Improved Startup and Slow-Speed Operation of Field Oriented Control (FOC) Algorithms

It’s well known that FOC is the king of motor control algorithms. It provides the highest energy efficiency and the maximum motor torque output when it is running in a closed loop. However, it’s Achille’s heel has always been that during motor startup and when running at very low speeds, the motor phase current feedback is not stable, and the motor runs open loop providing minimal torque and low efficiency.


For applications where the time spent at slow speeds in minimal, perhaps just on motor start-up as it ramps up to a high speed, this short open-loop control period is not a big concern. However, many applications need the maximum motor torque at startup and/or spend significant time at slower speeds. Image starting an e-bike from a parked position facing up-hill or using a battery powered drill to back out a lug nut sunk into a hard oak 4x4 board. Both situations require the maximum torque from the motor right at startup and both spend significant time running at slower speeds.


To solve these problems and help customers migrate applications from Hall switched 6-step block commutation to FOC, Microchip has developed an algorithm called Zero-Speed/Maximum-Torque (ZS/MT). This algorithm works in conjunction with FOC to provide the maximum torque from the motor at standstill (at start-up) and all through the slow-speed range until the normal Back Electromagnetic Force (BEMF) based sensorless estimator, typically a Phase-Locked Loop (PLL) can take over. In other words, the two work in a hybrid estimator arrangement.


In this session, details about how the ZS/MT algorithm and the hybrid estimator work will be presented and how it is deployed to customers. Also, the limitations and requirements for running ZS/MT will be discussed as well as enhancements in progress to expand the range of applications that can benefit from using ZS/MT.


Presenters

StanTrout_edited_edited.jpg

Patrick Heath

Motor Control Manager
Microchip Technology