The first time we put this supercharged Ford flathead on the dyno, it didn’t end the way anyone wanted.

The engine was climbing toward the 280-horsepower number Steve had predicted when the head gasket let go at 233 horsepower. One cylinder disappeared, the engine was running on seven, and the test was over.

That happened in May.

Four months later, Steve brought the same engine back—not with a convenient story about discovering one bad part, but with an honest admission:

He couldn’t say with certainty what caused the failure.

What he could do was improve every area that might have contributed to it, return to the same dyno and put the engine back under load.

That is where our upcoming video begins.

There are plenty of dyno videos built around one big number. This one is better than that.

It is a failure-and-redemption story, but it also shows what a dynamometer is actually for. Before this engine ever reached its destination—a 1925 Ford Model T nostalgia race car—the dyno exposed a cylinder-head machining problem, an ignition problem, two reversed plug wires, a cracked distributor cap and finally a failed head gasket.

Finding those things in the dyno room was frustrating. Finding them after the engine was installed in the car would have been much worse.

As Robert says in the video, there is a reason they call the dyno a “torture chamber.”

The changes Steve made are easy to miss because none of them looks particularly dramatic on camera.

The new head gaskets are three-layer Cometic MLS—or multi-layer steel—gaskets. You can identify them by the small tabs protruding from beneath the cylinder heads.

Steve also stoned and lapped the block and head surfaces, changed to thicker machined Edelbrock 1125 heads and fitted a known-good Stromberg E-Fire distributor. The ARP chromoly head studs were already on the engine when the gasket failed in May, so they were not one of the September changes.

The important detail is that Steve never presents any one of those parts as a magic cure. He improved the sealing surfaces, gasket construction, cylinder heads and ignition, then let the dyno determine whether the complete package worked.

That is how careful engine work is supposed to be done.

September also brought a fuel change: aviation gasoline replaced pump gas. It provided additional detonation margin during the dyno session and is also what Steve keeps in the Model T because it holds up better during storage.

Early in the process, Steve predicted that the engine would make approximately 280 horsepower.

On its successful September return, the Brawler four-barrel combination produced:

Watch Steve’s reaction when the corrected number comes up. It isn’t a staged celebration. It is the quiet satisfaction of someone seeing four months of work validated—and realizing his original estimate was almost exactly right.

Also listen carefully whenever “corrected” and “uncorrected” horsepower are discussed. Dyno numbers are affected by atmospheric conditions, so the corrected result is the proper figure for comparing performance. The biggest number visible on a screen is not automatically the number that belongs in the headline.

Here is another detail that will pass quickly: after the Brawler produced 285.1 horsepower, Robert installed his personal Holley 650 racing carburetor. That combination reached 296.6 corrected horsepower—the 296-horsepower result Steve now uses when describing the test.

The first supercharger is a Weiand 142. Later, the team changes to a Weiand 174. Those numbers describe the supercharger’s approximate displacement in cubic inches per revolution—not the engine’s horsepower.

The 142 was initially used with the Brawler because it was already a known, properly jetted combination. That removed one variable while everyone determined whether the repaired engine was healthy.

Once the engine proved itself, they installed the combination Steve actually wanted in the Model T: a Weiand 174 with two Big Stromberg 97 carburetors.

Robert’s racing carburetor made the biggest number, but it was never the final objective.

Steve wanted a configuration that looked appropriate in his nostalgia race car. The twin Strombergs delivered that—and still made approximately 271 horsepower at about 5,400 RPM.

That roughly 25-horsepower difference is real, but it does not make the twin-carburetor test a failure. The engine survived repeated pulls, the Strombergs were successfully tuned, and Steve ended with the period-correct combination he actually wanted in the car.

Sometimes the best dyno result isn’t the largest number. It is the combination that fulfills the purpose of the build.

Near the end, there is a deceptively interesting experiment involving the scoops on the two Strombergs.

From their appearance, you might expect the front scoop to obstruct airflow into the rear carburetor. Instead, the first result surprises everyone. The discussion turns to airflow, turbulence and reversion—the back-and-forth movement of the air-fuel charge created by camshaft timing.

They repeat the test and find only about a one-horsepower difference. In practical terms, the scoops impose no meaningful power penalty.

It is a small moment, but it is exactly why testing matters. A part that looks restrictive may behave differently once airflow, reversion and the shape of the inlet are involved.

Steve’s father is in the control room during the successful session.

Steve explains that the two of them have spent roughly 50 years working on old cars, engines and flatheads together. He isn’t presented as a major storyline, but watch his expressions during the final pulls and the 296-horsepower result.

Behind all the machinery, graphs and technical discussion, that may be one of the best parts of the entire video.

In May, the dyno found the engine’s weak link.

In September, the 59A-B came back with a stronger sealing strategy, survived the test, exceeded Steve’s prediction and proved the twin-Stromberg combination that will actually go into the Model T.

The number is impressive.

Finishing the job is the real story.

Coming soon to Barry T’s Garage.