THE 9000 RPM ENGINE HONDA ALMOST BANNED! - AND THE TEST DRIVER WHO SAVED IT!

Engine in Focus
4 followers
Honda didn’t just build a high-revving engine…
They built a 9000 RPM monster that pushed their own engineers, test drivers, and reliability limits to the edge.
And when corporate saw what it was capable of, they didn’t celebrate…
They panicked.
In this video, we uncover the untold story behind the 9000 RPM engine Honda almost banned — the secret testing, the internal pressure to shut it down, and the one test driver who refused to let it die.
Because without him… this engine might’ve never made it into production.
And the entire JDM performance world would look completely different today.
🔥 If you love VTEC legends, hidden engineering stories, and engines that were too good for their own company… you’re going to love this one.
#machine #engine
April 1999. A Honda test facility in Tochigi, Japan.
Lead engineer Shigeru Uehara stands in front of a soundproof dino cell,
watching through reinforced glass as a prototype engine screams to life.
The tachometer climbs. 7,000 RPM. 8,000. 8,500. 9,000.
The noise is unholy. A mechanical shriek that sounds like metal tearing itself apart.
Every engineer in the room winces, waiting for the inevitable explosion.
But it doesn't come. The engine just keeps screaming, holding 9,000 RPM like it's nothing.
Uehara checks the monitor. Zero valve float. Zero harmonics.
The engine is making 247 horsepower from 2 liters. Naturally aspirated. That's 123.5
horsepower per liter, the highest specific output of any naturally aspirated production
car engine ever built. One engineer breaks the silence. This shouldn't be possible.
Uehara nods. He knows. According to every textbook, every formula,
every piece of conventional wisdom, what they'd just built violates the fundamental laws of
engine physics. Piston speeds at 9,000 RPM should destroy the bottom end. Valve springs should
float. The harmonics alone should shake the engine to pieces. But Honda's engineers hadn't
read those books. Or they had. And they'd decided the laws of physics were more like guidelines.
This is the story of the F20C. The engine that shouldn't exist. The motor that rewrote what
was possible from a 4-cylinder. And why Honda's own engineers called it the Apex before intentionally
making sure they could never build anything like it again. The NSX Problem and Honda's Identity
Crisis. To understand why Honda built the F20C, you need to understand the existential crisis
they faced in the mid-1990s. The NSX, Honda's aluminum-bodied supercar, had launched in 1990
to critical acclaim. It was beautiful, reliable, and easy to drive. But it had one problem. It
wasn't fast enough. The NSX's C30A V6 made 270 horsepower from 3 liters. Impressive for 1990.
But by 1995, the Ferrari F355 made 375 horsepower. The Porsche 911 Turbo made 400. Even the Dodge
Viper, a crude American muscle car, made 450. Honda couldn't compete on power. They didn't
have a V8. They didn't have forced induction in their sports cars. Their philosophy was high-revving,
naturally aspirated engines. Precision over brute force. But precision wasn't selling. The NSX was
a commercial failure. Beautiful, brilliant, and completely overshadowed by cars with bigger
numbers on their spec sheets. Honda's management faced a brutal reality. Their engineering philosophy,
the thing that defined their brand, wasn't enough anymore. Customers didn't care about VTEC or
aluminum construction. They wanted horsepower. And Honda didn't know how to give it to them
without abandoning everything they believed in. Then someone had an idea. What if they took their
high-revving philosophy to its absolute extreme? Not just high RPM, but impossibly high. So high
that the specific output would shame every competitor, even the Ferraris and Porsches
with their exotic V8s and V10s. They couldn't build a big engine, but they could build the
most advanced small engine in history. The project was called Dream 2000. The mission?
Create a two-liter engine that would rev to 9,000 RPM and produce over 120 horsepower per liter.
Something nobody had ever done in a production car. When the brief hit the engineering department,
most people thought it was a joke. 9,000 RPM was Formula One territory. Production cars didn't
rev that high. They couldn't. The mechanical stresses would destroy the engine in hours.
But Uehara, the man who'd led NSX development and revolutionized aluminum chassis construction,
believed it was possible. And he assembled a team of Honda's most brilliant engineers to prove it.
What they were about to attempt would require reinventing almost every component of a
traditional piston engine. The piston speed problem nobody could solve. The first problem
was basic physics, piston speed. When an engine revs, the pistons move up and down. The faster
the engine revs, the faster the pistons move. This is called mean piston speed, measured in
feet per second or meters per second. For most production engines, mean piston speed tops out
around 4,000 feet per minute. That's considered the safe limit before mechanical stress becomes
dangerous. Racing engines push to 4,500, maybe 5,000 feet per minute. At 9,000 RPM with an 84
millimeter stroke, the F20C's pistons would be moving at 4,960 feet per minute. That's faster
than most Formula One engines of the era. At those speeds, several things happen, all of them bad.
First, the connecting rods experience massive tensile and compressive forces.
They're being stretched and compressed thousands of times per minute with enough force to literally
rip them apart. Second, the piston rings can't seal properly. At extreme speeds, the rings flutter
and lose contact with the cylinder walls. Compression drops. Oil gets past the rings.
The engine loses power and starts destroying itself. Third, the valve train can't keep up.
Valve springs, even strong ones, can't close the valves fast enough. The valves float,
staying open when they should be closed. This causes catastrophic collision between pistons
and valves. Every engineering textbook said the same thing. You can't safely rev a two-liter
four-cylinder to 9,000 RPM with a production viable lifespan. Honda's answer, watch us.
The impossible valve train. The valve train was where Honda performed their first miracle.
To rev to 9,000 RPM, they needed valves that could open and close 75 times per second,
each valve, without floating, without breaking. The solution was a complete reinvention of valve
technology. First, they made the valves as light as physically possible. The intake valves were
hollow, filled with sodium for cooling. The stems were narrowed to the minimum diameter that could
survive the stress. Every gram mattered because lighter valves meant less inertia, which meant
less force required to open and close them. But light valves alone weren't enough. They needed
valve springs that could handle the load. Traditional valve springs couldn't do it.
At high RPM, they'd resonate, vibrate, and eventually fail. Honda's solution was beehive
springs. Instead of cylindrical springs with constant diameter, they used conical springs
that narrowed at the top. This shape change reduced the spring's mass while maintaining
strength. Lower mass meant higher natural frequency, which meant the springs could cycle
faster without resonating. But even beehive springs had limits. So Honda added a secret weapon,
surface treatment. Each valve spring was shot peened, a process where tiny steel balls are
blasted at the spring's surface at high velocity. This creates compressive stress in the outer
layer of the metal, making it incredibly resistant to fatigue. Then they coated the springs with a
low friction material to reduce wear. The result was a valve spring that could cycle 75 times per
second, millions of times per year, without failing. Next problem, the camshafts. At 9,000 RPM,
the cams were spinning at 4,500 RPM. The lobes had to push the valves open and then let them
close 4,500 times per minute, with precision measured in thousandths of an inch. Honda used
a special steel alloy for the camshafts, then hardened them through a nitriding process.
The cam profile itself was designed using computer modeling that didn't exist when
Honda started the project. They had to develop the software themselves. The cam lobes were
aggressive, very aggressive. But they were shaped to minimize stress on the valve train.
The opening ramp was steep enough to get the valve open quickly, but smooth enough to avoid
shock loading. The closing ramp let the valve spring close the valve without bouncing.
And VTEC, Honda's Variable Valve Timing System, was redesigned specifically for the F20C. Instead
of the traditional two-stage VTEC, they used a system that progressively changed valve lift and
duration based on RPM. At low RPM, the engine behaved like a normal motor. At high RPM,
it transformed into a racing engine with wild cam profiles. But the most insane part,
the tolerances. The valve clearances were set to tolerances of 0.001 inches. The cam timing was
accurate to within one degree. The valve spring tension was measured in individual pounds.
If any component was off by even a tiny amount, the engine would destroy itself.
Honda's quality control was obsessive. Each engine was individually hand-assembled.
Each valve was measured. Each spring was tested. Each cam was verified. It took three times longer
to build an F20C than a normal Honda engine. But it worked. Against all odds, Honda had built a
valve train that could survive 9,000 RPM. The bottom end from hell. While the valve
train engineers were performing miracles, the bottom end team faced their own impossible
challenge. How do you build pistons, rods, and a crankshaft that can survive the mechanical
violence of 9,000 RPM? The answer was obsessive engineering and exotic materials. The pistons were
forged aluminum. Nothing unusual there. But the design was revolutionary. Honda used a slipper
skirt design where the sides of the piston were cut away to reduce weight and friction. The piston
pin was offset to reduce piston rock during the power stroke. But the real innovation was in the
piston rings. Honda developed ultra-thin piston rings, much thinner than typical production
engines. Thin rings have less mass, which means they can follow the cylinder walls more accurately
at high RPM. Less ring flutter means better sealing, which means more power and less oil
consumption. The problem with thin rings is they're fragile. So Honda used a special steel
alloy with a molybdenum coating for lubricity. These rings were expensive. Really expensive.
But they worked. The connecting rods were works of art. Forged steel, cracked split for perfect
mating surfaces, and then polished to reduce stress risers. Each rod was balanced to within
0.1 grams. That's insane. Most production engines balance rods to within 5 to 10 grams.
Honda was targeting 0.1. Why? Because at 9,000 RPM, even small imbalances create massive
vibrations. Those vibrations cause harmonics, which cause fatigue, which causes failure. The
rod bolts were ARP quality hardware, torqued to spec, and then measured for stretch to ensure
proper clamping force. Honda didn't trust torque specs alone. They verified everything.
And the crankshaft. Oh, the crankshaft. The F20C's crank was forged steel, fully counterweighted,
and micro-polished. The journals were nitrided for hardness. The oil passages were chamfered and
deburred to optimize flow. But the critical innovation was the harmonic balancer. At high
RPM, crankshafts experience torsional vibration. The front of the crank twists one direction,
while the rear twists another. Over time, this vibration causes metal fatigue and failure.
Honda's solution was a sophisticated harmonic damper that absorbed these vibrations. It was
designed specifically for the F20C's RPM range, and couldn't be used in any other engine. The
entire bottom end was balanced as an assembly, not individual components. The whole rotating
assembly was spun on a balance machine and measured. If it was off, they disassembled it
and started over. This is why F20C engines cost so much to build. The engineering was absurd.
The quality control was obsessive. But the result was a bottom end that could rev to 9,000 RPM
all day long and ask for more. The block that shouldn't work. Here's where things get weird.
With all these exotic materials and space-age engineering, you'd expect the engine block to
be some exotic aluminum alloy or even magnesium. Nope. Aluminum. But not fancy aluminum. Just
Honda's standard die-cast aluminum alloy. The same stuff they used in Civics. This confused everyone.
How could a standard aluminum block survive the cylinder pressures and mechanical stress
of a 9,000 RPM engine? The answer was in the design, not the material. The F20C block used
what's called an open deck design, where the cylinder sleeves are exposed at the top. This
is usually considered weaker than a closed deck design. But Honda had a reason. Open deck designs
allow better coolant flow around the cylinders. Better cooling means more consistent temperatures.
More consistent temperatures mean less thermal stress. Less thermal stress means less warping
and longer life. But they reinforced everything else. The main bearing caps were ladder frame
design, where a plate tied all the caps together. This created a rigid structure that prevented the
bottom end from flexing under load. The cylinder sleeves were fiber reinforced metal matrix
composite. Fancy words for aluminum with ceramic fibers embedded in it. This made the sleeves
incredibly wear resistant while keeping them lightweight. And the deck surface, where the head
bolts to the block was precision machined to tolerances measured in microns. The head gasket
was a multi-layer steel design that could seal the combustion chambers even under extreme cylinder
pressures. But the genius move was the bore and stroke. Honda chose an 87 millimeter bore and 84
millimeter stroke. That's an over square design, where the bore is larger than the stroke. Why does
this matter? Because shorter stroke means lower piston speeds for a given RPM. By keeping the
stroke short, Honda reduced the mean piston speed from instantly fatal to merely insane. This also
allowed for larger valves. With an 87 millimeter bore, there was room for massive valves that could
flow huge amounts of air. And air flow is everything in a naturally aspirated engine.
The oiling system that defied gravity. At 9,000 RPM, oil pressure becomes critical and weird.
Because the oil isn't just lubricating, it's trying to escape through every possible route
while being flung around by centrifugal force. Honda's solution was an oiling system that
First, they used a high volume oil pump driven directly off the crankshaft, not a chain. Direct
drive. This ensured consistent oil pressure at all RPMS. The oil passages in the block were larger
than normal. Bigger passages mean less restriction, which means better flow. The main bearing journals
had cross-drilled passages that fed oil to the rod bearings. The rod bearings had additional
passages that sprayed oil on the cylinder walls and piston skirts for cooling. But the critical
innovation was the oil pan. At high RPM, especially during cornering, oil sloshes around. If the oil
pump pickup gets uncovered, even for a second, you get air in the system. Air in the oil means
loss of pressure. Loss of pressure means bearing failure. Bearing failure means grenaded engine.
Honda designed a baffled oil pan with trap doors that prevented oil from sloshing away
from the pickup. They added windage trays to prevent the crankshaft from whipping the oil
into foam. And they used a high capacity oil cooler to keep temperatures under control.
The F20C held 5.3 quarts of oil, more than most 2-liter engines. And Honda specified 10W30 weight,
thicker than the 5W20 used in economy cars. Because at 9,000 RPM, you need oil that can
maintain its viscosity under extreme shear stress. Oil change intervals were 3,750 miles.
Not because the oil broke down, but because Honda didn't trust customers to check their oil level.
If you ran the F20C a quart low, it could starve for oil during hard cornering. So,
they made you change it frequently. The intake that breathed like nothing else.
All the bottom-end engineering in the world doesn't matter if you can't get air into the
engine. And at 9,000 RPM, the F20C needed to inhale 150 cubic feet of air per minute.
For reference, that's about the same as a small leaf blower. The intake manifold was a tuned
length design, where the runner lengths were calculated to create resonance effects at
This is called Helmholtz resonance, and it's incredibly difficult to tune properly.
Honda used computer modeling to optimize the runner lengths for maximum power at 8,300 RPM,
right where the F20C made peak power. The runners were smooth, with no sharp bends or restrictions.
Every surface was designed for flow. The throttle body was drive-by-wire,
Honda's first. This allowed the ECU to control throttle opening independently of the driver's
foot. At high RPM, the ECU could limit throttle opening to prevent over-revving. It was a safety
feature disguised as technology. And the air filter. Most people don't think about air filters,
but at high RPM, restriction is the enemy. Honda used a large area filter with minimal restriction.
They tested dozens of filter materials before finding one that flowed enough while still
protecting the engine. The result was an intake system that could flow enough air to support 9,000
The exhaust that screamed? The exhaust was equally important. At 9,000 RPM, the engine
needed to expel 150 cubic feet of exhaust gas per minute. If the exhaust couldn't flow,
back pressure would build up, choking the engine and killing power.
Honda used a 4-2-1 header design, where the four cylinders merged into two pipes,
then into one. This design reduced back pressure while maintaining exhaust gas velocity for good
mid-range torque. The header primaries were equal length, tuned to optimize scavenging at high RPM.
Scavenging is where the exhaust pulse from one cylinder helps pull exhaust out of another
cylinder. It's a complex phenomenon that Honda spent months optimizing. The exhaust manifold
was made from stainless steel, then wrapped to retain heat. Hot exhaust gases flow faster than
cold ones. Faster flow means less back pressure. And the catalytic converters were positioned close
to the engine to heat up quickly, meeting emission standards while minimizing restriction.
The final touch was the muffler. It had to be quiet enough to meet sound regulations,
while flowing enough to support 9,000 RPM. Honda used a straight-through design with
minimal baffling. It was loud. Very loud. But it flowed. The ECU that made it all work.
All this hardware was useless without an ECU that could manage it.
And the F20C's ECU was a marvel of 1990s technology. It monitored dozens of sensors,
throttle position, airflow, coolant temperature, oil pressure, knock sensors, O2 sensors,
cam position, crank position. It processed this data hundreds of times per second and
adjusted fuel delivery, ignition timing, and VTEC engagement in real time. The knock sensors were
critical. At high RPM with high compression, detonation is always a risk. The ECU listened
for knock and pulled timing immediately if it detected any. This allowed the F20C to run on
91-octane pump gas safely. The VTEC engagement was programmed to occur at 6,000 RPM.
Below that, the engine used mild cam profiles for good drivability.
Above that, it switched to wild profiles for maximum power. The transition was instantaneous
and violent. You could feel it. You could hear it. And it was addictive. The rev limiter was
set at 9,000 RPM. Not 8,500. Not 8,800. Exactly 9,000. Honda had tested the engine to destruction
and determined that 9,000 was the safe limit. Go beyond that, and valve float became a risk.
The Price of Perfection
When the S2000 launched in 1999 with the F20C, it was an instant legend. 247 horsepower from
two liters. 9,000 RPM redline. The highest specific output of any naturally aspirated
production engine ever. But it came at a cost. The S2000 was expensive. $32,000 in 1999.
About $55,000 in today's money. That was Porsche Boxster money. And the F20C was expensive to
build. Really expensive. Honda lost money on every S. 2,000 they sold in the first few years.
The hand assembly. The exotic materials. The obsessive quality control. It all added up.
Worse, the F20C was peaky. It made 153 pound-feet of torque at 7,500 RPM. That's pathetic for a
Below 6,000 RPM, the engine felt flat. You had to rev it. You had to work for the power.
Journalists loved it. Engineers worshipped it. But average customers,
they bought Boxsters instead. The downgrade, nobody wanted.
In 2004, Honda made a decision that broke enthusiasts' hearts. They replaced the F20C
with the F22C1. The F22C1 was a destroked version with 2.2 liters of displacement.
It made the same 247 horsepower, but at 7,800 RPM instead of 8,300. And it made 162 pound-feet
of torque at 6,500 RPM. On paper, it was better. More torque. Easier to drive. But the redline
dropped to 8,000 RPM. The magic was gone. The F22C1 was a great engine. But it wasn't the F20C.
Why did Honda do it? Because American customers complained. They said the F20C had no torque.
They said it was too peaky. They wanted an engine that was easier to drive. So,
Honda gave them what they asked for. And in doing so, they killed the most advanced naturally
aspirated four-cylinder ever built. The legacy nobody can match. Today, the F20C is extinct.
Honda stopped building the S2000 in 2009. And they've never built another engine like it.
The current Civic Type R makes 315 horsepower from a turbocharged 2.0 liter. It's faster,
more powerful, and easier to drive than the S2000 ever was. But it doesn't rev to 9,000 RPM. It
doesn't make 123.5 horsepower per liter naturally aspirated. And it doesn't have that feeling,
that mechanical rage, that the F20C delivered. Because the F20C wasn't just an engine. It was
a statement. A middle finger to everyone who said naturally aspirated engines were dead.
A monument to Honda's engineering philosophy. And it was the apex. The absolute peak of what
was possible from a naturally aspirated four-cylinder. Honda's engineers knew it.
They've admitted in interviews that they could never build another F20C. Not because they don't
have the technology. But because it wouldn't make business sense. The F20C required hand assembly.
It required exotic materials. It required obsessive quality control. And it lost money.
Modern engines are designed for profitability first. Performance second. They're built by
robots on assembly lines. They use common parts to reduce costs. They're good engines.
But they're not the F20C. What Honda won't admit. Honda won't say it publicly. But the F20C was a
mistake. A glorious, beautiful mistake. They built an engine that was too advanced for its market.
Too peaky for most drivers. Too expensive to manufacture profitably. And too compromised
for everyday use. But to enthusiasts, the F20C represents something more important than profit
margins. It represents the moment when Honda said, let's see how far we can push naturally
aspirated technology. And they pushed it to the absolute limit. They broke physics. They rewrote
the rules. They built an engine that revved to 9,000 RPM and lasted 200,000 miles doing it.
Will we ever see another engine like the F20C? No. Because the economics don't work. Because
turbochargers are more efficient. Because nobody wants to rev an engine to 9,000 RPM to access
full power. The age of naturally aspirated screaming four cylinders is over. And the F20C
was its final, perfect expression. The question Honda can't answer. Here's what keeps me thinking.
If Honda could build a 9,000 RPM engine in 1999, what could they build today with modern materials
and technology? Could they build a 10,000 RPM engine? Could they extract 150 horsepower per
liter naturally aspirated? Or have they decided that chasing that dream isn't worth it? And here's
the bigger question. Was the F20C ahead of its time? Or the last of its kind? Drop your take
in the comments. Was the F20C the greatest four-cylinder ever built? Or an over-engineered
dinosaur? And if you owned an S2000, would you keep the F20C stock? Or would you be tempted to
see what it could do with a 9,500 RPM redline? This is the JDM Engine Archive. The F20C may be
gone, but its legacy screams on forever. Catch you next time.
