Rotor position sensing has long relied on resolvers because of their reliable performance in demanding automotive environments. Resolvers are still reliable, but the technology landscape has changed, and equally dependable alternatives can now do the job in a smaller footprint.
Carsten Himmele, Segment Marketing Manager, Automotive, Allegro MicroSystems
Electric vehicle (EV) designers are working to make the total drive system smaller and lighter, and a wound component at the motor is an obvious candidate for downsizing.
Part I of this Q&A series offered a system-level view of position sensing in EV traction inverters. In Part II, Carsten Himmele, a segment marketing manager in the automotive space at Allegro MicroSystems, takes a closer look at the challenges engineers encounter most often, explaining why inductive sensing is emerging as a compact alternative and what it takes to keep the technology accurate over the life of a vehicle.
Here’s what Himmele had to say…
Why are traction inverter environments becoming more difficult for rotor position sensing?
Modern traction inverters increasingly use silicon carbide devices, with gallium nitride expected to become more common in the future. Both technologies switch much faster than conventional insulated-gate bipolar transistors (IGBTs).
Higher switching frequencies broaden the electromagnetic interference spectrum, while the move to 800-V architectures increases dv/dt and di/dt at the switching node. Those faster transitions create stronger coupling into nearby signal lines, including the position sensor. Both the resolver and inductive sensing systems can be affected by this coupled noise, but a resolver can’t reject or compensate for it at the sensor level.
An IC with on-chip compensation and calibration can mitigate some error components early on, and its digital communication is less affected and needs fewer wires than the four required for a resolver’s dual differential analog output. Mechanical tolerances and vibration add to the error budget as well, and the sensor has to maintain accuracy over a vehicle service life of 15 years or more.
Rotor-position feedback enables the traction inverter to precisely control motor torque and speed in an EV powertrain.
Why is maintaining sub-degree accuracy so challenging?
Accuracy is only part of the requirement. The angle measurement also has to stay accurate as temperatures swing and mechanical tolerances shift over the life of the vehicle. Every stage of the sensing chain contributes some delay and error, from the sensing element and analog front end through the signal processing to the communication interface.
The sensing chain breaks down into three segments, the transducer, the signal processing, and the communication interface, and the signal ages as it moves through them. In a dynamic system, the measured accuracy is a function of time, so getting the position information to the microcontroller quickly matters, and so does knowing how much time has passed along the way.
Simpler analog resolvers are strong here, but an IC can solve that challenge by reporting how long the signal took to travel from one point to another.
How does functional safety influence position sensor design?
The traction motor torque function is typically designed to meet an Automotive Safety Integrity Level D (ASIL D) safety goal, so the position sensor has to contribute its share of that system-level requirement.
The required safety architecture depends on how tier-one system designers allocate safety across the position sensor, current sensors, and other feedback devices. That means sensor suppliers need to understand the system architecture before defining the appropriate safety mechanisms.
Why are engineers looking beyond traditional resolvers?
Resolvers remain widely used today because they hold up well in harsh conditions and deliver high accuracy, but they are relatively bulky and heavy and require additional signal-conditioning hardware. Inductive sensing offers a more compact approach using PCB coils and an integrated circuit (IC).
Because much of the signal processing and compensation occurs inside the IC, the overall sensing system can become smaller while reducing component count.
Why do output interfaces matter in position sensor selection?
Resolvers typically use dual differential sine and cosine interfaces. Those signals perform reliably in harsh electrical environments, but they require analog-to-digital conversion and exactly four pins. Pin count relates directly to cost, so fewer pins are favorable as long as signal integrity, speed, and quality are not compromised.
A digital interface, such as serial peripheral interface (SPI), is also based on four pins. However, because SPI is a bus system, multiple devices can share those lines and optimize the overall interface. A combination of absolute and incremental outputs can also reduce interface complexity without sacrificing absolute angle information or speed.
Resolvers use wound electromagnetic components to measure rotor position, but their size and supporting electronics can complicate integration into increasingly compact EV drive systems.
How does inductive sensing help reduce system size?
Space is always limited in traction inverter applications. Power density continues to increase while inverter electronics become more compact. Even after adding the PCB coils, an inductive sensing solution is typically much thinner and lighter than a resolver. The integrated approach can also reduce external components and simplify the overall signal chain.
How are electromagnetic compatibility challenges addressed in inductive sensing designs?
Electromagnetic compatibility (EMC) is one of the biggest challenges for inductive sensing because the transmit and receive coils can behave like antennas. That makes careful coil design and IC design essential.
The sensing system is evaluated using automotive EMC testing, including bulk current injection, where energy is intentionally coupled into supply and signal lines. Beyond the IC, the PCB layout, schematic design, and system-level review all play an important role in maintaining signal integrity.
What advantages does digital signal processing provide?
Modern inductive sensing ICs already process the signal digitally, which allows compensation for known error sources such as temperature.
Digital communication also simplifies integration with the controller while enabling additional diagnostics and compensation features that are difficult to implement with traditional resolver systems. Those capabilities help maintain angle accuracy throughout the vehicle’s operating life.
Filed Under: Featured Contributions, Inverter, Q&As, Sensors