Here’s something most drivers never think about: in almost every car built since the mid-2000s, the gas pedal is just an electronic input — there is no cable between your right foot and the engine. You’re not pulling a throttle open. You’re moving a sensor, and the sensor sends a voltage. Your foot is a signal generator. I know this because on our student-built race car — a Formula Student EV we build at Chung-Ang University for the FSKorea EV class — I’ve measured that signal, argued with it, and once lost an afternoon to it.
The Pedal Is Just a Sensor
Our race car uses an aftermarket pedal, and when I probed it with a multimeter, the output did exactly what a textbook says a potentiometer should do: press slowly, and the voltage sweeps smoothly through a range — on ours, 0.85 volts released to 4.2 volts floored. The controller reads that voltage and translates it into a torque request. Released means “give me nothing,” floored means “give me everything,” and every position in between is a percentage.
Notice something odd there: released is not zero volts. The pedal at rest sits at a calibrated idle voltage, and the controller is programmed to treat everything below a certain threshold as “foot off.” That detail sounds boring. It’s the reason our car once refused to start.
The Day Our Car Pressed Its Own Pedal
At a test day, the motor controller threw error code 14 the moment we powered up: pedal pressed at startup. Nobody was touching the pedal. The car was convinced someone was.
Here’s what had happened, and the pedal was innocent. Buried in the controller’s settings is a parameter called throttle minimum valid voltage — the voltage at which the controller decides the driver is asking for torque. Ours was set to 0.8 V. Our pedal, untouched, rests at 0.85 V. The released pedal was permanently sitting above the line. From the moment we switched the car on, the controller saw a foot on the throttle, and it refused to arm.
So we measured the pedal end to end (0.85 V released, 4.2 V floored) and raised that parameter to a conservative 1.2 V — well clear of the resting voltage, but still early enough in the pedal travel that the car responds immediately. The error vanished, the contactor clicked, and the car drove off like nothing had ever been wrong.
One more setting worth stealing: we told the controller to treat 3.2 V as 100% output, even though the pedal actually reaches 4.2 V. That volt of headroom means a slightly worn or drifting pedal still commands full power. Production cars do the same thing, which is why a healthy throttle sensor never needs the last few millimeters of travel to give you everything.
This is also why some production cars want a throttle relearn procedure after you disconnect the battery: the computer needs to re-learn what “released” and “floored” mean before it can trust the pedal again. When a shop says your car “forgot its idle,” this is what they’re talking about.
One more lesson from that season. We originally wanted an organ-style pedal. On the bench, fed 5 V from a power supply, it behaved perfectly. Wired to our controller, its output collapsed to somewhere between 0.5 and 1.2 volts — a fraction of the range the controller needed — and the car crawled. We burned most of a test day blaming our wiring and our parameters before accepting it was a compatibility problem and going back to the pedal built by the controller’s own manufacturer. Then at a competition we met another university team running that same organ pedal successfully. They’d put an op-amp on it to reshape the signal into the range their controller wanted. A textbook part, applied to a real problem. We’d been treating classroom knowledge and race car problems as two separate worlds. They weren’t.
Kickdown: The Hidden Switch at the Bottom
If you drive an automatic, try this on an empty road: floor the pedal completely. On a lot of cars you’ll feel the transmission suddenly drop two gears and the engine leap toward redline. I first felt this in a rented Hyundai Avante — floored it, and the gearbox instantly kicked down two gears like I’d flipped a hidden switch. Later, driving a Genesis G70 2.5T at a manufacturer driving experience, same thing: full throttle, two-gear kickdown, and a very different car underneath me.
That’s the kickdown function. On older automatics it literally was a switch at the end of the pedal’s travel; on modern drive-by-wire cars it’s software watching for 100% pedal, sometimes with a mechanical detent added just so your foot can feel the “click.” Either way, the message is the same: the driver wants everything — shift accordingly.
Electric cars take this to its logical extreme. I got to launch a Kia EV6 GT at a track event, and the pedal-to-shove delay is essentially zero — no gears to drop, no revs to build. The voltage goes up, the torque shows up. It rearranged my expectations of what a pedal can do.
Drive Modes: Same Pedal, Different Personalities
Here’s the part I find genuinely clever: a drive mode changes software, not hardware. My daily driver is my dad’s old Hyundai Equus, and in Sport mode the throttle gets noticeably more eager. The pedal hardware hasn’t changed. The voltage range hasn’t changed. What changed is the map — the software curve that converts pedal percentage into torque request. Eco mode stretches the curve out so the first half of the pedal does very little. Sport mode front-loads it so the same half-press delivers much more. Automakers are shipping one pedal with several personalities, and the difference between them is a lookup table.
Once you know this, marketing claims get easier to read. A “sharper throttle response” mode isn’t necessarily making more power — it’s often just spending the pedal travel differently.
If You’re Wiring a Pedal on a Formula Student Car
If you’re on a student team about to wire an aftermarket pedal to a motor controller, measure the pedal’s real output range before you trust a single parameter in the software. Every setting downstream is built on those two numbers, and no datasheet knows what your pedal does on your car. Here’s the order that would have saved us an entire test day.
- Measure released and floored, with the pedal wired to the controller. Ours read 0.85 V at rest and 4.2 V flat to the floor. Write both numbers down before you open the parameter software.
- Set the minimum valid voltage clear of the resting voltage. Ours was 0.8 V against a pedal that rests at 0.85 V, so the controller saw a foot on the throttle from the instant we switched on — error 14, no arming, nobody in the seat. We moved it to 1.2 V and the car woke up.
- Leave headroom at the top. We told the controller that 3.2 V means 100%, even though the pedal reaches 4.2 V. That volt of margin means a worn or drifting pedal still commands full power on the last day of the competition, not just the first.
- Never assume a pedal is compatible because it worked on a bench. Our organ pedal was flawless on a bench supply and collapsed to somewhere between 0.5 and 1.2 volts once it was wired to the controller. We burned most of a test day blaming our own wiring and our own parameters before we admitted it was a compatibility problem.
The part of that story I’d want a first-year to hear: at the competition we met another university team running the exact organ pedal we had given up on. They’d put an op-amp in front of it to reshape the output into the range their controller wanted. A textbook part, applied to a real problem, by people our age. We had been treating classroom knowledge and race car problems as two separate worlds. They were never two worlds. We just never thought to look in the first one.
Next time you press the gas, picture what’s actually happening: a small voltage rising, a computer doing math, and a torque request going out. You’re not pulling a cable anymore. You’re writing a signal — so write a clean one.
This is the latest post in the series.
Start from the beginning: Student EV Race Car: 3 Dead Controllers and What a Brutal Season Taught Me
Series: This is part of Field Notes — everything that broke on our Formula Student EV car, in the order it broke.