Tom Wilson
January 27, 2010

Horse Sense:
In our first article on MagnaFlow's 3-inch exhaust, we mentioned the interior of Jason Cenora's '06 Mustang GT. We were just having fun, but Jason's car was torn up inside because it was in the middle of a makeover. That didn't stop him from feeling bad about it. This time, let us make a special note of saying the interior was transformed beautifully for our second test. In fact, it's so nice, you'll be seeing more of it in a feature article.

Time flies when you're having horsepower. While it seems that only last month we were trying MagnaFlow's 3-inch exhaust for S197 GTs on the dyno, it was really back in our February '07 issue when we made our initial report on these dreamy, big pipes ("Big Three," p. 122).

Or were they pipe dreams? In our first review, we lost power almost all the way across the powerband when we replaced the stock exhaust with MagnaFlow's 3-inch system. The only gain was at the power peak, where we saw an extra 19 hp-not enough to offset the string of double-digit losses on the way up the tach. Clearly, something was wrong. Sure enough, the MagnaFlow-equipped engine with the Novi 2200 was able to process so much more air, it was running lean after we bolted on the MagnaFlow exhaust system. If you read the story, you may recall how we easily spotted the lean air/fuel mixture, but didn't have any tuning capability on hand that day to do anything about it. We decided to have another dyno day and see what gains could be had with some tuning.

Today is that day, but before we verbally spin the dyno rollers again, a quick review of the test car and MagnaFlow exhaust is in order. The car is Jason Cenora's blue '06 GT, a showy number he uses to promote Hillbank Motorsports, a mail-order speed shop. While this car is dripping with neat details from the headliner to the stereo, in the powertrain department the players include only a Novi 2200 supercharger with air-to-air charge cooling. An electric water pump has been on board since day one, which was the first test since the fuel rails were upgraded. The fuel injectors, pumps, controller, and all other fuel-system components-save one-were left unchanged. They were same in both tests and therefore shouldn't influence the testing-neither should the larger fuel rails. We'll talk more about the change in the fuel system soon.

MagnaFlow's exhaust system is a 3-inch mandrel-bent assembly of two separate kits. A cat-equipped X-shape crossover makes up the first part; an after-cat with 5-inch diameter and round, flow-through mufflers make up the second half. Both halves are gorgeous, pinch-free, stainless steel parts designed for maximum airflow in high-output engines. Practically speaking, that means well-boosted, supercharged Mustangs as MagnaFlow has other, less pricey kits for stock, bolt-on, and lightly boosted blower cars. As it is, the 3-inch system can be found on the street for slightly more than $1,300, which is a solid hit to the wallet. But it's bright stainless steel, mandrel-bent, and wears two catalytic converters.

We must also note that although the system sports a pair of cats, it isn't CARB-exempted. Most importantly to us, it scrubs the exhaust so it minimizes environmental damage, even if it isn't a legally exempted part. We expect law enforcement may have other priorities.

For our retest, MagnaFlow brought in tuning whiz Steve Ridout of Powertrain Dynamics. Steve is an old friend; he has been e-tuning Fords on his Dynojet in Huntington Beach, California, seemingly forever, going well back into our Super Ford days. We knew that as long as we didn't detour into discussions on surfing, Steve would have this tuning situation licked in short order.

With Jason's car strapped to MagnaFlow's Dynojet chassis dyno, the car was quickly re-baselined as a check against the results obtained in our first test, which had been run several months earlier. In the previous test, the stock-exhaust baseline peaked at 395 rwhp, literally just hours after the Novi 2200 supercharger had been installed and driven to the MagnaFlow test center. That combination was green, including the base car, which had only a couple hundred miles on it at the time.

After bolting on the MagnaFlow exhaust in the first test, we saw 411 rwhp at the peak. Below that, the system lost power compared to stock because the engine was running lean.

For the second test we're reporting on here, the baseline run was with the 3-inch MagnaFlow exhaust installed. That enabled us to see if there was any change in the same run from the previous 411-rwhp test of the same combination. Steve found that although the car was alternating between rich and lean throughout, its basic fuel curve wasn't bad. At its richest, the A/F ratio was in the low 10:1 range, while the lean sections topped out at 12.2:1. Steve immediately noted that the run tended to lean at high rpm. He knew he could pick up some power, but he didn't think it would be a huge amount more than the baseline already gave us.

That was OK because this new baseline run was far better than the one during the original testing months earlier. When we say far better, we mean 22 hp to the good, topping out at 433 rwhp. What was different between the baseline run from the first test session and our runs for this retest article to get such a huge gain?

Earlier we said all the fuel system components-save one-were the same. The new fuel system component was the gasoline. The first tests were run on 91-octane pump gas. For the second test, Jason had filled the tank with 100-octane unleaded Rockett Brand racing gasoline. Why? Because Rockett is a new sponsor of his. It's a good reason, for sure, but it did introduce a variable.

While octane itself doesn't make one whit of horsepower, our first test was run in September, and that means the 91-octane pump gas used during that test was summer-blend California fuel, which is 3 percent oxygenated. It has less energy density and runs 3 percent leaner than the Rockett race gas, accounting for some power right there.

Extra octane helps by supporting more ignition timing before detonation sets in. As our quick tests weren't sophisticated enough to lock the ignition timing at one value, the higher-octane Rockett racing gas could have been supporting an extra degree or two of ignition timing. That would easily increase 20 hp worth of power.

Secondly, it was warm during the first tests in September-96 degrees-while the second round of tests was run in February with 30-degree-cooler air. Interestingly, the humidity was nearly the same at approximately 20 percent both days. The correction factor is supposed to account for the difference in air density, but this may have been enough to keep engine temperatures cooler, thus keeping the Mustang's com-puter from pulling ignition timing. In any case, the cooler air meant more raw power was being made.

Another variable was the mileage on the blower installation and car overall. As noted, the blower had been on the car about two hours and the car had less than 300 miles on it when we first tested it, so it was still tight. Before our retest, Jason racked up a couple thousand miles on the S197, so the engine was freed up from a tolerance point of view. Furthermore, the adaptive electronics in the engine management computer had a chance to tune to the specific engine package. Both factors contributed to the power increase observed over the two baseline runs.

Add all the factors up-the non-oxgenated, higher octane fuel combined with the uncontrolled ignition timing, greater air density, and significant time for adaptive learning to take place in the computer-and the 22 rwhp increase in baseline is accounted for.

With that explained, Steve began to tune. He had no more noted that Paxton's factory tune on the supercharger kit was fairly good when he found out the Diablo tuner and software already installed didn't have the correct template from DiabloSport for Jason's car. There are several software templates Ford uses for the S197s and a quick trip to Diablo's Web site had the necessary template downloaded. Steve finally changed fuel and spark parameters.

Steve had an interesting way of describing his tuning process. He tunes for the air/fuel ratio and sees power gains as a byproduct of managing the fuel. The same is true of the spark, but to a lesser extent, as it's dangerous to be greedy with ignition timing on a supercharged car. Thus, Steve was keen to examine the datalogging of the air/fuel ratio provided by the Diablo tuner, and he didn't pay much attention to the power figures.

Steve's goal for a blower car such as this is mid-11 A/F readings. He believes the seemingly rich mixture is the best compromise between power and safety (no detonation). Detonation is to be avoided at all costs on the Three-Valve engine. Similar to most V-8s in Ford's modular engine family, the Three-Valve piston is weak in the top piston-ring groove and will break off a chunk of the upper-ring groove/piston dome at the first hint of supercharge-assisted detonation.

In practice, the ideal mid-11 A/F readings begin at 12:1 mixtures at lower rpm, then richen to 11:1 for the majority of the tach. They top out at a fat 10:1 A/F ratio at high rpm for maximum detonation protection at the high cylinder pressures and temperatures associated with the power peak.

After several runs sneaking up on the fuel curve, Steve blasted a 451-rwhp pass out of Jason's car for a solid 18-rwhp gain over the day's 433-rwhp baseline. That was on a cooled run, but not iced or otherwise tricked up. It shows the power of electronic tuning, not to mention getting everything to run right with the correct air/fuel ratio.

How much did the MagnaFlow exhaust pipes gain compared to stock? After all this, we can't say directly. The stock pipe combination was never optimized via electronic tuning, so we can't precisely separate the gains from the exhaust pipes from those gained electronically. If we take the original 395-rwhp figure from the Sep-tember test and add 22 hp to it, figuring that was the same tuning advantage from the better fuel/more advanced ignition timing our February test benefited from, a 417-rwhp baseline for the stock exhaust pipes is achieved, had that combination been e-tuned. That would give the Magna-Flow 3-inch exhaust system credit for a stout 34-rwhp gain at the 451-rwhp reading. That's an excellent gain for an exhaust system on a street-driven car, and it seems right for a blown engine at this power level.

It's too bad we didn't have electronic tuning available when the MagnaFlow exhaust first went on Jason's car. But who would have thought a computer would be needed to install a set of exhaust pipes? When they're 3 inches in diameter and handling supercharged airflow, apparently one is.

2,800 159.52 299.22 163.45 306.60 160.77 301.57 171.20 321.13 11.68 21.91
3,000 167.16 292.65 168.58 295.13 171.86 300.88 183.42 321.12 16.26 28.47
3,200 180.68 296.54 181.43 297.78 187.51 307.77 200.18 328.56 19.50 32.02
3,400 200.92 310.35 200.69 310.00 212.78 328.69 229.02 353.76 28.10 43.41
3,600 227.19 331.44 229.49 334.79 239.70 349.70 258.07 376.51 30.88 45.07
3,800 245.51 339.33 240.59 332.52 259.01 357.99 279.01 385.62 33.50 46.29
4,000 266.04 349.31 256.15 336.33 274.70 360.67 301.12 395.37 35.08 46.06
4,200 284.47 355.72 277.00 346.38 299.19 374.14 326.79 408.65 42.32 52.93
4,400 307.23 366.73 297.70 355.35 316.90 378.28 347.49 414.80 40.26 48.07
4,600 318.77 363.96 311.02 355.12 324.85 370.91 357.25 407.89 38.48 43.93
4,800 333.14 364.52 318.24 348.21 342.01 374.22 376.10 411.52 42.96 47.00
5,000 350.43 368.09 336.91 353.89 369.99 388.64 397.91 417.96 47.48 49.87
5,200 364.75 368.41 353.06 356.59 390.02 393.93 412.47 416.62 47.72 48.21
5,400 377.47 367.14 358.84 349.01 401.29 390.30 417.02 405.61 39.55 38.47
5,600 393.38 368.94 378.56 355.05 402.24 377.25 436.34 409.23 42.96 40.29
5,800 395.01 357.69 394.21 356.97 418.38 378.87 451.38 408.73 56.37 51.04
5,900 n/a n/a n/a n/a 433.44 379.44 n/a n/a n/a n/a
6,000 n/a n/a n/a n/a 334.12 284.35 n/a n/a n/a n/a

All data was taken on MagnaFlow's in-house Dynojet chassis dyno. The ultimate power achieved, 451 rwhp, dream-wheels out at 519 hp at the flywheel using the industry standard guesstimate of 15 percent drivetrain loss. That's 1.8 hp/ci or 185 hp/liter. Those are nice numbers for a mass-produced engine augmented by a bolt-on supercharger kit.