Two bots line up. The faster motor loses. Why?

Most builders answer a slow lap by buying a bigger number: more RPM, more voltage, more torque. That works until the wheels break loose in the first corner. Now the fast motor is spinning tires while the lighter bot rolls past.

Speed isn't a spec you buy. It's a budget you balance between weight, torque, and grip. Here's what sets your top speed: the traction ceiling, why more RPM has a real cost, where weight helps and where it doesn't, and the one trick that breaks the trade-off for good.

For the buying-decision version of this, see Choosing a Chassis for Speed. This post is what's underneath it.

The Traction Ceiling: How Much Force Your Wheels Can Actually Put Down

Every wheel has a hard limit on how much forward push it can put into the track before it slips. That limit is set by two things only: the friction between your wheel and the surface, and how much weight is actually pressing that wheel down. Engineers write it as force equals friction coefficient times normal force, or F = μN. Normal force is just the technical name for the weight pushing straight down on the contact patch.

Say your bot weighs 150 grams, about what a Mark 1 or Mark 2 weighs stock. Pull it across the track with a fish scale until the wheels start to slide, and you'll get your actual friction coefficient, not a number off a spec sheet. A rough placeholder for rubber on a matte competition track sits somewhere around 0.5 to 0.7. At 0.6, your traction ceiling comes out to about 90 grams of push. That's the hard stop.

No motor, no gear ratio, and no amount of voltage gets you past that ceiling. Push more torque at the wheel than the ceiling allows and the extra doesn't turn into speed. It turns into wheelspin, heat, and a tire that's worn out by the third heat.

Why More RPM Isn't Free

A motor's gear ratio decides where its output sits on the speed-versus-torque line. Push the ratio up and torque climbs while top speed drops, roughly in proportion, since the gearbox itself is only about 70 to 85 percent efficient at converting one into the other. Drop the ratio and you get the reverse: a fast, low-torque motor that spins wheels for a living.

This is why Techgeeks' own line splits the way it does. The Mark 1 runs 600 RPM N20 motors on a standard grip chassis, geared for torque the wheels can actually use. The Mark 2 jumps to 2000 RPM motors, more than three times the speed, but that only works because the suction chassis pairs it with well over 600 grams of downforce. Put a 2000 RPM motor on a standard chassis with no extra downforce and you haven't built a faster bot. You've built one that hits its traction ceiling sooner and burns the rest of that RPM as wheelspin.

Pick your gear ratio for the corner exit, not the straightaway. The corner is where your traction ceiling gets tested first.

Where Weight Actually Costs You (and Where It Doesn't)

Weight cuts two different ways, depending on which limit you're actually running into.

If your motor is the bottleneck, meaning the wheels grip fine and never break loose, every gram you add costs you acceleration directly. That's plain F = ma: the same motor force has to move more mass, so it moves it slower. Most beginner builds live in this regime.

If your wheels are the bottleneck, meaning they already spin under hard acceleration, the math changes. Your traction ceiling is μN, and N is roughly your total weight. Add mass and you raise the ceiling at the same time you raise what needs pushing. In the simple version of the physics, those two effects cancel: maximum grip-limited acceleration works out to roughly μg, a number that doesn't depend on how heavy the bot is.

That's not the whole story. Heavier bots still take more force to reach that acceleration, drain more current doing it, and put more stress on the gearbox with every launch and stop. Weight isn't automatically the enemy. Misplaced weight is. Where you put the heaviest single part, almost always the battery, changes how your total weight actually splits across the wheels, without changing the total weight at all. A study on PID-tuned line followers found exactly this: shifting the battery toward the front increased slipping, shifting it too far back made the front end lift in sharp turns, and centering it relative to the frame gave the most stable result. That's a five-minute chassis tweak, not a redesign.

Downforce: The One Way to Cheat the Trade-Off

There's a way around the cancellation above, and it comes down to one idea: raise N without raising m.

Downforce does exactly that. Instead of adding weight to the chassis, you use moving air to pull the chassis down onto the wheels. Grip goes up. The mass you have to accelerate does not. That breaks the trade-off that plain weight can't.

This isn't a robotics-only idea. In 1978, the Brabham BT46B Formula 1 car ran a large rear-mounted fan, officially there for engine cooling, that also pulled air from under the car and generated a huge amount of extra downforce. It won its only race, at the Swedish Grand Prix, and was withdrawn from competition within weeks after rival teams objected to how much grip it bought. Robotics builders have used the same trick since: Pololu's Suckbot line follower ran a sealed-skirt fan for the same reason, less sliding through corners.

Techgeeks' Advanced Suction Chassis runs a small coreless motor and impeller under a sealed skirt to generate over 600 grams of downforce, on a robot that weighs about a quarter of that. The Spirit chassis pushes it further, past 700 grams, five to six times the robot's own weight. That's the entire reason the Mark 2's 2000 RPM motors put power into the track instead of just spinning the wheels. Weight raises your ceiling at a cost proportional to what you add. Downforce raises it at a fraction of that cost, since the impeller's own weight is small next to the grip it buys.

The Wheel Is Where All of This Pays Off, or Gets Wasted

Every gram of torque and every gram of downforce still has to go through the wheel to become forward motion. This is the last link, and it's where careless choices quietly cancel out good ones upstream.

Grip starts with the rubber. A smooth, hard tire wastes contact area. A soft, micro-textured surface uses more of it. Techgeeks' N20 Wheels use a grip layer modeled on cricket bat rubber for exactly this reason: more real contact with the track per gram of wheel.

Size is a second lever. A 20 by 20 mm wheel has low rolling resistance and suits a fast, low-torque, motor-limited build. A 30 by 30 mm wheel puts more rubber on the track and suits a high-torque build that's already fighting wheelspin. Picking the bigger wheel on a motor-limited bot just adds rolling resistance you don't need.

There's a third factor that's easy to miss: rotational inertia, how much a wheel resists changing its own spin speed. A lighter wheel hub spins up and slows down faster, which means a PID correction reaches the track sooner instead of getting absorbed by a wheel that's slow to change speed. Match the wheel to whichever bottleneck you found two sections ago, not to whichever one looks fastest on the shelf.

Conclusion

Speed on a line follower isn't one spec you max out. It's a budget spread across four things: how much torque your wheels can actually use before they slip, how your gear ratio trades speed for torque, where your weight sits and whether it's helping or costing you, and whether you've found a way to add grip without adding mass.

The bots that win aren't the ones with the biggest motor. They're the ones that spent every gram of traction on purpose.

So which one is actually holding your lap time back: the motor, the weight, or the grip you haven't measured yet?

If it's grip, the Impeller Setup adds the downforce trick above to almost any existing chassis, no full rebuild required.