MM32SPIN360PF

✓ In Stock

Cortex-M3 motor control MCU with 128KB Flash, hardware cordic, and advanced FOC support for PMSM motors

Product Overview

Description

The MM32SPIN360PF is an advanced motor control microcontroller featuring ARM Cortex-M3 core with hardware cordic accelerator and comprehensive motor control peripherals. It is designed for high-performance PMSM and BLDC motor control applications.

This MCU includes 128KB Flash memory and 16KB SRAM with advanced features including hardware cordic for fast trigonometric calculations, three advanced-control timers with complementary PWM, and fast 12-bit ADC with 2Msps sampling rate.

The integrated features support sensorless FOC algorithms with sliding mode observer or extended Kalman filter. The device operates from 4.5V to 16V with comprehensive protection features for safe motor operation.

Product Series

MM

Primary Application

High-performance servo drives

Key Features

  • ARM Cortex-M3 up to 96MHz
  • Hardware cordic accelerator
  • 128KB Flash with read protection
  • 16KB SRAM
  • Advanced PWM with dead-time
  • Fast 12-bit ADC with PWM sync
  • Analog comparators with break
  • Quadrature encoder interface

Specifications

Core ARM Cortex-M3
Frequency 96 MHz
Flash 128 KB
RAM 16 KB
Hardware Cordic Yes
PWM Resolution 16-bit
ADC 12-bit, 2Msps
Package LQFP-64

Applications

High-performance servo drives

Motor drive and control systems

Appliance inverter motors

Motor drive and control systems

Drone propulsion

Electronic system design

Industrial motor drives

Motor drive and control systems

Electric vehicle auxiliary motors

Motor drive and control systems

Documents & Resources

FAE Expert Insights

D

"The MM32SPIN360PF is MindMotion's flagship motor control MCU and it delivers exceptional performance for demanding applications. The hardware cordic is a game-changer for FOC implementations - it accelerates Park/Clarke transformations and SVPWM calculations by 10x compared to software implementation. I've successfully implemented sensorless FOC for PMSM motors using this MCU with excellent results. The 96MHz Cortex-M3 provides ample processing power for complex observer algorithms while maintaining reasonable cost. The PWM-ADC synchronization is precise, enabling accurate current measurement critical for torque control. I particularly like the quadrature encoder interface with hardware decoding reducing CPU overhead. For a high-speed spindle application, we achieved 0.1% speed accuracy using this MCU with sensorless FOC. The protection features including fast over-current detection provide peace of mind for industrial applications."

Advanced motor control MCU with hardware cordic for high-performance FOC applications

— Dr. James Chen, BeiLuo

Frequently Asked Questions

What is the hardware cordic and how does it help motor control?

The hardware cordic (COordinate Rotation DIgital Computer) is a hardware accelerator for trigonometric and vector calculations: (1) Functions supported - Sine, cosine, arctangent, vector magnitude, and vector rotation. (2) Performance - Completes calculations in approximately 20 clock cycles vs hundreds of cycles for software implementation. (3) FOC applications - Accelerates Park transformation (3 sine/cosine), inverse Park, and SVPWM calculations. (4) Precision - 16-bit precision suitable for motor control applications. (5) CPU offload - Frees CPU for other tasks like observer calculations and communication. In FOC algorithms, the cordic accelerates the most time-critical calculations. For example, a typical sensorless FOC loop requires 4-6 trigonometric calculations. With hardware cordic, these complete in ~120 cycles vs ~600 cycles in software, enabling higher control loop frequencies or lower CPU utilization. This is essential for high-performance servo applications requiring fast current loops.

Contact LiTong for cordic utilization in FOC algorithms and performance optimization.

hardware cordic trigonometric calculation FOC acceleration vector rotation
What sensorless FOC algorithms are supported?

MM32SPIN360PF supports various sensorless FOC algorithms through its processing power and peripherals: (1) Sliding Mode Observer (SMO) - Robust observer estimating rotor position from current measurements. Good dynamic performance and parameter robustness. (2) Extended Kalman Filter (EKF) - Optimal state estimator providing accurate position and speed estimates. Higher computational load but excellent performance. (3) Flux Linkage Observer - Estimates rotor flux position directly. Simple implementation with good low-speed performance. (4) High-Frequency Injection (HFI) - Injects high-frequency signal for position detection at standstill and low speeds. Useful for startup and low-speed operation. MindMotion provides reference implementations of SMO and flux linkage observers in their motor control library. The 96MHz Cortex-M3 with hardware cordic can execute these algorithms with control loop frequencies of 10-20kHz. For startup from standstill, high-frequency injection or open-loop startup sequences are typically used.

Contact LiTong for sensorless FOC algorithm selection and implementation support.

sensorless FOC sliding mode observer EKF flux observer
How accurate is the quadrature encoder interface?

The MM32SPIN360PF includes a hardware quadrature encoder interface (QEI) with the following capabilities: (1) Resolution - Can decode quadrature signals up to 1MHz edge rate per channel, supporting high-resolution encoders at high speeds. (2) 4x decoding - Counts on both edges of both channels providing 4 counts per encoder line. (3) 32-bit position counter - Wide counter range eliminating overflow concerns. (4) Index pulse handling - Optional index pulse input for absolute position reference. (5) Velocity calculation - Hardware velocity capture or software delta-position calculation. (6) Direction detection - Automatic direction detection from quadrature phase relationship. For a 1000-line encoder at 4x decoding, this provides 4000 counts per revolution. At 6000 RPM (100 rev/s), the edge rate is 400kHz, well within the 1MHz capability. The hardware QEI significantly reduces CPU overhead compared to software decoding, freeing processor cycles for control algorithms.

Contact LiTong for encoder interface configuration and high-speed encoder applications.

quadrature encoder QEI encoder interface position feedback
What is the control loop frequency achievable?

The MM32SPIN360PF can achieve high control loop frequencies suitable for high-performance motor control: (1) Current loop - Typically 10-20kHz for FOC applications. At 96MHz with hardware cordic, current loop execution time is approximately 5-10μs leaving significant headroom. (2) Speed loop - Typically 1-5kHz, easily achievable with the available processing power. (3) Position loop - Typically 100Hz-1kHz for servo applications. (4) PWM frequency - Usually matched to current loop at 10-20kHz. The achievable loop frequency depends on algorithm complexity: (1) Simple BLDC six-step - Current loop possible at 50kHz+. (2) Sensored FOC - Current loop at 20kHz typical. (3) Sensorless FOC with SMO - Current loop at 10-15kHz typical. (4) Sensorless FOC with EKF - Current loop at 5-10kHz due to higher computational load. Higher loop frequencies provide better dynamic response and lower current ripple but require more CPU time.

Contact LiTong for control loop frequency optimization and algorithm tuning.

control loop frequency FOC loop current loop control bandwidth
Does MM32SPIN360PF support single-shunt current sensing?

Yes, MM32SPIN360PF supports single-shunt current sensing for cost-effective motor drives: (1) Single-shunt principle - Reconstructs three-phase currents from DC bus current measurement using PWM state information. (2) ADC timing - Fast 2Msps ADC with precise PWM synchronization captures current during active vectors. (3) Sampling windows - Hardware identifies valid sampling windows when current flows through shunt. (4) Current reconstruction - Software algorithms reconstruct phase currents from samples. (5) Limitations - Some PWM states don't provide valid current information, requiring special handling at high modulation indices. Single-shunt sensing reduces hardware cost by eliminating two current sensors but requires more complex software. It works well for most applications but may have limitations at very high speeds or high modulation indices. For highest performance and simplest software, dual-shunt or triple-shunt sensing is preferred. The MindMotion motor control library includes single-shunt current reconstruction algorithms.

Contact LiTong for current sensing scheme selection and implementation guidance.

single-shunt current sensing current reconstruction shunt resistor