2003 Ford Mustang SVT Cobra Adds Horsepower

After a back-up run verified the power results of the baseline, we opted to start with the time-honored tradition of increasing the boost pressure. To that end, we removed the 3.5-inch blower pulley and replaced it with a 3.0-inch pulley. The pulleys used were prototypes of a system offered by Kenne Bell. As expected, the 3.0-inch pulley offered a dramatic increase in boost pressure by changing the drive ratio from 2.14:1 (with 3.5-inch) to 2.50:1. The pulley change increased the blower speed of the M112 from 13,715 to 16,250 rpm, and the peak boost from 8.6 to 11.9 psi. At 6,000 rpm (near the power peak), the boost pressure increased from 7.8 to 10.3 psi, upping the power peak from 417 to 454 hp. The additional 2.5 psi was worth 37 hp, or 14.8 hp per pound of boost. Though the 3.0-inch pulley increased the peak boost pressure by 3.3 psi, the boost drop was more acute with the smaller pulley. The boost dropped from a peak of 11.9 psi at 4,500 rpm to just 10.0 psi at 6,500 rpm, compared to a drop of just .8 psi with the 3.5 pulley. The additional boost helped power, but the returns were definitely diminishing.

The next item on the list was a Kenne Bell cold air intake system. Knowing that supercharged motors are extremely sensitive to inlet restrictions, we decided to test the effectiveness of replacing the factory air filter assembly. Though our test motor was already equipped with a free-flowing K&N filter, we suspected there was more power to be had by replacing the cone-shaped air box assembly. After producing 454-rear-wheel horsepower, we installed the Kenne Bell air intake system consisting of a billet aluminum adapter (for the MAF), a length of hose (to run into the fenderwell) and a large K&N cone filter. The Kenne Bell system allowed the supercharged motor to breathe cold air from the fenderwell, while minimizing the restrictions inherent in the stock air filter system. Adding the cold air system increased the boost slightly (a good sign) to 12.2 psi and upped the power peak to 473 hp. As you will note from the curves, the majority of the power gains were above 4,500 rpm, where the restrictions would be the greatest.

God bless data logging, as we found out some interesting things during the test sessions at Kenne Bell. One thing that frightened us was the air/fuel ratio, a situation that only got worse once we installed the smaller blower pulley. Naturally we were logging the air/fuel ratio during testing, and the meter indicated a lean condition (at wide open throttle) from 1,500 rpm to 2,500 rpm. From 3,000 rpm to 4,000 rpm, the air/fuel ratio hovered in the mid-12s, dropping down to 11.7:1 at 4,500 rpm. The air/fuel ratio continued its descend with rpm, finishing up at a rich 10.3:1 at 6,500 rpm. Once we installed the 3.0-inch pulley, the situation became much worse, with the blower motor running leaner than 14.2:1 from 2,000 to 4,000 rpm. Remember, the SVT motor makes peak boost at 4,000 rpm. A 14.2:1 air/fuel ratio and 11.9 psi of boost are usually not a good combination. As with the larger blower pulley, the motor richened up with rpm, ending up at 11.5:1 at 6,500 rpm.

Another bit of valuable information gathered by the data logging was the spark curve. Apparently SVT wanted to eliminate any chance of ignition-related detonation. The full-throttle timing curve indicated the supercharged motor ran with just 11 degrees of total timing at 2,000 rpm. By 3,500 rpm, the timing had increased to 15 degrees, winding up at a solid 23 degrees by 6,000 rpm. Adding the smaller blower pulley actually reduced the total timing by 1-2 degrees, at least down low. By 4,000 rpm, the timing curves were identical and remained so through 6,500 rpm. It is logical SVT would choose to run so little timing at wide open throttle, as they have no idea what kind of folks might be behind the wheel of the supercharged Stang. Imagine some idiot pumping the tank full of 87-octane unleaded fuel and then running at full-throttle up a hill in fifth gear. While we all laugh and quickly dismiss such antics, rest assured this situation has occurred, probably more than we would care to admit. The cautious timing curve is also no doubt a function of the lack of a functioning knock sensor. Odd such a motor would not be so equipped.

Regardless, we knew from looking at the timing and fuel curves there was plenty of power to be had from programming a chip. All we had to do was add some fuel at the appropriate rpm points and then add a few degrees of timing and we should have some serious horsepower. It should be noted we did not expect the peak power numbers to change, as both the air/fuel and timing numbers were pretty optimized near the horsepower peak. What we did expect were big-time gains down low and through the mid-range. What the Kenne Bell Performance chip gave us was exactly that. There was no peak power gain, but check out the (chip) graph for the serious gains elsewhere. The chip increased the power output by as much as 35 hp and over 40 lb-ft of torque. The gains were consistent from 2,000 to 4,000 rpm, tapering off until 5,500 rpm. The boost pressure remained at 12.2 psi, while the air/fuel ratio dropped to a much safer 12.2-12.5:1 from 2,000 5,500 rpm, tapering off ever so slightly thereafter.

At this point, we had nearly exhausted our box of bolt-ons, with one exception--we still had one prototype Kenne Bell supercharger pulley left, which measured 2.8 inches. With our safe air/fuel, we felt more than confident installing the smaller blower pulley. Besides, the SVT Cobra comes equipped with one heck of and air-to-water intercooler. Testing has shown the charge temperature can reach near 270 degrees once the boost pressure exceeds 12 psi. The air-to-water intercooler quickly drops the elevated charge temperature down to a more desirable 108-110 degrees. Knowing we were working with such an efficient cooler gave us the confidence to up the boost even further.

Installing the 2.8-inch pulley upped the boost pressure sure enough, but not (unfortunately) the power output. The peak boost pressure jumped to 14.4 psi, but the power peak remained the same at 475 hp. There were some gains registered across the rev range, but the gains were not in line with the higher boost pressure. The peak boost pressure was up by more than 2 psi, but at 6,500 rpm, the boost pressure was up by just 1.3 psi. All signs pointed to the fact that we were getting very close to the flow limit of the M112 Eaton supercharger.

In addition to running out of blower, the data logging also indicated we were topping out the factory 90mm mass air meter. According to the data logging, the mass air voltage peaked at 5 volts by 6,000 rpm. The system was not designed to operate at or above 5 volts, so it is likely more chip tuning may be necessary before we can proceed further. We know the fuel pumps and injectors are good for well over 500-rear-wheel horsepower, especially when augmented by a Kenne Bell Boost-a-Pump. Next time on the dyno, we hope to run some testing on the Ford Racing cat-back exhaust along with a complete header and off-road pipe system. From there, we will venture into the installation of a Kenne Bell replacement supercharger along with a host of other performance goodies, including an AccuFab throttle body and much more boost. Stick around if all goes well, we hope to find out just how much power can be coaxed from a stock SVT long-block.