In the previous powertrain instalment, we moved the traction inverter from silicon IGBTs to 800V Silicon Carbide MOSFETs. That decision carries through to the motor.

SiC switches at a higher frequency with lower loss. The motor sees a phase current much closer to a pure sinusoid, with lower current ripple. Less ripple means lower harmonic copper loss in the windings and less eddy-current and hysteresis loss in the laminations and magnets. The motor runs cooler. A cooler motor keeps its magnets further from their demagnetisation limit and its copper at lower resistance. It is more efficient at every operating point.

This final Built in Britain instalment designs the machine to take full advantage of that: a 150 kW electric drive unit built for cost, efficiency and supply-chain sovereignty.

Drive unit specification at a glance

  1. Topology: Interior Permanent Magnet Synchronous Motor (IPM-PMSM).
  2. Peak power: 150 kW.
  3. Peak efficiency: 97%.
  4. Stator: compressed copper hairpin windings, with more than 70% slot fill.
  5. Magnets: sintered NdFeB, with grain boundary diffusion (GBD) of dysprosium and terbium.
  6. System voltage: 800V, fed by a SiC inverter integrated into the EDU.
  7. Insulation and bearings: PD-resistant enamel on the hairpins, and ceramic-hybrid bearings.
  8. Battery: 55 kWh sodium-ion, 152 Wh/kg at pack level.

Motor topology: why IPM-PMSM beats induction

An induction motor needs no magnets, which makes it attractive on paper for sovereignty. For this platform, however, the physics does not support it.

Where induction falls short

  1. Rotor copper loss. An induction machine generates rotor current through slip, and that current heats the rotor. A permanent magnet rotor needs no current to produce its field.
  2. Magnetising current. The stator must supply the whole magnetic field. Power factor falls, and so does part-load efficiency.
  3. Power density. For the same torque, an induction motor needs more active material. It ends up larger, heavier and harder to package.
  4. Rotor cooling. Heat trapped in a spinning rotor is the hardest heat to remove.

Where IPM-PMSM wins

  1. Two sources of torque. Buried magnets create saliency, so the motor produces magnet torque plus reluctance torque. That extra torque comes without any extra magnet material.
  2. Field weakening. The inverter injects negative d-axis current to extend the constant-power speed range without oversizing the magnets.
  3. Part-load efficiency. Everyday driving at 10–40 kW on urban roads and motorways falls in the IPM's high-efficiency region. That is where range is won.
  4. Mechanical retention. Magnets sit in stamped rotor pockets, not on the surface, so they need no retention sleeve at high rotor speeds.

This matters for the pack. As the battery instalment established, a 55 kWh sodium-ion pack has little energy to spare. Energy lost as rotor heat is range the car does not have. IPM-PMSM is the range-optimal topology for a car of this size and cost.

Trade-offs to manage

  1. Spin loss. The magnets induce back-EMF and iron loss while coasting. Low-loss thin-gauge electrical steel and a flux-optimised rotor keep drag torque low.
  2. Uncontrolled generation. At high speed, the back-EMF can exceed the 800V DC link. A fault could then feed current back into the pack. The SiC inverter's active short-circuit (ASC) safe state handles this, and the rotor design keeps the resulting short-circuit current within the thermal limits of the windings and magnets.

The stator: hairpin windings

Conventional stators use round-wire, pull-through windings: many thin strands pulled into each slot. Round conductors cannot pack tightly. The gaps are filled with resin or left as air, and both conduct heat poorly.

A hairpin stator replaces them with rectangular copper bars. Each bar is formed into a U-shape, inserted axially, twisted and laser-welded at the connection end.

The fill-factor jump

  1. Round wire: about 45% slot fill.
  2. Compressed hairpin: more than 70% slot fill.

That change affects the whole machine:

  1. Lower resistance. More copper in the same slot means lower DC resistance and lower I²R loss at a given torque.
  2. Lower thermal resistance. Solid copper sits flat against the slot liner and the stator teeth. Heat passes into the stator core and cooling jacket through a short, conductive path instead of through resin and air.
  3. Better heat dissipation. Continuous power rises towards peak power. The motor can sustain high output without derating.
  4. Short, rigid end turns. Hairpin end turns are compact and precise, which cuts axial length and the copper that produces no torque.
  5. A smaller, lighter motor. With higher current density and better cooling, the same 150 kW needs less stack length and less active mass.

The high-frequency caveat

Large rectangular conductors suffer from skin and proximity effects at high electrical frequency, which increases AC resistance at top speed. The fix is geometric: use more, thinner conductor layers per slot, typically six to eight, and transpose them across the slot. SiC's cleaner current waveform also helps by reducing the high-frequency harmonic content that drives AC loss.

Magnet strategy: sovereignty by design

The main supply-chain risk in any PMSM is the magnet. Sintered NdFeB provides the energy product that 150 kW in a compact package requires. The risk lies in the heavy rare earths.

The heavy rare earth bottleneck

  1. Dysprosium (Dy) and terbium (Tb) increase coercivity, the magnet's resistance to demagnetisation at high temperature.
  2. Traditional grades add them by bulk alloying throughout the magnet volume, including the core, where they do little.
  3. Both elements come from highly concentrated sources and are subject to export controls and sharp price swings. For a cost-constrained car, that is an unhedgeable risk.

Grain boundary diffusion

Grain boundary diffusion (GBD) changes where the heavy rare earths go.

  1. Dy or Tb is applied to the magnet surface. Heat treatment then drives it inwards along the grain boundaries.
  2. It forms a thin, high-anisotropy shell around each Nd₂Fe₁₄B grain, concentrated towards the magnet edges and surfaces. That is exactly where reverse fields hit hardest and where demagnetisation begins.
  3. The grain cores keep their high remanence. Coercivity rises, and flux is barely reduced.
  4. Heavy rare earth content falls to a fraction of what bulk-alloyed grades need for the same temperature rating.

What this protects

  1. The bill of materials. Less Dy and Tb per motor means less exposure to supply shocks and export controls.
  2. High-temperature capability. Coercivity remains high at peak operating temperatures, protecting against irreversible demagnetisation during sustained high load and short-circuit events.
  3. Rotor design freedom. The IPM geometry uses flux barriers and pocket placement to shield the magnets from armature reaction. That lets GBD-grade magnets run with margin.
  4. A route to UK sourcing. NdFeB magnets remain the drive unit's main imported input. The UK's emerging rare-earth alloy and magnet-recycling capability, including hydrogen-based recovery of magnets from end-of-life scrap in the West Midlands, offers a route to more domestic supply over time.

800V dielectric stress: protecting the machine from its inverter

The inverter instalment flagged the cost of SiC's speed: high dV/dt switching edges. Voltage transitions that take tens of nanoseconds put severe dielectric stress on the motor.

The failure mechanisms

  1. Voltage overshoot. Fast edges reflect at the impedance mismatch between cable and motor. Terminal voltage can overshoot the DC link substantially, and at 800V the peaks exceed the inception level of ordinary winding insulation.
  2. Uneven voltage distribution. A steep edge concentrates voltage across the first turns of each phase, where turn-to-turn stress is highest.
  3. Partial discharge (PD). Once the local field exceeds the partial-discharge inception voltage, micro-discharges erode the enamel. The insulation fails progressively.
  4. Bearing currents. Common-mode voltage couples capacitively onto the rotor, producing shaft voltage. It discharges through the bearing's lubricant film as electrical discharge machining (EDM) currents. The result is pitting, frosting and fluting of the races, and premature bearing failure.

The countermeasures

  1. High-performance hairpin insulation. A thick, partial-discharge-resistant enamel system, with an extruded high-temperature polymer layer over the enamel where the voltage stress requires it. Hairpins are well suited to this, because the rectangular bar takes a uniform, thicker insulation build than fine round wire.
  2. Phase and slot insulation. Upgraded slot liners and phase separators rated for 800V PD-free operation, plus vacuum-pressure impregnation to remove voids.
  3. Ceramic-hybrid bearings. Silicon nitride rolling elements are electrical insulators, which break the discharge path through the bearing and eliminate EDM currents at the source. The platform can use full-ceramic or hybrid bearings, backed by a shaft-grounding element where the application needs one.
  4. Integration. The SiC inverter mounts directly on the motor housing inside the EDU. Minimal phase-conductor length means minimal reflection and overshoot, so the fix starts with the layout.

Local manufacturing fit: the Midlands drive unit

The drive unit is assembled in the UK Midlands automotive engineering corridor, the same cluster that assembles the SiC power electronics. Each stage adds value in the UK:

  1. Precision lamination stamping. High-speed progressive-die stamping of thin-gauge electrical steel for the stator and rotor cores, with bonded or interlocked stacks to keep iron loss low.
  2. Hairpin forming. Automated conductor cutting, insulation stripping, U-bending, insertion, twisting and laser welding, followed by impregnation and PD testing.
  3. Rotor assembly. Magnet insertion, bonding, balancing and magnetisation.
  4. EDU integration. Motor, SiC inverter and single-speed reduction gearbox combined into one unit with a shared cooling circuit, then given full end-of-line electrical and NVH testing.

Rules of Origin

From 2027, the UK–EU Trade and Cooperation Agreement caps non-originating materials in an EV at 45% of ex-works price. The drive unit is designed with that limit in mind:

  1. Stamping, winding, rotor build, inverter assembly and EDU integration are all UK-originating value added.
  2. Imported inputs are limited to magnets, bare SiC dies and commodity electrical steel, all small in value compared with the finished unit.
  3. Together with the UK-converted sodium-ion pack, the whole powertrain keeps the vehicle comfortably inside the 45% cap. That means tariff-free export to the EU.

Built in Britain: the complete powertrain

This instalment completes the powertrain. Across the series, each decision reinforces the others:

  1. Battery: 55 kWh sodium-ion, 76% Cell-to-Pack efficiency, 361.8 kg, £2,258, which is 9.0% of the build.
  2. Inverter: 800V SiC, up to 70% less inverter loss, +4–6% system efficiency.
  3. Motor: 150 kW IPM-PMSM, hairpin stator with more than 70% fill, GBD magnets, 97% peak efficiency.
  4. Manufacturing: Coventry cells and Midlands power electronics and drive units, compliant with post-2027 TCA rules.
  5. Budget: within the £25,000 net manufacturing budget for a sub-£30k retail car.

The low-cost chemistry depends on an efficient inverter, and the inverter's efficiency depends on a motor designed to use it. Localised production is what lets all three reach the EU tariff-free.

This concludes the core Built in Britain powertrain triad. Explore our complete structural layout engineering models in the FutureWatt Platform Blueprint Index.