The real fight is not boost pressure or headline CFM. It is air density and restriction. If we improve the air going into the compressor, the turbo does less work to move the same mass flow. If we reduce backpressure after the turbine and after the DPF, drive pressure drops, exhaust pumping losses go down, and more of the engine’s power stays at the crankshaft. That is why intake design, boost-tube sizing, intercooler pressure loss, and even exhaust-tip shape matter. A part that looks bigger is not automatically better. The right part preserves velocity, reduces pressure loss, and improves density where the engine actually uses it. In the discussion here, the intake and exhaust changes increased airflow enough to lean the air-fuel ratio from about 20:1 to 21:1 while adding power on the same tune. That is the kind of result that shows the engine is breathing easier, not just working harder. The bigger point is simple: stop treating airflow parts like cosmetics. Measure mass flow, pressure loss, and density. When restriction comes out of the system, the engine makes power easier, runs cleaner, and puts less stress on the hardware.
This episode of Speed School features Gale Banks in conversation with Corey Willis of PPEI, a Louisiana-based tuner and calibration specialist. After some brief personal background, the discussion quickly settles into what both men care about most: airflow, combustion, testing discipline, and how modern diesel performance has to be understood as a complete system rather than a collection of isolated parts.
Banks describes why he values conversations with younger tuners like Willis. Experience tends to make engineers more conservative, while younger builders often push into ideas that initially sound risky but can reveal new opportunities. Willis, in turn, explains that studying older engineering principles remains valuable because an engine is still fundamentally an air pump. Whether the platform is old or new, the same core rule applies: get the air into the engine efficiently before adding fuel.
The technical discussion begins with diesel air-fuel limits and the consequences of exceeding them on emissions-equipped trucks. Banks explains that once a diesel drops below roughly a 17.5:1 to 18:1 air-fuel ratio, soot loading rises enough to burden the particulate filter. That increases regeneration frequency, wastes fuel during regen events, and raises the nominal soot percentage in the filter. A dirtier filter also creates more backpressure than a clean one.
That backpressure matters far more than many enthusiasts assume. Banks points out that pressure reduction after the turbine has a multiplied effect upstream. If pressure behind the turbine drops by only a couple of PSI, turbine inlet pressure may fall by 8 to 9 PSI when the turbine expansion ratio is around 4:1 to 4.5:1. At higher power, where expansion ratio increases, reducing downstream backpressure becomes even more important. The result is lower drive pressure, less pumping loss on the exhaust stroke, and more usable power at the crankshaft for the same fuel rate.
Willis confirms this with recent dyno testing. He had previously been skeptical of intake-and-exhaust gains on modern trucks, but after testing Banks components he concluded that the system worked measurably better than he had believed. His results closely matched Banks' own data, within only a couple of horsepower, despite being gathered on separate dynos.
A substantial part of the conversation focuses on test quality. Willis explains that he uses both a Dynojet and a Dimsport dyno, and both are equipped with eddy-current absorbers so they can be properly loaded rather than used only for quick inertial sweeps. His Dimsport has four eddy brakes and can handle very large vehicles, while the Dynojet has two. Because both are calibrated and loaded consistently, they produce repeatable numbers.
He criticizes the common practice of making very short dyno pulls to generate attractive peak numbers before heat has time to build in the system. His normal peak-power pulls last about 14 seconds, and when developing race calibrations he often bases pull duration on expected quarter-mile elapsed time plus another 2 to 3 seconds. For towing calibrations, he may sustain load for more than five minutes.
Banks adds his own example from differential-cover development. During a long-duration pull at roughly 550 rear-wheel horsepower, a stock differential cover allowed lubricant temperature to climb above 300 degrees F, and one of the dyno absorbers eventually caught fire. He notes that sustained testing exposes weaknesses that short pulls hide. He also explains why large single rollers are better for long-duration testing: they flex the tire less than smaller rollers, reducing heat buildup and distortion during extended loaded runs.
Willis then details what he observed when testing a Banks intake and exhaust on an L5P Duramax. Beyond the horsepower increase, he saw that the truck required less turbo vane position to achieve the same desired boost set point. That indicated the turbocharger was not working as hard on the compressor side and that the engine was flowing more freely overall. Lower vane percentage also implied lower drive pressure, so the gains were not just from reduced pipe restriction but from reduced turbocharger effort as well.
Banks explains the mechanism from the intake side forward. Between ambient air and the intake valve, nearly every component except the turbocharger tends to reduce air density. A well-designed intake system, such as Banks Ram-Air, aims to preserve ambient density at the compressor inlet by minimizing temperature rise and pressure loss. If compressor-inlet density improves, the compressor needs less shaft speed to move the same mass flow. That unloads the turbine, lowers drive pressure, and reduces the pumping work required during the exhaust stroke. The horsepower saved from that pumping loss either appears as additional output at the crankshaft or allows the driver to maintain the same speed with less throttle.
Willis provides a concrete example. On a street tune for the L5P, the truck originally made 544 horsepower at about a 20:1 air-fuel ratio. After installing the intake and exhaust, the same tune produced 564.67 horsepower while the air-fuel ratio leaned out to about 21:1. In other words, the truck made over 20 additional horsepower while using the same calibration and becoming leaner. To bring power back down to 544 horsepower, he would have to remove fuel. Banks emphasizes that comparing systems at the same air-fuel ratio is the proper way to isolate what the hardware changed. As the mixture leans out, exhaust gas temperature also drops, improving safety margin.
The discussion then narrows to one of Banks' more unusual claims: that the exhaust tip itself can materially affect performance. He explains that the stock Chevrolet and GMC L5P tailpipe tip is highly restrictive because it uses a brute-force method of blending ambient air with hot exhaust gas. Banks wanted a design that would reduce backpressure after the DPF without relying on restriction.
His first prototype was a conventional round air-blend tip with slots that mixed ambient air into the exhaust stream. It worked, but it did not create the negative pressure he wanted. He then developed an obround, slash-cut tip that expands the gases and uses vehicle motion to create a scavenging effect. According to Banks, this design can actually pull a slight negative pressure in the tailpipe, helping evacuate the system after the DPF.
He explains that the tip geometry depends on maintaining sufficient gas velocity while also maximizing outlet surface area. Packaging matters too: a 5-inch tube is difficult to route cleanly over the axle, and bend radius limitations on production tube benders constrain what can be built without damaging the material or hurting flow. The final tip shape balances packaging, velocity, and scavenging behavior. Willis admits that years earlier he would have argued that tips were purely aesthetic, but after seeing the data and hearing the design rationale, he now accepts that the tip can be a functional flow device rather than decoration.
Banks next turns to a broader measurement problem: the industry standard of rating airflow in CFM rather than mass flow. He is reworking his flow-bench methodology to measure what actually matters to engine performance. Traditional benches report cubic feet per minute, but CFM alone does not capture air density. Since engines burn air mass and fuel mass, the meaningful quantity is pounds per minute, not just volume.
He describes using a mass airflow sensor with the flow bench to measure density loss through components such as the stock air horn, stock heater, and Banks Monster-Ram. The goal is to quantify how much mass airflow improvement the upgraded hardware provides while retaining emissions-legal features such as the cold-start heater. On a Cummins HO application, he cites a measured stock flow of about 52.4 lb/min at 2,800 rpm for the 400-horsepower engine, then compares performance at 60 and 70 lb/min as tuning and hardware increase demand.
Banks frames every component as either a density machine or a density throttle. Turbochargers, superchargers, and intercoolers can increase density; restrictive ducts, horns, and poorly designed charge-air components reduce it. He argues that enthusiasts often fixate on boost pressure when density is the real target. Pressure is only one factor in density; temperature and humidity matter too. The engine displaces a fixed volume every two crank revolutions, but the mass contained in that volume depends on manifold density.
This leads into charge-air system design. Banks says oversized boost tubes do not meaningfully hurt throttle response when their added volume is compared with engine displacement. On a Cummins, adding the equivalent of two cylinders' worth of volume to the boost tubes represents only about two-thirds of a crank revolution, which is not perceptible in normal driving. The larger gain comes from reducing throttling losses. He also notes that the intercooler is often the biggest offender in the system: it helps density by cooling the air but hurts density through pressure drop. A well-designed charge-air cooler must exceed stock cooling performance while reducing pressure loss, and in some cases it can improve density more than the compressor itself.
Willis reinforces Banks' density argument with examples from both diesel trucks and side-by-side racing. He often tells customers that not all 1,000-horsepower combinations are equal. One setup may reach the number by forcing air through restrictive hardware and wasting energy, while another could make the same power more easily, with less fuel, lower stress, and more headroom. In his view, if a truck makes 500 horsepower with a tune alone, adding the right airflow components may allow it to make 530 or 540 horsepower with the same fuel quantity while improving fuel economy and durability.
He also describes testing more than six intercoolers on Can-Am vehicles after leaving a previous Banks visit with density on his mind. The team eventually found one intercooler that clearly outperformed the others. At the next race, they ran 2 PSI less boost yet achieved stronger launches and better overall performance. Lower boost meant less engine stress, but improved intercooler efficiency allowed more airflow and more power.
At the same time, Willis cautions that bigger is not always better. One large custom intercooler looked impressive and performed reasonably well at high speed, but on the low end its added volume badly hurt throttle response on the small engine. That experience reinforced the idea that there is always an optimum size range. Banks agrees and recalls a customer insisting on a 4-inch-thick intercooler simply because it sounded desirable. Effective intercooler design requires balancing cooling effectiveness, pressure drop, and system volume rather than chasing size alone.
The conversation briefly shifts to Banks' PedalMonster and the broader issue of throttle mapping. Banks admits he was initially skeptical of throttle boosters, believing drivers could simply press the pedal farther if they wanted more response. After driving the prototype, however, he concluded that a properly designed system genuinely improves drivability by removing lag.
What changed his mind was not just quicker response but the shape of the pedal curve. Many throttle boosters compress the usable range so aggressively that full throttle arrives early, leaving a dead-pedal zone from roughly 50 to 100 percent pedal travel. Banks says PedalMonster was designed differently: 0 percent and 100 percent remain in the same pedal positions, but the rate of change between them is reshaped so the pedal stays linear and usable across the full range. He also notes safety features, including bypass behavior in the event of failure and stock-like reverse response to avoid overly aggressive backing maneuvers.
Willis says this product changed customer expectations enough that he had to revisit his own calibrations. Once drivers became accustomed to improved pedal response, they perceived stock-style mapping as laggy even when engine output had not changed. He responded by rescaling portions of his throttle tables, particularly in the midrange, to restore the sharper feel customers now expected.
A major portion of the episode addresses emissions compliance, tuning legality, and the future of the aftermarket. Both men reject the idea that emissions-compliant performance is inherently incompatible with meaningful power gains. Banks notes that modern tuning, intake systems, exhaust systems, turbocharging, and intercooling can produce substantial gains while remaining legal. Willis agrees and says 600 to 700 horsepower is now achievable on emissions-compliant vehicles, which is more performance than many enthusiasts seem to appreciate.
Banks argues that many complaints about DPF systems are really complaints about poor tuning. If a calibrator dumps excessive fuel into the engine, soot loading rises, regeneration frequency increases, and the owner blames the emissions hardware rather than the bad tune. Willis adds that calibrators must understand combustion fundamentals, especially NOx formation, particulate matter, hydrocarbons, oxygen content, and pressure relationships across the exhaust system. Simply turning emissions functions back on in a former delete file does not produce a correct calibration.
Willis says that pursuing compliance forced him to become a better calibrator. Learning what drives NOx, soot, and other emissions sharpened his understanding of diesel combustion. He advises tuners to begin by fingerprinting a stock vehicle: measure oxygen content, soot load, and pressure behavior in factory form, then modify the calibration while keeping those parameters in a healthy range. Even without the most expensive equipment, that discipline leads to better results than blindly copying aggressive files.
Both men also discuss the business and regulatory side. Banks recounts how, in the late 1970s, work on a turbo package for the Pontiac Sunbird led him and Hugh MacInnes into repeated meetings with the California Air Resources Board. Those efforts helped establish the EO process under Vehicle Code 27156, which still governs legal aftermarket emissions parts. Willis says that shouting at regulators accomplishes little; progress comes from persistence, technical credibility, and working to improve the process. He notes that more manufacturers are now seeking compliant turbochargers and hardware, and that the requirement to gather real data is pushing product quality upward.
The episode closes with Banks reaffirming his support for the RPM Act and clarifying that he has long worked with SEMA on emissions and racing issues. He distinguishes between deleting emissions equipment from street vehicles and building dedicated race or military engines that were never emissions-equipped in the first place. His broader point is that the aftermarket survives by understanding the rules, improving the engineering, and continuing to fight for workable paths that preserve both performance and legality.