Project BMW M3 Part 2: Cooling system and airflow sensor

Boy, this project is "fun." I never thought I'd experience so many headaches deciding which aftermarket companies to use. The competition is so great, the quality offerings so abundant, that lengthy research is inevitable in order to make an intelligent decision.

In the past several months I'd learned from a couple of tuners that E36 BMW factory cooling systems tend to last around 5 years or 60k miles, depending on the nature of driving. Not knowing what the previous owner of this '95 M3 was like, yet knowing how much testing and high-performance driving this vehicle will go through, I opted to replace it all--radiator, belts, hoses, water pump, thermostat and thermostat housing. Even though the car had less than 40k at the time, the car was approaching the 6-year mark. So, I thought, "Better safe than sorry!" In this case, deciding on which supplies to use wasn't an issue, as 90 percent of the cooling system components were ordered from the dealer.

Installing the radiator is not an easy task, but I was fortunate enough to have friend Jonathan Caldito, a promising, young certified BMW tech from United BMW in Duluth, Ga., do the work on the car. His technical expertise with these cars led to a quick and clean installation.

The factory water pumps supplied on the '95 M3s had plastic impellers and were more prone to failure than the metal ones supplied in the '96 and later M3s. For this reason, BMW now supplies only water pumps with metal impellers for all E36 M3s. I went ahead and ordered one, as this is something that's been known to plague these M3s--if your '95 M3 has more than 50k miles and you haven't replaced the water pump yet, now might be a good time to do so.

Active Autowerke's thermostat housing replaced the original plastic one that came with the car. Active's heavy-duty thermostat housings are aluminum and CNC machined to an exact shape for a perfect fit. These housings are less prone to cracking under heat and pressure like the factory ones and, frankly, made more sense to use in the project. After the install I noticed this thermostat housing truly had a factory-like fit and look to it. The thermostat itself came from Active Autowerke as well.

When it came time to replenish the cooling system, a 20-percent coolant/80-percent water mixture was used. Living in southern California, I've had to worry more about overheating than I do the radiator freezing, and since water cools more effectively than coolant, this combo made more sense. A small bottle of Redline "Water Wetter" was added to further help reduce heat and corrosion.

With the new cooling system now installed, I was more confident in boosting the straight-line performance of the M3. Now, the logical way to modify a car is to first do the brakes, followed by the suspension and lastly the engine, but who cares--I couldn't wait, and these parts had already arrived. Besides, there's a big-brother edition of this car in Europe handling 80 more horsepower on the same chassis, so I knew this car would be able to handle a few more ponies.

As mentioned in the previous write-up, this M3 delivered 218 bhp to the wheels. I figured that getting a normally aspirated car that's running so well to put out even more horsepower would be no easy feat--notice the surge in bhp and torque at 6100 rpm. But it was accomplished with a few simple modifications to the intake, exhaust and ECU. To verify the results, the car went to Dominic Conti at the McMullen Argus Tech Center, where he put the car through several dyno runs.

The first modification we tested was an old intake system that came out of my late '97 M3. This MSDS cone filter sat on top of the brake duct, which had a hole cut out for additional air. The filter was attached to a tube that was fitted to an adapter to fit the HFM (Hot Film Mass, also referred to as the air-flow meter). Although the peak bhp didn't increase very much, the filter made a noticeable difference on the '95 M3 throughout the rev band, shining in the midrange with a 6.9 bhp and 8.7 lb-ft peak increase over stock.

When I first picked up the car, a Dinan chip was already installed. I decided to test that with the conical intake mentioned above. Again, there wasn't much of an increase over stock peak bhp, but there was additional improvement throughout the rev range, widening the torque curve with a healthy 190 lb-ft of torque from 3550 to 6100 rpm (compared to 3950 to 5900 rpm for stock).

Naturally, the next modification we tested was an exhaust system to complement the intake and ECU modifications. For this, Active Autowerke sent its new stainless-steel Generation III exhaust system. My obsession with weight reduction led me to immediately weigh this shiny new piece. The cat-back exhaust tipped the scales at only 35 lb (compared to 45 lb for the stock muffler), so it passed my first test. The muffler also fit perfectly on the car, which should have meant a short installation. Unfortunately, the factory rubber exhaust hangers were failing (the AA exhaust mounts directly to the exhaust hangers, so make sure yours are in good condition), and waiting for the local BMW dealer to get a couple of new hangers in stock set me back a day and a cool $85--a small price to pay to prevent the thing from falling off!

The twin 3-in. rounded-tip exhaust gives the M3 a much more aggressive look and provides a deep exhaust note that makes the sound of the inline six music to the ears. An Active Autowerke exhaust is also very easy to spot from the others by its hexagonal shape and shiny polished finish. When I first drove the car with the exhaust installed, I noticed a cleaner, smoother pull to redline. Previously I was tempted to shift a little sooner; I could now hold the throttle down through 6500 rpm without the power falling off.

When the M3 was put on the dyno, my driving impression was confirmed: With a few more bhp throughout the range over the previously tested stock muffler with Dinan/intake kit, the exhaust did a wonderful job with a max increase of 6.1 bhp at 6550 rpm. It seemed this was just what the car needed--taking into consideration, of course, the intake and chip currently operating on the engine.

Karl Hugh of Active Autowerke offered to take it a step further in the race for more horsepower and sent an ECIS (East Coast Induction Systems) intake kit, which is comprised of a larger air flow meter from a BMW 540, an air filter and a heat shield. Karl also burned a custom chip for this application to deliver the proper fuel the larger meter demanded and avoid any possible leaning-out. Active Autowerke had expanded its R&D into chip tuning since its acquisition of a Mustang "loading" dynamometer, and successfully teamed up with ECIS' expertise in BMW intake systems. Looks like AA is also catering to the folks who are deathly afraid of the power those turbo kits make (check out our cover story in last month's issue!) with some simple bolt-on components. AA's new projects include software tuning for the new 3 Series BMWs, including the 328 and 330.

The AA/ecIS HFM kit made a dramatic improvement in the project M3's performance. The midrange power increase was now very noticeable in everyday driving, and the difference in the upper powerband was quite apparent as well. The first thing I noticed driving the car was the more aggressive intake sound it now made. Once the car hit 4500 rpm it would growl with fury, kicking in an extra 16.8 lb-ft of torque over stock at a very usable 5350 rpm. At this stage, the car didn't sound like a 3.0-liter anymore. The rev limiter was also increased to 7250 rpm, giving me an extra 650 rpm to work with over stock. But the M3, at this time, wasn't making the power necessary to have to rev it that high, so hard acceleration shifts at 6500 rpm seemed the most effective.

Now I had a car that was making over 224 bhp at the wheels, and things were getting a lot more fun. But why stop there? In his Jamaican accent Karl mentioned his firm's 21.5-lb injector setup for the '95 M3: "We basically take injectors from a stock 3.2-liter M3, put them on a 3.0 liter and make a chip to give the extra fuel, mon." He said the stock 3.0-liter M3s deliver fuel via 17-lb injectors, which need to be upgraded when certain modifications are done to the 3.0 motor. "Once a 3.0-liter M3 makes over 225 bhp at the wheels, larger injectors are needed because the duty cycle of the injectors is too high, mon."

At first the injector swap seemed a little complicated, but Karl patiently walked me through it over a few phone calls. Since they were factory 3.2-liter injectors, there was no worry of an imperfect fit. The new chip that Active Autowerke supplied was different this time, however. The chip came attached to a green board. "The chip is encrypted," Karl explained, "and this green board unique to Active Autowerke is the only way it will work. It's like a key."

Back at the McMullen Argus Tech Center again, Dominic and I ran the car through some more dyno runs and, without surprise, the car continued to make more power. If you look at the graph, you will see the power is very similar to the previous AA/ecIS HFM/stock-injector setup, until the car hits 6000 rpm, making considerably more power through redline. At 6900 rpm the M3 put out an additional 10.8 rpm over the AA/ecIS HFM/stock-injector setup! Peak torque and horsepower were achieved at considerably higher rpm than the other tested setups, and the car didn't lose any substantial low-end torque or horsepower at the same time.

So, we got the best of both worlds. By the end of the day, the M3 put out a healthy 229.3 hp at the wheels at 6560 rpm (over 410 rpm higher than stock), and 214.1 lb-ft of torque, which was accomplished at a high, yet often used 4700 rpm. The torque curve was improved substantially over stock, with over 190 lb-ft of torque at the wheels as early as 3480 rpm (compared to 3950 rpm) all the way to 6320 rpm (vs 5900 rpm).

Although it's hard to notice in everyday driving situations, the infamous "Vanos dips" that happen around 4000 rpm showed up on our dyno charts. And for some reason, the AA 21.5-lb-injector setup did not eliminate the dip in power as effectively as the other setups. These dips, however, happen so fast and in such a short rpm range that it's not very noticeable in real-world driving situations. In addition, the rpm cutoff came down an extra 150 rpm from the AA HFM stock injector setup. If anything, I figured the 21.5-lb-injector setup would have more use for the extra rpm over the stock-injector setup.

When asked about this, Karl simply said this chip was one he may have overlooked in correcting to a 7000-rpm limit, which is what he decided to do for all his normally aspirated chips. "Taking into account the valve retainer problem the early M3s had, we decided to keep our rev limiter down to 7000 rpm," he said. "Too many people are cranking their cars with stock retainers to 7400 rpm and are breaking motors! I feel a 7000-rpm limit is much safer." But Karl assured me he always has people wanting more rpm: "Sure! I'll give them more rpm if they demand it. But if they want to blow the engine up they're on their own, mon!"

After all of these modifications and tests, I have found myself left with an Active Autowerke-tuned setup that makes this M3 much more fun to drive. The driveability has improved greatly, and the car is simply faster and more responsive. With these simple bolt-on mods, my 3.0L M3 was transformed into a more powerful, V8-threatening machine that pulls hard all the way to 7000 rpm--a fact witnessed by an unsuspecting Ram-Air Camaro SS a few late nights ago (remember, they're rated at 330 crank bhp!); I won't go into detail about that, but I think Karl may have made the top speed governor magically disappear.

What's next in line for european car's project BMW M3? Let's just say we haven't even begun to have fun yet! Stay tuned.

Active Autowerke
9940 S.W. 168th Terrace
Miami, FL 33157
(305) 233-9300
Fax: (305) 253-8921
www.activeautowerke.com