and detonation resistance. The GT35R opened up more potential and greater efficiency at high boost pressures. Though the compressor side was oversized for the application it was still operating within its map at a reasonable level of efficiency.
On the road, this new system proved to be quite a performer. Spool characteristics were surprisingly good despite the oversized turbo. It would make some boost even at very low RPM. It didn’t really get going until about 3500 RPM, and then it got going all at once. The system exhibited boost creep as well, due to the compromised location of the wastegate. It was branched off of the collector just before the turbine inlet, and the turbine of the GT35R was so efficient that it presented an insufficient restriction to divert enough flow through the wastegate. It could control boost pressure to the set value up until about 5500 RPM, at which point boost would begin to climb. By 6000 RPM, it was making 1.2 bar of boost pressure, regardless of the springs installed in the wastegate. This worked out just fine in a way. Peak cylinder pressure in this engine with it’s current camshaft and intake manifold occurred around 5000 RPM. Beyond that point, additional boost could be safely dialed in without risking detonation. The additional boost pressure presented little additional risk of engine damage.
This system was run at a boost pressure of 1.4 bar for the majority of the time I owned the car. It would spool up so dramatically that the car would accelerate as if it were being “booted” down the road. Imagine a goliath-sized giant literally kicking the car forward. This was the sensation experienced in higher gears. In first, second and sometimes third, the tires would simply break traction and the engine would bounce off of the rev limiter as the car struggled to put the power down in a long, rolling burnout. Final drive gearing was very tall, second gear would clear 70 MPH. Shifting into third would generate either tremendous acceleration or a highway speed drift depending on the condition and quality of the pavement. Usually it hooked up though, and the car would move out into the triple digits in a manner that seemed to completely disregard the exponentially increasing drag and the nearly three thousand pound weight of this ridiculously overpowered automobile.
At this point it was 2009 and the car had become a monster. It was nearly impossible to launch. As soon as the boost came on it would just blow the tires away. On the next trip to the dyno, the car spun the tires on the rollers, requiring extra straps to get the tires to overcome the massive inertia of the dynamometer rather than convert horsepower into a quantity of tire smoke which was impossible to measure. The result was outstanding. Without any correction factors applied, it produced 384 rear wheel horsepower and 382 foot-pounds of torque. Between 3500 and 4000 RPM, torque output doubled as boost pressure jumped from the single digit range to the full 25 pounds of boost. Even at this altitude, this was 450 horsepower at the crank. Not bad at all for a 12 valve 2.7 liter inline six that left the factory making about one quarter of that. Best of all, at this altitude, this 384 rear wheel horsepower would be comparable to about 560 crank horsepower at sea level. This crusty, cosmetically challenged e28 was able to embarrass some very exotic cars. The only cause for concern on the day was the large puddle of oil under the dyno after the pull. It had been carried out by crankcase gases despite the best efforts of the crankcase ventilation system. I had not seen this in previous dyno runs, but I had noticed an increase in oil consumption. Perhaps this was the nature of running high boost through an M20. Who could be concerned with that though, the car was running great.
One such exotic car was encountered late at night one evening, going eastbound across the Rio Grande. This scenario was tilted very much in my favor. We were cruising at highway speeds, where gear ratios were tall enough that the tires only had to handle a fraction of the power and torque delivered in lower gears. The victim was a bright yellow Corvette, a C6 Z06. This was the top of the line variant of the Corvette, with a 7 liter V8 engine producing 505 horsepower and 400 foot-pounds of torque. Quite unlike mine, this car was designed, built and bred for high performance. I recognized the LED rings visible in its taillights, as well as the vacuum controlled butterfly valves in its exhaust outlets. The “Z06” badge on the front fender just behind the tire subtly indicated what its other features had already disclosed. This was General Motor’s fastest version of the Corvette currently available.
The challenge for me was trying to set the hook. My car was completely invisible to people like this. An old BMW, unassuming to say the least. I would usually spool the turbo up briefly and give them a quick “jump” to provoke their interest and communicate my intentions. Sometimes they would bite, sometimes they were far more responsible than I and opted to mind their business within the bounds of the posted speed limit. For the Vette, I showed him what third gear would do. He didn’t have any advance notice, and he was left far behind. This wasn’t a race, just an offer to engage in a bit of extralegal tomfoolery. He pursued, and I slowed down. Then he opened it up. I pursued, then surpassed the Corvette. We had the road completely to ourselves, and we went back and forth a few times before reaching the interstate. He had bought the fastest Corvette available at the time and it wasn’t fast enough. As we headed south on I-25, a police car appeared on the frontage road with it’s lights on, moving at about the same speed as us and in the same direction. Could have been a complete coincidence, as neither one of us ended up being pursued and we went our separate ways.
Nobody believed me about this encounter, but the math is very simple. I’m making 384 rear wheel horsepower, measured at this altitude. The Z06 is producing 505 at the crank at sea level, of which about 404 make it to our elevation, and after being transmitted through the driveline about 344 reach the rear wheels. I had a large advantage in torque, 382 foot-pounds at the wheels. The Z06 only had 272 left of its original factory rating of 400 foot-pounds. It would have been a battle at sea level but this guy, along with everyone else breathing the thin atmosphere, was missing about 20% of his get-up-and-go at our elevation. Even then, I had a substantial advantage in torque, and even turbocharged engines run more efficiently at sea level. The turbo doesn’t have to work as hard to make any given manifold pressure with denser air available at the compressor inlet, and it operates at a lower pressure ratio. All things considered, the Z06 would have almost certainly taken the L there too.
I never went back to the dragstrip. The car wasn’t suited for it. It was a street car and it did pretty well out there. Standing starts were fun but not particularly fast. I once got smoked, for the lack of a better word, by a stock e60 535i. We had lined up at a traffic light, and it was driven by some of my co-workers at BMW. They knew about the e28, and my buddy in the passenger seat was hyping it up. Said the car was sick or nasty or something along those lines. I tried to temper expectations, saying the car was “hard to launch”. I had handily beat-ass all over one of these twin turbo models previously, with the GT2871R at 16 psi, but now I was driving an entirely different car. The light turned green, the e60 hooked up and accelerated hard. I laid down rubber in a feeble attempt to translate horsepower into forward motion, but after spinning through first and second gear I had been left behind. Not my best work. Sometimes I missed the old days of having usable power. The more I built into this thing, the less of that I had.
I don’t remember the year, but at some point I found another e28. This one was a 1986 528e, sitting in a field in Corrales. Had a blown heater core and was just as ugly as the ’88, but with a straighter body. The price was right, another $500 car. I bypassed the heater core, dunked a battery in and drove it home. The transmission didn’t like to stay in fourth gear, couldn’t do a lot with that, but it did come with some other good parts. The clutch was brand new, although completely insufficient for the current power level. I planned to skeletonize the ’88 and swap everything into the shell of the ‘86. One issue that I couldn’t resolve with the ’88 was a matter of clearance on the left rear wheel. There had been some body damage at some point in the car’s life which had resulted in the inner wheel well being pushed in. They fixed the outside but not the inside, so it rubbed heavily on the 265/40R17 tires I was running in an attempt to improve traction.
This must have been a tremendous amount of work. I can’t imagine doing this again, though I thought nothing of it at the time. It did present an opportunity to go through this motor I had been beating the shit out of for the last three years. I had a spare engine block I could use to build a fresh engine and I had found a set of pistons on eBay for a mere $200. The seller had them listed at $600, but the seller accepted a substantially lower offer. I was delighted to have found them, and I’m sure they were amazed anyone out there was looking for them. This was a rare build. Few people were doing anything with the M20. Most of the builds were with the 24 valve M50/M52, such as the one I was planning on using, but with the results I had with the GT35R on the M20 I couldn’t justify it.
Upon disassembly of this motor, the original one that came with the ’88 from the barn in Tome, I found some very surprising things. Most of the connecting rod bolts were very loose. None had loosened completely, but they should have been much tighter. This engine had a rev limit of 4700 RPM originally. I was running it up to 7200 at times, and reciprocating loads increase exponentially with RPM. That could do it. I also recall one instance where I mis-shifted and went from redline in second to first gear. This should have resulted in a catastrophic, likely explosive engine failure. I don’t know how high it overrevved but it was higher than it ever went of its own accord. This was years ago though. It seemed implausible that the motor would have lasted that much longer afterward if it had overstressed the rod bolts at the time.
With the rod caps off I was able to pop the pistons out of the block. This is where I found the truly scandalous details. In the rear cylinders, 5 and 6, the piston ring lands were destroyed. On number 6 the piston rings themselves were broken into multiple pieces. These pieces were trapped in place only by the remaining upper half of the top ring land and the skirt of the piston. When did this happen? That would have been the last trip to the dragstrip, about 18 months prior. That would also explain the oil consumption and large puddle of oil under the dyno. Would have probably made a few more horsepower with six properly sealed combustion chambers but I suppose I’ll never know for sure. What was really amazing was that these were the original pistons that had accumulated 238 thousand miles before I even acquired the car. I must have put another 30 or 40 thousand miles on them since the rebuild in 2006. If not for the completely avoidable incident at the dragstrip in 2008, they would have withstood everything else I had thrown at them.
Fortunately, I had other cars to drive. I had purchased an e39 528i from one of our customers at the dealer. The car had been overheated, the Achilles heel of the M52TU and M54, and the engine had been damaged. I got it for $2000 with about 90 thousand miles on the odometer. In continuation of my tradition of finding some way to drive these things home, I tied off the radiator hoses to prevent coolant loss through the gaping hole in the radiator end tank and I drove it home. This was an interesting experience to observe the various stages of overheating as indicated by the e39 instrument cluster. These temperature gauges are “buffered”, which means the computer evaluates the coolant temperature signal and displays its own simplified version to the driver. Most drivers will only ever see the needle at the “cold” and “normal” positions. Beyond that, there is a position at three quarters, indicating elevated temperatures. Then the needle goes into the red, indicating overheating outside of any normal range of operating temperatures. Then an orange light illuminates next to the red zone, in case you didn’t notice the needle position. In its final stage, the needle moves beyond the red zone and the light turns off. In practice, any indication above normal indicates serious overheating and potential engine damage. If by any chance one wasn’t cooked before I brought it home, it certainly was now.
A replacement engine cost about $1350. All of the other things that should be replaced while the engine is out cost another thousand or so. I was working at the dealer at the time, where I took tremendous advantage of the employee discount on genuine BMW parts. I must have spent tens of thousands of dollars over the course of my employment there.
The engine was removed with the transmission left in the car. This e39 was a different animal than I was used to, not as pleasant to work on as the e28. Somehow, I got the replacement engine in and buttoned everything up. The car started and ran beautifully. This was certainly far nicer than what I was used to driving, although quite a bit more ordinary in terms of performance. It did have air conditioning, a contemporary interior and superior sound insulation. This thing was a nice commuter. A bit more practical and enjoyable than driving a turbo e28 with the temperament of Dr. Jekyll.
I would by lying if I said I didn’t consider installing a turbo M20 into the e39. I think I made the right decision though. I still had the e28, and the ’86 was fleshed out with the very nice interior pieces from the ’88, the subframe and suspension, and eventually a new engine and transmission. I had a used M20B25 bottom end that came from somewhere or another, and I thought it would be an interesting experiment to install that engine into the ’86 with the existing turbo system. I had always run a modified M20B27 with what was essentially an M20B25 top end, so this was an interesting experiment to me. The engine was assembled as is, without anything more than a new oil pan gasket, and installed in the car. It ran well, with perhaps a slight advantage in top end power. This is not out of the realm of possibility, considering the same intake manifold was used on the 2.5 as I was using on the 2.7. On the smaller engine, this manifold would make power to a higher RPM due to the length of it’s runners and their cross sectional area.
It also made more boost though, at one point pegging the manifold pressure sensor at 260kPa. This was .05 bar higher than it should have read, and it was seeing even higher pressure than that. In conjunction with the slightly higher compression ratio generated by the M20B25 bottom end, this was too much to ask from the head gasket. I don’t know how many I had been though by this time, but I pulled the top end off and replaced the head gasket, throwing the old one on top of the stack of the others I had hanging on a nail against the wall. About a three hour job these days. That evening, I finished assembly and cranked the engine which nearly fired on the first try. On the second try, the engine stopped abruptly and would not crank. On subsequent attempts, the engine freed itself and turned over, but with no compression and only an occasional thump of resistance. Something terrible had happened to this engine, and that was the last of the M20B25 experiment.
I never really figured out where this went wrong. It clearly lost time at some point, allowing the crankshaft to turn independently of the camshaft. Valves hit pistons, one valve broke off, flipped over and was punched through the crown of the piston below as it collided with the cylinder head. During the process of installing the timing belt, the engine must be turned over twice to ensure proper tension on the belt and check timing. It was timed properly at this point, but something happened afterward that destroyed the engine. Oh well.
The new engine was well on its way. The spare engine block I had, which interestingly enough came out of the ’83 528e I rode around in when I was a child, was bored to fit the Mahle pistons I scored on eBay. Considering what I had done to the crank from the ’88, I figured I would use the ’83 crankshaft as well. These are cast iron, and the common wisdom about them among the BMW community was that they were weak and would break if revved over 5500 RPM. I never experienced that, and it wasn’t for a lack of trying.
This engine was assembled in what was soon going to be my daughter’s bedroom and wheeled out of the house on a furniture dolly. I was actually trying to get the car ready for a buyer that was coming in out of Michigan. I had convinced myself that the car was going to inevitably get me in trouble. It was impractical. We were looking at moving into a new house with less parking available and something had to go. The price was a fraction of what I had put into it but it was never built for profit. Perhaps I was getting tired of working on it. I had certainly done my fair share of that over the years.
Right on schedule, the car was running again, with a fresh M20B27 in the ’86 chassis. I spent some time shaking down loose ends and breaking the car in. This was a new bottom end that I had not yet developed full confidence in, but there was little time for that. The buyer had flown in and we had two days to go over the car. He seemed somewhat disappointed in the appearance of the car. I thought I had been pretty clear about it being a “turbo shitbox”. His first impressions of the power were interesting. He didn’t seem terribly impressed, which I could not understand. The next day, we went over a crash course in tuning and an introduction to forced induction. This was a hopeless matter, one could not possibly communicate or comprehend a significant portion of this material in just a few hours. Luckily the tuning was done, the research and development was baked in. You just had to put gas in it and drive it.
That night we went out for a highway drive. This time the buyer seemed to notice what he was getting for a fraction of what I spent to build it. We hit about 140 on the interstate, which I usually tried to avoid, although the car was just getting going in fifth gear. It would probably do at least 160 despite the complete lack of aerodynamics. These top end pulls subject the engine to maximum stress, and I had not yet done this during the break in period. The guy was excited to a state approaching delirium, which provided me a sense of validation. I knew the car was fast but it was nice to see someone else realize it too.
The ’86 had an unfortunate tendency to lose fuel pressure if the level in the tank dropped too low. The fuel would slosh away from the pickup under hard acceleration, clearly not a problem with the stock 528e, and the engine would cut out under boost. To avoid this, I suggested we stop for some fuel after we got off of the interstate. Having gone along for the ride, I felt like popping the hood and having a look around this still freshly built engine. As the buyer fueled the car, I opened the hood and looked toward the coolant reservoir, the first indicator of trouble in the event of a compromised head gasket. I saw what I really didn’t want to see, the cooling system had purged a small amount of fluid, indicating excessive pressure which results from leakage of combustion gases.
The guy thought I was screwing with him. He didn’t believe it, but it was not ready to drive back across the midwest. I guess I should have taken it up to 140 before he got there and blown the head gasket myself. For this build I had decided to try another multi-layer steel gasket, this time in stock thickness. It held fine for the week or so I drove the car, but I hadn’t run it hard enough. I told him I could get a gasket and install it in a day, but I was not able to ensure that the issue wasn’t a defect in the cylinder block, which I had never run with 1.4 bar of boost pressure. He opted to return home without the car, which I repaired shortly afterward with a new, stock head gasket. I never had another head gasket issue with it.
Eventually, the car did sell. I met a kid working at another dealership that had a certain appetite for excitement. Jerry bought the car for what I had originally been asking for it from the previous buyer. A tremendous bargain. Jerry somehow survived the experience, despite being completely unprepared for a car of this caliber. He mentioned an experience going 150 miles an hour on Wyoming, when he realized the car was resistant to steering input at these speeds due to aerodynamic lift under the front of the body. Apparently the police caught up to him about five minutes later, at which point he also had to explain why the license plate on the car belonged to a different vehicle, how he had just purchased it and had not yet registered it under his name. The officer let him go without as much as a ticket, according to Jerry, and he lived to drive another day. Jerry did ask me to lower the boost pressure, not an unreasonable request, as it helped maintain traction in second gear if it was only hitting 1 bar on spool-up. He drove the car for a year or so and ended up selling it to an unknown party.
I received sporadic inquiries over the years from people on the internet saying they purchased my old car. It moved around a bit and I haven’t seen or heard anything in a few years. In the wrong hands, the car would not last long. I can only assume the worst. It doesn’t cause me to much distress in any case. These cars were all brought back from the dead. If they are doomed to the scrap yard, that’s where they would have ended up anyway had I never taken an interest. Their detour from destruction was a hell of a ride.
The officer was polite and professional. I provided my documents. He returned to my window and had me step out of the vehicle. He asked about the car. I briefly described what had been done to it, eager to spill the beans but averse to digging myself an even deeper hole. The officer seemed perhaps somewhat interested in high performance vehicles and even intrigued to some extent. He said that the fastest he was able to pace me was 77 miles an hour, so he wrote the ticket for that. We both knew I was going quite a bit faster and I certainly could have been in a much worse predicament. This was the only time I was ever caught having too much fun, which is truly incredible considering the frequency with which I indulged in this pleasure.
As 2007 neared its end, the summer of reckless good fun and pushing boundaries had eventually exacted some consequences. The head gasket I installed in late 2006 had been through hell, periods of severe detonation were inevitable during this process. My learning curve was steep and the margins of safety were diminished. I believe it was October when the head gasket had seen enough. Combustion gases had breached the sealing rings around the cylinders and begun to displace the liquid circulated through the cylinder head to wick away the wasted heat of combustion. This is not a failure of the same magnitude experienced by the ’85 about a year prior, but it required repair nonetheless. The top end of the engine was disassembled, the head gasket was replaced, and I was operating under the assumption that I was destined to be back on the road after the weekend.
Upon the initial test drive, problems arose. The coolant level was not stable, but no leaks were apparent. This is when I learned something very specific about this M20 engine. Coolant mixing with the oil in the crankcase is very common among head gasket failures in a wide variety of engines. The M20, on the other hand, does not mix these fluids unless it is through a crack in the cylinder head. Removal of the valve cover revealed a crack under the camshaft in the area of cylinder number 5. Time to remove the cylinder head again, not to replace the head gasket but the entire cylinder head. The original head that came with the car, from the barn in Tome, with close to 250 thousand miles on it, had made its last horsepower.
This was mostly just an inconvenience. It would have been nice to discover this crack before building the engine with a doomed cylinder head. A valuable lesson had been learned though. Next time I see coolant in the oil, just throw the whole head away and get another one. One would hope this was the last time I destroyed a cylinder head, but one would be disappointed in doing so.
By 2008, I was still tinkering. With the obstacle of fuel injection comprehensively addressed, I was only inhibited by the integrity of the engine and my ability to calibrate the tune to safely support any given boost level. The Megasquirt ECU could read up to 255 kPa of manifold pressure, which would be 172 kPa of boost above atmospheric. So, I could run up to about 1.7 bar of boost before the computer would reach the limit provided by the range of its sensors. This is 25 pounds of boost, which is very nearly as much as the pressure in most car tires. A hell of a lot to run through an engine designed to run on ordinary atmospheric pressure. I didn’t get there right away, it was always a methodical, incremental process.
The next major change was a different turbocharger and manifold from Todd at TCD. This was a manifold designed with a more performance-oriented application in mind. The manifold was still cast iron but used larger internal cross-sectional area and a larger overall volume. It also utilized an external wastegate, which offered improved control and easy adjustability of boost pressures. Todd had a turbo and manifold laying around that weren’t part of one of his complete kits, so I decided to give this system a try. The modified turbo diesel manifold was never ideal and it had the potential to be contributing to the head gasket destroying detonation I had learned to fear.
This system required some alteration to fit. A new exhaust system was fabricated with a larger diameter all the way back. The intake and charge plumbing were revised to connect to the new location of the compressor. The rest of the system was unaffected, so this was another weekend job. In conjunction with this kit, I decided to install a multi-layered steel head gasket. The stock gasket was susceptible to damage and this reinforcement seemed like a very good idea.
This system did offer more power but the tradeoff in response was substantial. The previous Garrett turbochargers used advanced, ball bearing cartridges and were remarkably responsive. The new turbo system used a journal bearing cartridge and could only be described as sleepy below the boost threshold. I could no longer put my foot into third gear at 3000 RPM and expect anything to happen. Once I got used to it, this was not a big deal. The car did go fast once it got going. Some tuning was required to suit the change in the engine’s compression ratio resulting from the new head gasket. Not only was it a multi-layer steel gasket, it was 1.4mm thicker than the original composite gasket. This lowers the compression ratio, which reduces the amount of heating that occurs during compression and improving the engine’s resistance to detonation.
This process should have been more meticulous than it was. I recall it was a Thursday evening and I had gone to a car meet. The car was sluggish and I had no patience for that, so I dialed up the ignition advance until it started making power. Throwing caution to the wind produced results, and there was a rematch between the previously mentioned 335i and the ’88 528. My buddy had modified his 335i to generate additional boost pressure and was eager to see how it compared to the turbo shitbox. I don’t recall to what extent he was aware that the e28 had been turned up, but I think he at least had some idea. The second matchup went down the same as the first, though the 335i made a better showing this time around. Trey’s hopes weren’t the only thing that got smoked that night, my high-performance clutch had met it’s match as well. It got the job done but it was slipping under power and could no longer be trusted.
Todd at TCD expressed some surprise that this unit had failed. I don’t know exactly how much torque it was being subjected to but it was operating close to its rated capacity. Upon disassembly I closely inspected the flywheel, which comprises one half of the clamping surface that the friction disc uses to transmit power. Here I discovered part of the issue, the face of the flywheel had been worn in a conical manner, and it tapered inward toward the center of the flywheel. Imagine a very shallow ice cream cone. The new clutch was a very exotic unit consisting of a segmented Kevlar disc. It would only be appropriate to install this clutch onto a brand-new flywheel assured to be machined to the optimal dimensions.
The new clutch was a strong unit, and I never again experienced slippage. This new turbo system was better at maintaining full boost pressure across the full RPM range. The car was going faster than ever. The next trip to the dragstrip produced an ET of 13 seconds flat, again, but with a trap speed of 114 miles per hour. This was not as fast as it should have been, and I was up against the limits of fuel delivery at the time with the size of fuel injector I was running. On this day, I must have been in a certain mood. I was going to turn it up until something happened, good or bad. Something did indeed happen. I don’t know how much boost it was making, the manual boost controllers are very touchy. Minor changes make big results. The clutch held, but the engine experienced a quarter mile of the most severe detonation I’ve ever heard. I did not lift but instead held the throttle open and ran it for what it was worth. The ET was not the best of the day, and the coolant light illuminating informed me that I had, at very least, blown that steel head gasket.
Sometimes these things happen. This should have been avoided, but it is what it is. The car was hurt but I was able to drive it back home from the dragstrip. It only ran slightly hot on the gauge as I neared my destination. Perhaps it wasn’t too bad. Upon my arrival at home, which was near Constitution and San Pedro at that time, I found signs of what I already knew was a severe and damaging event. There was again coolant in the oil. Head was cracked. Removing the intake manifold revealed a green/blue crust on the intake valves near the back of the engine. The head gasket was compromised to the extent that coolant was blowing into the combustion chamber and being drawn up into the intake tract by manifold vacuum. This was remarkable.
The tear down was completed and more parts were ordered. I was perhaps motivated to run this thing to destruction by a sense of dissatisfaction with its performance. With larger fuel injectors it might have safely made more power. I never really bothered to properly explore this avenue. Maybe I was bored and felt the need to buy an even larger turbo. That’s what I ended up doing. I didn’t really know where I was relative to the GT2560R I started with, or the GT2871R I had used prior to this current system. I know it was running a T04E 50 trim compressor which is certainly large enough. The turbine side was a bit of a mystery. The decision was made at that time to order a turbo that was, without a doubt, too large for this engine. In the future I was already planning to replace the M20B27 with something like an M52B28, a 24 valve engine from the family that replaced the older 12 valve design I was current working with. The Garrett GT3582R was more suitable for this M52, and it was again a ball bearing turbo. I longed for the responsiveness of the ball bearing units I had run previously.
This required a new manifold, this GT35R would not fit the bottom mount cast iron unit I was running with the T04E. At the time there were a large quantity of very cheaply made tubular units available on eBay, but they were supposed to fit the M20 engine and they did use a T3 turbine flange. The manifold was undoubtedly as crappy as expected when it arrived and required substantial reinforcement over the coming months as it would repeatedly crack, be repaired, and crack again. This manifold also had no provisions for an external wastegate, which was required as the GT35R had no internal wastegate. I added an appendage to the manifold, that could only be described as intestinal in it's appearance, which laid the external wastegate down under the turbine outlet. This was a rather slick matter of packaging, though it left little room for a wastegate outlet pipe to be connected. For now, this was satisfactory.
The thoroughly abused engine was repaired with a new cylinder head. This time I opted to use the original head gasket. My experience with the MLS gasket was not outstanding. Nothing is det proof, and the MLS was challenging to successfully seal up after installation. Overall I was not impressed. I did utilize stronger hardware, some quality head studs manufactured by Raceware. These were robust and reusable. They also required less torque to generate tremendous clamping force, on account of the finer thread pitch. These studs and the stock head gasket proved sufficient for all future purposes.
This system had a certain appeal for me. I had gotten back to an entirely homemade recipe. The TCD kit was nice but it wasn’t mine. I doubt I had achieved its full potential but I was ready for the next step either way. I also had a large turbocharger with known technical specifications. The turbine side of the GT35R was twice as large as the original GT2560R. I had come to understand over the course of this project that this had a substantial effect on power output and detonation resistance. The GT35R opened up more potential and greater efficiency at high boost pressures. Though the compressor side was oversized for the application it was still operating within its map at a reasonable level of efficiency.
On the road, this new system proved to be quite a performer. Spool characteristics were surprisingly good despite the oversized turbo. It would make some boost even at very low RPM. It didn’t really get going until about 3500 RPM, and then it got going all at once. The system exhibited boost creep as well, due to the compromised location of the wastegate. It was branched off of the collector just before the turbine inlet, and the turbine of the GT35R was so efficient that it presented an insufficient restriction to divert enough flow through the wastegate. It could control boost pressure to the set value up until about 5500 RPM, at which point boost would begin to climb. By 6000 RPM, it was making 1.2 bar of boost pressure, regardless of the springs installed in the wastegate. This worked out just fine in a way. Peak cylinder pressure in this engine with it’s current camshaft and intake manifold occurred around 5000 RPM. Beyond that point, additional boost could be safely dialed in without risking detonation. The additional boost pressure presented little additional risk of engine damage.
This system was run at a boost pressure of 1.4 bar for the majority of the time I owned the car. It would spool up so dramatically that the car would accelerate as if it were being “booted” down the road. Imagine a goliath-sized giant literally kicking the car forward. This was the sensation experienced in higher gears. In first, second and sometimes third, the tires would simply break traction and the engine would bounce off of the rev limiter as the car struggled to put the power down in a long, rolling burnout. Final drive gearing was very tall, second gear would clear 70 MPH. Shifting into third would generate either tremendous acceleration or a highway speed drift depending on the condition and quality of the pavement. Usually it hooked up though, and the car would move out into the triple digits in a manner that seemed to completely disregard the exponentially increasing drag and the nearly three thousand pound weight of this ridiculously overpowered automobile.
At this point it was 2009 and the car had become a monster. It was nearly impossible to launch. As soon as the boost came on it would just blow the tires away. On the next trip to the dyno, the car spun the tires on the rollers, requiring extra straps to get the tires to overcome the massive inertia of the dynamometer rather than convert horsepower into a quantity of tire smoke which was impossible to measure. The result was outstanding. Without any correction factors applied, it produced 384 rear wheel horsepower and 382 foot-pounds of torque. Between 3500 and 4000 RPM, torque output doubled as boost pressure jumped from the single digit range to the full 25 pounds of boost. Even at this altitude, this was 450 horsepower at the crank. Not bad at all for a 12 valve 2.7 liter inline six that left the factory making about one quarter of that. Best of all, at this altitude, this 384 rear wheel horsepower would be comparable to about 560 crank horsepower at sea level. This crusty, cosmetically challenged e28 was able to embarrass some very exotic cars. The only cause for concern on the day was the large puddle of oil under the dyno after the pull. It had been carried out by crankcase gases despite the best efforts of the crankcase ventilation system. I had not seen this in previous dyno runs, but I had noticed an increase in oil consumption. Perhaps this was the nature of running high boost through an M20. Who could be concerned with that though, the car was running great.
One such exotic car was encountered late at night one evening, going eastbound across the Rio Grande. This scenario was tilted very much in my favor. We were cruising at highway speeds, where gear ratios were tall enough that the tires only had to handle a fraction of the power and torque delivered in lower gears. The victim was a bright yellow Corvette, a C6 Z06. This was the top of the line variant of the Corvette, with a 7 liter V8 engine producing 505 horsepower and 400 foot-pounds of torque. Quite unlike mine, this car was designed, built and bred for high performance. I recognized the LED rings visible in its taillights, as well as the vacuum controlled butterfly valves in its exhaust outlets. The “Z06” badge on the front fender just behind the tire subtly indicated what its other features had already disclosed. This was General Motor’s fastest version of the Corvette currently available.
The challenge for me was trying to set the hook. My car was completely invisible to people like this. An old BMW, unassuming to say the least. I would usually spool the turbo up briefly and give them a quick “jump” to provoke their interest and communicate my intentions. Sometimes they would bite, sometimes they were far more responsible than I and opted to mind their business within the bounds of the posted speed limit. For the Vette, I showed him what third gear would do. He didn’t have any advance notice, and he was left far behind. This wasn’t a race, just an offer to engage in a bit of extralegal tomfoolery. He pursued, and I slowed down. Then he opened it up. I pursued, then surpassed the Corvette. We had the road completely to ourselves, and we went back and forth a few times before reaching the interstate. He had bought the fastest Corvette available at the time and it wasn’t fast enough. As we headed south on I-25, a police car appeared on the frontage road with it’s lights on, moving at about the same speed as us and in the same direction. Could have been a complete coincidence, as neither one of us ended up being pursued and we went our separate ways.
Nobody believed me about this encounter, but the math is very simple. I’m making 384 rear wheel horsepower, measured at this altitude. The Z06 is producing 505 at the crank at sea level, of which about 404 make it to our elevation, and after being transmitted through the driveline about 344 reach the rear wheels. I had a large advantage in torque, 382 foot-pounds at the wheels. The Z06 only had 272 left of its original factory rating of 400 foot-pounds. It would have been a battle at sea level but this guy, along with everyone else breathing the thin atmosphere, was missing about 20% of his get-up-and-go at our elevation. Even then, I had a substantial advantage in torque, and even turbocharged engines run more efficiently at sea level. The turbo doesn’t have to work as hard to make any given manifold pressure with denser air available at the compressor inlet, and it operates at a lower pressure ratio. All things considered, the Z06 would have almost certainly taken the L there too.
I never went back to the dragstrip. The car wasn’t suited for it. It was a street car and it did pretty well out there. Standing starts were fun but not particularly fast. I once got smoked, for the lack of a better word, by a stock e60 535i. We had lined up at a traffic light, and it was driven by some of my co-workers at BMW. They knew about the e28, and my buddy in the passenger seat was hyping it up. Said the car was sick or nasty or something along those lines. I tried to temper expectations, saying the car was “hard to launch”. I had handily beat-ass all over one of these twin turbo models previously, with the GT2871R at 16 psi, but now I was driving an entirely different car. The light turned green, the e60 hooked up and accelerated hard. I laid down rubber in a feeble attempt to translate horsepower into forward motion, but after spinning through first and second gear I had been left behind. Not my best work. Sometimes I missed the old days of having usable power. The more I built into this thing, the less of that I had.
I don’t remember the year, but at some point I found another e28. This one was a 1986 528e, sitting in a field in Corrales. Had a blown heater core and was just as ugly as the ’88, but with a straighter body. The price was right, another $500 car. I bypassed the heater core, dunked a battery in and drove it home. The transmission didn’t like to stay in fourth gear, couldn’t do a lot with that, but it did come with some other good parts. The clutch was brand new, although completely insufficient for the current power level. I planned to skeletonize the ’88 and swap everything into the shell of the ‘86. One issue that I couldn’t resolve with the ’88 was a matter of clearance on the left rear wheel. There had been some body damage at some point in the car’s life which had resulted in the inner wheel well being pushed in. They fixed the outside but not the inside, so it rubbed heavily on the 265/40R17 tires I was running in an attempt to improve traction.
This must have been a tremendous amount of work. I can’t imagine doing this again, though I thought nothing of it at the time. It did present an opportunity to go through this motor I had been beating the shit out of for the last three years. I had a spare engine block I could use to build a fresh engine and I had found a set of pistons on eBay for a mere $200. The seller had them listed at $600, but the seller accepted a substantially lower offer. I was delighted to have found them, and I’m sure they were amazed anyone out there was looking for them. This was a rare build. Few people were doing anything with the M20. Most of the builds were with the 24 valve M50/M52, such as the one I was planning on using, but with the results I had with the GT35R on the M20 I couldn’t justify it.
Upon disassembly of this motor, the original one that came with the ’88 from the barn in Tome, I found some very surprising things. Most of the connecting rod bolts were very loose. None had loosened completely, but they should have been much tighter. This engine had a rev limit of 4700 RPM originally. I was running it up to 7200 at times, and reciprocating loads increase exponentially with RPM. That could do it. I also recall one instance where I mis-shifted and went from redline in second to first gear. This should have resulted in a catastrophic, likely explosive engine failure. I don’t know how high it overrevved but it was higher than it ever went of its own accord. This was years ago though. It seemed implausible that the motor would have lasted that much longer afterward if it had overstressed the rod bolts at the time.
With the rod caps off I was able to pop the pistons out of the block. This is where I found the truly scandalous details. In the rear cylinders, 5 and 6, the piston ring lands were destroyed. On number 6 the piston rings themselves were broken into multiple pieces. These pieces were trapped in place only by the remaining upper half of the top ring land and the skirt of the piston. When did this happen? That would have been the last trip to the dragstrip, about 18 months prior. That would also explain the oil consumption and large puddle of oil under the dyno. Would have probably made a few more horsepower with six properly sealed combustion chambers but I suppose I’ll never know for sure. What was really amazing was that these were the original pistons that had accumulated 238 thousand miles before I even acquired the car. I must have put another 30 or 40 thousand miles on them since the rebuild in 2006. If not for the completely avoidable incident at the dragstrip in 2008, they would have withstood everything else I had thrown at them.
Fortunately, I had other cars to drive. I had purchased an e39 528i from one of our customers at the dealer. The car had been overheated, the Achilles heel of the M52TU and M54, and the engine had been damaged. I got it for $2000 with about 90 thousand miles on the odometer. In continuation of my tradition of finding some way to drive these things home, I tied off the radiator hoses to prevent coolant loss through the gaping hole in the radiator end tank and I drove it home. This was an interesting experience to observe the various stages of overheating as indicated by the e39 instrument cluster. These temperature gauges are “buffered”, which means the computer evaluates the coolant temperature signal and displays its own simplified version to the driver. Most drivers will only ever see the needle at the “cold” and “normal” positions. Beyond that, there is a position at three quarters, indicating elevated temperatures. Then the needle goes into the red, indicating overheating outside of any normal range of operating temperatures. Then an orange light illuminates next to the red zone, in case you didn’t notice the needle position. In its final stage, the needle moves beyond the red zone and the light turns off. In practice, any indication above normal indicates serious overheating and potential engine damage. If by any chance one wasn’t cooked before I brought it home, it certainly was now.
A replacement engine cost about $1350. All of the other things that should be replaced while the engine is out cost another thousand or so. I was working at the dealer at the time, where I took tremendous advantage of the employee discount on genuine BMW parts. I must have spent tens of thousands of dollars over the course of my employment there.
The engine was removed with the transmission left in the car. This e39 was a different animal than I was used to, not as pleasant to work on as the e28. Somehow, I got the replacement engine in and buttoned everything up. The car started and ran beautifully. This was certainly far nicer than what I was used to driving, although quite a bit more ordinary in terms of performance. It did have air conditioning, a contemporary interior and superior sound insulation. This thing was a nice commuter. A bit more practical and enjoyable than driving a turbo e28 with the temperament of Dr. Jekyll.
I would by lying if I said I didn’t consider installing a turbo M20 into the e39. I think I made the right decision though. I still had the e28, and the ’86 was fleshed out with the very nice interior pieces from the ’88, the subframe and suspension, and eventually a new engine and transmission. I had a used M20B25 bottom end that came from somewhere or another, and I thought it would be an interesting experiment to install that engine into the ’86 with the existing turbo system. I had always run a modified M20B27 with what was essentially an M20B25 top end, so this was an interesting experiment to me. The engine was assembled as is, without anything more than a new oil pan gasket, and installed in the car. It ran well, with perhaps a slight advantage in top end power. This is not out of the realm of possibility, considering the same intake manifold was used on the 2.5 as I was using on the 2.7. On the smaller engine, this manifold would make power to a higher RPM due to the length of it’s runners and their cross sectional area.
It also made more boost though, at one point pegging the manifold pressure sensor at 260kPa. This was .05 bar higher than it should have read, and it was seeing even higher pressure than that. In conjunction with the slightly higher compression ratio generated by the M20B25 bottom end, this was too much to ask from the head gasket. I don’t know how many I had been though by this time, but I pulled the top end off and replaced the head gasket, throwing the old one on top of the stack of the others I had hanging on a nail against the wall. About a three hour job these days. That evening, I finished assembly and cranked the engine which nearly fired on the first try. On the second try, the engine stopped abruptly and would not crank. On subsequent attempts, the engine freed itself and turned over, but with no compression and only an occasional thump of resistance. Something terrible had happened to this engine, and that was the last of the M20B25 experiment.
I never really figured out where this went wrong. It clearly lost time at some point, allowing the crankshaft to turn independently of the camshaft. Valves hit pistons, one valve broke off, flipped over and was punched through the crown of the piston below as it collided with the cylinder head. During the process of installing the timing belt, the engine must be turned over twice to ensure proper tension on the belt and check timing. It was timed properly at this point, but something happened afterward that destroyed the engine. Oh well.
The new engine was well on its way. The spare engine block I had, which interestingly enough came out of the ’83 528e I rode around in when I was a child, was bored to fit the Mahle pistons I scored on eBay. Considering what I had done to the crank from the ’88, I figured I would use the ’83 crankshaft as well. These are cast iron, and the common wisdom about them among the BMW community was that they were weak and would break if revved over 5500 RPM. I never experienced that, and it wasn’t for a lack of trying.
This engine was assembled in what was soon going to be my daughter’s bedroom and wheeled out of the house on a furniture dolly. I was actually trying to get the car ready for a buyer that was coming in out of Michigan. I had convinced myself that the car was going to inevitably get me in trouble. It was impractical. We were looking at moving into a new house with less parking available and something had to go. The price was a fraction of what I had put into it but it was never built for profit. Perhaps I was getting tired of working on it. I had certainly done my fair share of that over the years.
Right on schedule, the car was running again, with a fresh M20B27 in the ’86 chassis. I spent some time shaking down loose ends and breaking the car in. This was a new bottom end that I had not yet developed full confidence in, but there was little time for that. The buyer had flown in and we had two days to go over the car. He seemed somewhat disappointed in the appearance of the car. I thought I had been pretty clear about it being a “turbo shitbox”. His first impressions of the power were interesting. He didn’t seem terribly impressed, which I could not understand. The next day, we went over a crash course in tuning and an introduction to forced induction. This was a hopeless matter, one could not possibly communicate or comprehend a significant portion of this material in just a few hours. Luckily the tuning was done, the research and development was baked in. You just had to put gas in it and drive it.
That night we went out for a highway drive. This time the buyer seemed to notice what he was getting for a fraction of what I spent to build it. We hit about 140 on the interstate, which I usually tried to avoid, although the car was just getting going in fifth gear. It would probably do at least 160 despite the complete lack of aerodynamics. These top end pulls subject the engine to maximum stress, and I had not yet done this during the break in period. The guy was excited to a state approaching delirium, which provided me a sense of validation. I knew the car was fast but it was nice to see someone else realize it too.
The ’86 had an unfortunate tendency to lose fuel pressure if the level in the tank dropped too low. The fuel would slosh away from the pickup under hard acceleration, clearly not a problem with the stock 528e, and the engine would cut out under boost. To avoid this, I suggested we stop for some fuel after we got off of the interstate. Having gone along for the ride, I felt like popping the hood and having a look around this still freshly built engine. As the buyer fueled the car, I opened the hood and looked toward the coolant reservoir, the first indicator of trouble in the event of a compromised head gasket. I saw what I really didn’t want to see, the cooling system had purged a small amount of fluid, indicating excessive pressure which results from leakage of combustion gases.
The guy thought I was screwing with him. He didn’t believe it, but it was not ready to drive back across the midwest. I guess I should have taken it up to 140 before he got there and blown the head gasket myself. For this build I had decided to try another multi-layer steel gasket, this time in stock thickness. It held fine for the week or so I drove the car, but I hadn’t run it hard enough. I told him I could get a gasket and install it in a day, but I was not able to ensure that the issue wasn’t a defect in the cylinder block, which I had never run with 1.4 bar of boost pressure. He opted to return home without the car, which I repaired shortly afterward with a new, stock head gasket. I never had another head gasket issue with it.
Eventually, the car did sell. I met a kid working at another dealership that had a certain appetite for excitement. Jerry bought the car for what I had originally been asking for it from the previous buyer. A tremendous bargain. Jerry somehow survived the experience, despite being completely unprepared for a car of this caliber. He mentioned an experience going 150 miles an hour on Wyoming, when he realized the car was resistant to steering input at these speeds due to aerodynamic lift under the front of the body. Apparently the police caught up to him about five minutes later, at which point he also had to explain why the license plate on the car belonged to a different vehicle, how he had just purchased it and had not yet registered it under his name. The officer let him go without as much as a ticket, according to Jerry, and he lived to drive another day. Jerry did ask me to lower the boost pressure, not an unreasonable request, as it helped maintain traction in second gear if it was only hitting 1 bar on spool-up. He drove the car for a year or so and ended up selling it to an unknown party.
I received sporadic inquiries over the years from people on the internet saying they purchased my old car. It moved around a bit and I haven’t seen or heard anything in a few years. In the wrong hands, the car would not last long. I can only assume the worst. It doesn’t cause me to much distress in any case. These cars were all brought back from the dead. If they are doomed to the scrap yard, that’s where they would have ended up anyway had I never taken an interest. Their detour from destruction was a hell of a ride.
Re: Got bored at work, wrote up my life story
Wow, even if you type really fast, you had some free time.
Good story, I was happy to see it only related, at least 99% related to BMWs. I really didn't want to see marriages, divorces, kids, jail, etc. At least not here.
Good story, I was happy to see it only related, at least 99% related to BMWs. I really didn't want to see marriages, divorces, kids, jail, etc. At least not here.
Re: Got bored at work, wrote up my life story
I'm surprised anybody read all that.
I did try to keep it about the cars. It was fun going back through the projects in such detail. Brings back some vivid memories. I may have to go back through and work in all the things I left out. The e34 540i project was omitted entirely, and this story wraps up well before the current e28 build began. This may end up being a novel when it's all sorted out.
I did try to keep it about the cars. It was fun going back through the projects in such detail. Brings back some vivid memories. I may have to go back through and work in all the things I left out. The e34 540i project was omitted entirely, and this story wraps up well before the current e28 build began. This may end up being a novel when it's all sorted out.
-
austin8753
- Posts: 1490
- Joined: May 16, 2010 1:37 AM
- Location: Portland, Oregon, USA
Re: Got bored at work, wrote up my life story
that was quite the journey. cool to see the natural progression of events. thanks for sharing. looking forward to part 3.
Re: Got bored at work, wrote up my life story
I'd definitely watch the movie.
AI anyone?
-
CanadianMiniFan
- Posts: 140
- Joined: Oct 16, 2011 6:31 PM
Re: Got bored at work, wrote up my life story
Another member who read all that. Very interesting. Any reason you were never lured to try an m30 turbo set up?
Re: Got bored at work, wrote up my life story
I could have gone that way early on. Now that you mention it i remember that I did have a 533i that I bought for the shell. The whole car was clapped out and I ended up scrapping it, M30 and all. That one never got off the ground. I'll have to write that part in...CanadianMiniFan wrote: Oct 10, 2024 6:16 PM Another member who read all that. Very interesting. Any reason you were never lured to try an m30 turbo set up?
The M20 has advantages in weight and fuel efficiency. Even with boost, I get way better mileage than an M30 ever could. The M20 seems to hold up to boost better than the M30 too. I am currently running a stock head gasket and stock head bolts at .8 bar. The original build was running 1.4+ on a stock gasket with studs. I don't think I've ever heard of an M30 putting up with that.
Either motor can make all the power you want. I never found the limit of the M20. If I ever did try another platform it wouldn't be another 12 valve motor.