Browse Prior Art Database

Self-Configuring, Adaptive DSP Controller for Multiple AC Induction and Sensorless BLDC Motors Disclosure Number: IPCOM000015715D
Original Publication Date: 2002-May-02
Included in the Prior Art Database: 2003-Jun-21

Publishing Venue



Our invention is a "Self-Configuring, Adaptive DSP Controller for Multiple AC Induction and Sensorless BLDC Motors." The system includes hardware and custom embedded software which automatically configures itself for the appropriate motor type based on the circuit it is designed into, thereby supporting multiple motor configurations including dual Sensorless Brushless DC motors, Dual AC Induction motors, or a combination of each. The problem solved is that multiple controllers are no longer needed to support both cases of single motors of varying types as well as multiple motors of the same or differing type. Traditionally, multiple versions of microcode were used exclusively to control a single motor of a single type. Our solution involves both microcode and hardware which tailors itself to the motor control application it is designed into. Our software system adapts to the hardware attached to it. A single stream of code is maintained for all possible motor combinations. Hardware is simply configured with the existence of a resistor network. The DSP detects the BEMF sense networks for both motor powertrains, and sets a configuration state variable according to the results. According to the motor configuration state, motor drive control registers are initialized , DSP hardware interrupts are configured accordingly, and appropriate motor type specific control algorithms are called from the mainloop of code. This invention is implemented in a prototype, as part of the refrigeration controller for future Z Series high-end servers. Working hardware and lab data are available. The implementation looks for a network of counter-EMF sense resistors on drive phase outputs by applying a known sequence of pulses to drive transistors. Control algorithms are selected based on feedback to a Digital Signal Processor. Please refer to Figure 1. Figure 1 System Implementation 1