EV Motors Aren't Done: Inside the Next Wave of EV Drive Units

ELI5/TLDR

Everyone obsesses over EV batteries, but the motor is where the quiet progress is happening. A teardown engineer walks through how an EV motor actually works, why nearly every carmaker has independently arrived at the exact same design, and why that sameness is a flashing sign that disruption is coming. The punchline: there are two challengers waiting in the wings — a magnet-free motor (insurance against China cutting off rare-earth supply) and a pancake-thin “axial flux” motor that does more with less steel. Also, cheaper sodium batteries are arriving and they work in the cold.

The Full Story

The speaker is Paul Turnbull, a technical specialist at Munro & Associates — the firm famous for buying cars, taking them completely apart, and reporting on what every manufacturer is actually doing under the skin. His beat is the motor and drive unit, the part that turns electricity into motion. But he opens with the battery, because that’s the elephant in the room.

The battery, briefly

Three things have always been true of EV batteries: too big, too heavy, too expensive. The good news is that the price has been falling. Carmakers now pay around $100 per kilowatt-hour at the pack level (that’s their cost, not what you’d pay buying cells off Amazon). The interesting twist is a new chemistry now in mass production.

Sodium ion battery chemistry works at sub-zero temperatures. Doesn’t suffer the same issues that lithium ion batteries have at very low temperatures.

These sodium-ion packs cost carmakers roughly $60 per kilowatt-hour — a big drop — and they behave better in winter. Think of it like swapping a fussy ingredient for a cheaper one that also happens to survive the freezer. Expect them in budget EVs soon.

Two other battery trends, quickly. First, voltage. Cars run their electrical system at either 400 or 800 volts. Higher voltage charges faster and runs a touch more efficiently, but most public chargers are still 400-volt, so the benefit only shows up at the rare 800-volt charger — and getting to 800 volts costs extra. So premium cars go 800, budget cars stay at 400, and Turnbull expects the split to settle around half and half. Second, how packs are built. The old way nests cells into modules, then modules into a pack — more parts, but each module stays under 50 volts (safer to handle) and the pack stays serviceable. The new way, pioneered by Tesla, glues cells straight into the pack with a urethane foam:

It makes the entire pack essentially disposable. However, it makes it very robust… if it lasts for a million miles, then you don’t necessarily need to service it.

Why efficiency is the whole game

Here’s the mental model that makes the rest click. A more efficient motor lets you go the same distance on a smaller battery. The battery is the expensive, heavy part — so shaving a dollar off efficiency in the motor can save ten or twenty dollars in battery. And it compounds:

If you can reduce the battery size, you reduce the battery weight which reduces the power required to accelerate the vehicle which helps you reduce the motor size and so on.

A smaller battery means less weight means less power needed means a smaller motor — a virtuous loop. So every design decision in the motor is really a decision about cost.

How the motor works

The drive unit is three things now fused into one aluminum casting: motor, inverter, gearbox. Fusing them removes parts and removes cables — and cables in a high-power EV are not innocent. A motor pulling hundreds of kilowatts is, electrically, a little broadcast station spraying interference. Keeping everything inside one conductive aluminum box keeps that noise contained.

The motor itself has two parts. The rotor spins and carries permanent magnets. It sits inside the stator, which stays still (hence “stationary” → stator). You pour alternating current into the stator and it becomes an electromagnet whose magnetic field rotates — even though the metal doesn’t move. The rotor’s permanent magnets chase that spinning field around, and that chase is the motor turning. This trick — a magnetic field that rotates inside a stationary part — is the foundational invention, credited to Nikola Tesla.

The inverter sitting on top converts the battery’s DC into the three-phase AC the stator needs. The gearbox has just one gear — you’re effectively in first gear from zero to top speed — because these motors stay efficient across a huge range of speeds. It also holds the differential, the part that lets the two driven wheels turn at different rates through a corner.

Why everyone’s motor now looks identical

This is the heart of the talk. Three things have converged across nearly every manufacturer.

The winding. For a century, stators were wound with bundles of fine round strands. The shift now is to square wire (“hairpin” winding). Square wire packs tighter — more copper, less wasted air gap. That matters because the energy lost as heat scales with current squared times resistance, so lower resistance means less heat, less waste, more range. The machines to do hairpin winding used to be exotic; carmakers like GM have been deliberately handing the technology to suppliers to create competition and drive the price down for everyone.

The rotor. After twenty years of varied designs, the industry has landed on one shape — a “double V,” where magnets sit near the rotor’s outer edge in two V-shaped slots. The appeal is tunability: during design (not in real time) engineers can adjust the geometry to hit almost any torque-and-speed target while keeping efficiency high at both low and high speed.

So the consensus design has a name: a radial-flux interior-magnet synchronous motor, double-V magnets, bar-wound stator. And Turnbull’s tell:

As soon as you see a technology where everyone’s copying everyone else and everyone has pretty much the same design, that is a technology… ripe for disruption.

What’s coming next

Two challengers.

The first is an old idea returning: the wound-field synchronous motor. Instead of permanent magnets, you put electromagnets in the rotor — which means no neodymium, the rare-earth magnet material sourced mostly from China. The hard part is feeding electricity into a spinning rotor; the cheap way is brushes, but BMW uses a clever brushless “rotating transformer.” Nissan and BMW are already in production with these — not because they’re better, but as insurance against a political disruption to the magnet supply chain.

The second, and Turnbull’s pick for the real disruptor, is the axial flux motor. The everyday motor sends its magnetic field outward, radially (rotor inside stator, like a pencil in a tube). An axial flux motor sends the field along the axle instead, sandwich-style, which makes it very thin — a pancake rather than a can. British firm Yasa (now owned by Mercedes-Benz) builds one, the YM360, slotted into hybrids where the torque converter used to sit because it’s so slim. The economics are the draw: it uses about the same copper and neodymium but dramatically less steel, and steel is a big chunk of a motor’s cost and weight at volume. Lighter, smaller, cheaper.

His timeline: permanent-magnet motors dominate through 2030 because the factories are already tooled and humming. Axial flux arrives first in high-performance cars where size and weight command a premium, then trickles down to affordable cars.

Key Takeaways

Motor efficiency is a cost lever for the battery: saving a dollar in the motor can save $10–20 in battery, because you can then carry a smaller pack — and the savings compound (lighter car → less power needed → smaller motor).
Heat loss in a motor scales as current² × resistance. Packing more copper into the stator slots (square “hairpin” wire vs. round stranded wire) lowers resistance, cuts loss, extends range.
The modern drive unit fuses motor + inverter + gearbox into one aluminum casting — fewer parts, shorter cable runs, and the conductive box contains the electromagnetic interference a high-power motor radiates.
Stator = stationary part you energize; rotor = spinning part with permanent magnets. The stator’s magnetic field rotates without the metal moving, and the rotor chases it. That rotating field is Tesla’s foundational invention.
EV gearboxes typically have a single gear because the motor stays efficient across the full speed range from standstill to top speed.
Sodium-ion batteries (CATL, ramping in 2025) cost carmakers ~$60/kWh vs. ~$100/kWh for current packs, and they keep working at sub-zero temperatures where lithium-ion struggles.
400V vs. 800V architecture: 800V charges faster and is slightly more efficient but only pays off at scarce 800V chargers and costs more — so premium cars go 800V, budget cars stay 400V.
Tesla/BYD glue cells directly into the pack (cell-to-pack) for robustness at the cost of serviceability — the pack is essentially disposable but built to outlast the car.
Industry-wide convergence on one motor design (radial-flux, interior permanent magnet, double-V rotor, bar-wound stator) is itself the signal that disruption is near.
Wound-field motors (Nissan, BMW) drop permanent magnets entirely — insurance against a China-driven neodymium supply shock, not a performance upgrade.
Axial flux motors (Yasa / Mercedes) use far less steel for similar copper and magnet content — lighter, smaller, cheaper. Turnbull’s bet for the dominant disruptor post-2030.

Claude’s Take

This is a teardown engineer talking shop, and it shows in the best way — concrete, unhyped, numbers attached. The framing that motor efficiency is really a battery cost decision is the single most useful idea here, and it reframes why carmakers sweat over copper packing density that sounds trivial. The “everyone has the same design, therefore disruption is coming” heuristic is a clean piece of pattern-recognition that travels well beyond motors.

Where to keep a little salt handy: the price figures ($60 and $100 per kWh, the 40-to-50% voltage split) are presented as settled facts without sources, and battery price quotes are notoriously slippery depending on chemistry, volume, and who’s counting. The sodium-ion optimism is real but the “near future” has been doing a lot of work in EV forecasts for a decade. None of it is wrong, but treat the specifics as a knowledgeable practitioner’s read rather than audited data.

Scored a 7: genuinely clarifying on how the hardware works and where it’s heading, delivered by someone who physically takes these things apart. It loses points only for being a single-source talk with unsourced numbers and a slightly promotional close (he plugs his own axial-flux video). Worth the half hour if the question “how does an EV actually move” has ever nagged at you.