Engine bay teardown
OK, I was extremely embarrassed and not at all pleased with the dyno testing results of my R28 presented in Part Three of this project car series. The awesome Advanced Motorsports (AMS) four-wheel MAHA dyno beat me fair and square. Sadly, I still have 'My R32 Blues'. Perhaps I was asking a bit too much of the little 2.8L VR6 24V (with cams, chip, and numerous bolt-ons), to put out power and torque numbers approaching the 3.2L R32 engine. Anyway, after such a defeat I felt it was time to go back to the drawing board. It did not, however, take me long to determine that my next plan of action was to go all out. No more modest gains with bolt-ons. This time I would open my wallet for some real performance. That is, to a performance level well beyond even the factory's mighty R32. I reckoned an extra 150bhp or so would turn my lightweight R28 into a lethal track weapon, while still maintaining desirable sleeper status. The time had come for forced induction, and the name of the game was Turbo.
In the summer of 2004, Marcel Horn, president of HPA Motorsport, took me for an exciting ride in his infamous 550 bhp twin-turbo R32. Having driven and been a passenger in many fast cars as a track day enthusiast and performance driving instructor, I climbed into the passenger seat, buckled my safety belt, and thought: "How fast can it be?" It was not only fast, it was beyond a doubt the fastest car I have ever been in (exotics, Vipers and Porsche Turbos included). The twin-turbos' power was bordering on unbelievable and the R32 platform's ability to put this extreme power to the road was equally astonishing. I made a mental note: "I've got to buy one."
The junk pile (or classified section) !
New Turbo Power
Having spent a good deal of hard earned cash converting my front-wheel drive GTI to the R32 4Motion driveline, stepping up to a twin turbo was completely out of the question. A more affordable and conservative solution was required. This was after all, my daily driver. Fortunately, in mid 2005, HPA Motorsport released several new single turbo systems for 24-valve VR6 engines. These use an R30 Garrett ball bearing turbocharger with custom HPG compressor housing, integrated recirculation valve, and custom HPG exhaust housing with integrated wastegate, for reduced turbo lag and therefore improved throttle response. The system I chose was HPA's top-line FT410 Series.
This is configured with a proprietary HPG short-runner cast aluminum intake manifold, tuned cast exhaust manifold with external oxygen sensors, dual side-mount TT intercoolers, custom stainless steel boost piping, a compression-reducing spacer (down to 8.5:1), high-flow fuel injectors, a larger diameter mass air flow (MAF) unit, an auxiliary inline fuel pump, a 100-cell-count 70mm race cat (dual R32/TT-specific 100-cell race cats were used in my 4Motion application for proper ground clearance), and race-grade connecting rod bearings and bolts to deal with the extreme forces of the application.
All this is put together in a beautiful OEM quality package, the software tuning is seamlessly integrated into the factory ECU, and maximum boost set to 14psi. By request, my software was tuned to run on readily available 91-octane premium fuel. The result is a ferocious brute with excellent driveability for a single turbo application. Figures 1 to 8 show the progression from dyno-dud to dyno-dude.
Dyno Results and Interpretation
Head spacer install
HPA provided pre- and post-turbo-install dyno runs on the same two-wheel 248C Dynojet. Peak values for these runs were: 181whp/178lb-ft in normally aspirated form (as previously tested in Part Three), and 314 whp/329lb-ft at 14 psi boost, with the turbo system installed. Please note that while the gains in peak power and torque are quite significant, these values are only a small part of what can be learned from the charts. This will be further explained below. Figure 9 shows all pre- and post- curves nicely overlaid on the same plot for your viewing pleasure. This compilation was created by reading the data off the extremely smooth raw curves at 250rpm increments. Also, both dyno runs were done in fourth gear with the Haldex unit's electrical circuit disconnected for obvious safety reasons.
There is a great debate in the VW/Audi tuning scene with respect to converting such peak values from a two-wheel dyno run to actual horsepower at the flywheel-which I'm not going to continue here. The range of conversion factors for MkIV 4Motion Haldex-equipped vehicles is, at present, 0.78 to 0.85. The reason for this variance is due to the unknown additional drag on the measured value caused by the mechanically-pumping Haldex unit. The final drive attached to the front drive is split mechanically and is spinning the center prop shaft during the dyno run. The rear section, even when unplugged, is pumping oil, generating back pressure and drag on the system. The 'chosen' conversion factor is divided by the measured peak wheel-hp value, thus peak horsepower at the flywheel in this case ranges from 369 to 403bhp (depending on who you believe).
Conversion factors aside, this is still not the full story, as fourth gear dyno pulls of turbocharged engines on chassis dynamometers without proper wind tunnels are most definitely not representative of real-world output. It is critical to understand the effect of measuring the generated load in a fourth gear run (simulating 32mph to over 107mph) when reduced air flow is traveling across the radiator, intercooler and firewall.
New short runner intake manifold
Let's see what else can be learned from the dyno curves. First, look at the relative gains at the wheels: 133 wheel-hp (up a very satisfying 73 percent), and 151lb-ft of torque (up an incredible 85 percent). The initial surge in the turbo TQ curve demonstrates the amazing efficiency of the Garrett/HGP R30 turbo, this is really the key in punching the car off the line. Next, it is important to notice the enormous gains in area under the turbo curves. In fact, the turbo curves reveal that from 3500-6400rpm, while the engine is under full thrust, it does so without dropping below a whp or TQ value of 290 at the wheels (see the red line traced on the chart in Figure 9).
Remember that "torque gets you going and horsepower keeps you going". They work together to propel you down the road. Further interpretation of the turbo whp curve shows that shifting before just 6000rpm will maximize a fourth gear pull. Note that this will differ for each gear as the load on the motor will vary. During a real-world pull in fourth gear, I didn't detect any loss in power until about 6500rpm. This helps substantiate the aforementioned need for proper airflow to achieve real-world power outputs.
The HPG/Garrett R30 turbocharger and cast exhaust manifold
Real World Driving
How does the new turbo system work in the real world? Regardless of what the numbers say, I can tell you the delivery of power is smooth and strong. After starting the engine, I was pleased to learn there were no codes thrown or engine lights brightening the dashboard once the ECU finished its checks. It idled as smoothly and quietly as it did before, albeit a little higher at 750rpm (when warm). I assume this is necessary due to the reduction in engine compression. One of the aspects of this system I was quite surprised with (besides the power) was that I have lost nothing off-boost (this is demonstrated nicely by the dyno curve overlays in Figure 9).
With the engine fully warmed up, the car is just as easy to drive around town. Throttle sensitivity is nice, linear yet responsive, making the power easy to modulate. As I use a bit more pedal, I enter into the lower boost range. There are some interesting but satisfying whirrs and swooshes up front. The steep torque curve between 2100 and 3800rpm does appear ominous, but, because of the excellent throttle sensitivity and response, accelerating at any desired rate through this rev range in any of the lower gears is not an issue.
Dual TT intercoolers with beautiful stainless steel boost piping
Full stab on the throttle in second gear builds boost nearly instantaneously and you're rewarded by a rocket-like thrust squashing you into the seat. Be absolutely sure to have a clear, straight road and both hands firmly on the wheel. The sound begins as the gentle VR6 hum we all know and love, then transforms from growl to roar to shriek, like a cross between a race car and jet fighter. The car accelerates feverishly toward 60 mph, as the tach races towards to the new 7000rpm redline in a blink. Fortunately, with 4Motion, there is no lack of grip, so none of this power is wasted on burning up tires. A quick shift to third is rewarded by a similar sound and an almost equally forceful thrust of acceleration when hard on the throttle. Before you can say "80mph" you're doing it.
Stepping off the throttle in either gear will throw you forward, due to the loss of acceleration, and cause the recirculation valve to make a minor rumbling and rubbing noise that is slightly audible through the dashboard. This is normal. For full throttle runs starting in first gear, the rate of acceleration is obviously sky-high. In these instances, be sure to have the corner of your eye on the tach.
The finished product
I identified the stock clutch as a weak point on getting my new-found power to the pavement. To rectify this I ordered the HPA/Sachs R32 Racing pressure plate and unsprung organic friction disc, for strong, yet streetable, clamping power (see Figure 10). This setup is designed to work with the stock dual mass flywheel and does so extremely well. When driving around town, it is virtually indistinguishable from the stock unit, due to its light pedal feel. The unsprung organic friction disk provides smooth and quiet operation. Note that this is the same pressure plate used for HPA's early 450-plus bhp twin turbos, so I am not worried about slippage in higher gears.
In order to retain the expensive aftermarket stainless exhaust system (2.5 inch diameter) described in Part Three, HPA recommended the installation of a computer-activated exhaust bypass valve. This works via vacuum line in much the same way as the OEM R32 exhaust flapper. Once activated at 3100rpm, the bypass valve redirects the exhaust flow, half through the muffler and half through the bypass dump (see Figure 11). The change in sound is fairly minimal (but nice all the same) and this releases the back pressure on the exhaust system, thus increasing power in the top end. Unfortunately, this gain cannot be seen realistically on the dyno chart (which drops quite drastically), due to the artificial loading and airflow during the dyno run.
Next, the lower rearward engine/transmission mount, often called the dogbone, was another item to be addressed and strengthened. HPA has a new solution for this as well. A proprietary set of bushings was designed to replace the OEM rubber in the dogbone (see Figure 12). This modification was noticeable the first time out of the parking lot, as I could feel the more direct connection of the drivetrain to the road in first and reverse gears. An additional plus is that it provides as much vibration damping from the engine as the stock rubber, so your eyeballs remain unshaken.
With the increased performance, it was now necessary to more closely monitor my engine's condition, so I decided to add some aftermarket gauges. In a MkIV VW it is possible to install up to seven gauges: two in a steering column pod, two in an A-pillar pod-mount, and three above the stereo (if you are planning to upgrade the stock double-decker stereo unit with a single-level aftermarket piece). However, it should be noted that A-pillar pods should not be used on MkIVs equipped with A-pillar airbag systems (like my car), as they will become projectiles during airbag deployment.
A closer look
With the guidance of Rob Leech at Tunerworks Performance Parts House in Calgary, I selected gauges from Auto Meter's Sport Comp line. Having strong confidence in HPA's tuning abilities, I opted not to add EGT and A/F gauges necessary in most custom turbo applications. The simple solution was the Auto Meter two-gauge steering column pod mount with the bare minimum. This comprises a combination vacuum/boost gauge and an oil temperature gauge (see Figure 13).
The extremely stealth R28 Turbo is proudly shown in Figure 14. I would like to say that, as a VR6 and R32 enthusiast, this turbo system is well beyond my expectations. Power delivery is ultra smooth and surprisingly easy to modulate at any engine speed, and it can be absolutely fierce when desired, with around town off-boost driving as silky as the stock VR6 engine (with an aftermarket chip, i.e. no drive-by-wire delay). It is truly like driving a stock car with a 'magic' throttle-controlled power release. Also, in this application, turbo lag is negligible, and gas mileage was over 27mpg (US gallon) during a day of highway driving with many trips into the boost range. With all of this performance, I think 'My R32 Blues' have now completely faded away.
Stay tuned for Part Five, the last installment of the series, when the R28 Turbo gets some new shoes, thanks to Neuspeed and Toyo Tires, and a few more finishing and handling touches. Also, a full track review will be included and the complete specs listed.
The R28 Dyno plot compilation
Sachs/HPA performance clutch assembly
Computer-controlled exhaust by-pass
HPA dogbone mount
Steering column gauge pod
The stealth R28 with 17-inch winter wheel setup