Step-motor-control algorithms for efficiency

Step-motor-control algorithms for efficiency

For maximum efficiency, the boundary conditions of the complete electro-mechanical system should be mapped

BY MATTHEW TYLER
ON Semiconductor
Phoenix, AZ
www.onsemi.com
and
MATS SANDVIK
Stegia
Västerås, Sweden
www.stegia.se

During the optimization of a stepper-motor-based motion control system design an engineer must look at cost, performance, efficiency, unanticipated feedback complications (such as mechanical resonance), and development time. Modern motor control systems are challenged to work in a number of adverse environments and the classical solution’s total efficiency is usually limited by the worst-case conditions encountered by the overall system. Adaptive control algorithms are essential to extract the full efficiency of the optimized electromechanical system.

System mapping

When maximum efficiency is desired the boundary conditions of the complete electro-mechanical system must be mapped. All system variables must be considered: temperature, mechanical degradation, acceleration, velocity, supply voltage, to list a few. System architecture will also have an impact.

In an open loop system it will typically be necessary to stimulate the motor with the worst case current drive and velocity profile, so it is safe to assume that efficiency is not the first design goal for such systems. This type of testing is very time consuming since the system has to be verified for all possible supply voltages, temperatures and velocities at which it will operate in order to minimize the risk of resonance. The potential for resonance exists in every stepper motor system and is often caused by operating at (or close to) the motor’s natural frequency. Avoiding these regions is critical as resonance may cause the motor to lose steps or go into a stall condition. However, identifying these regions can be quite difficult for an open loop system.

Closed-loop controls generally assume two forms: sensor-based systems (optical or hall effect) and sensorless. Sensorless, also known as “semiclosed loop,” most often use the voltage generated by the motor windings for feedback. Sensor-based control systems are widely used, but the additional variation of the sensor must be considered during the mapping exercise. Sensorless systems have a primary advantage in that the information relative to the physical motion of the motor is all that needs to be read. Another important advantage is the reduced system cost of a closed or semiclosed system, as well as reduced system complexity due to the lack of external sensors. Successful design requires an understanding of the nature of back EMF.

SLA mapping

Back EMF can be used to easily extract detailed information about the motion of the electro-mechanical system and to provide diagnostic data. Between motor drive current pulses, the motion of the motor windings produce a voltage as they move through the motor magnetic field. This information is commonly referred to as the speed and/or load angle (SLA) of the motor. The amplitude of the Back EMF can be monitored to provide a good approximation of the stepper motor angular velocity.

Figure 1 illustrates a mapping of the SLA pin of an AMIS-30522 micro-stepping stepper motor controller driving a conventional stepper motor mounted in a mechanical system. This information was collected during a frequency sweep of the NXT input (clock input that determines the speed of the motor stimulus). The frequency of stimulus increases as it moves left to right and different regions of operation can be clearly seen. The ability to measure the motor characteristics of a complete system is a very powerful feature available to the AMIS-305xx series – in particular it addresses the conventional design challenge where a system designer only analyzes the resonance behavior of the motor, unaware that the regions might change once the complete mechanical setup is put together.

The system controlling the motor can continuously sample the SLA voltage and then apply appropriate actions if unusual behavior is encountered. Since the back EMF is proportional to the rotational velocity of the rotor it can easily be used to sense the external load on the output shaft and regulate the current supplied to the motor. Another area where the data from the SLA pin is of great help is when the motor is about to enter a resonance region. By designing an algorithm which quickly identifies this behavior, the system controlling the stepper motor can immediately accelerate through this region to a new, safe velocity.

Fig. 1. A frequency sweep of the NXT pin while monitoring the SLA pin.

The red square on the left of Figure 1 highlights a resonance in the system. This can be the result of the physical mounting of the motor, the fundamental resonant frequency of the motor between steps, or other 2nd order factors. These are usually regions of commutation speed to be avoided and can be easily mapped in minutes with ON Semiconductor back EMF technology. This will help to reduce stress on the electro-mechanical system. This is important as system stress may result in increased noise, reduced performance and, potentially, reduced system reliability. The highlight of this data collection method is that it is not necessary to physically change the system to complete the mapping process. The only sensor is the motor itself, so there is no additional mechanical complexity.

The red square on the right in Figure 1 shows the region where the current drive is exceeding the RLC time constant of the system leading to residual current in the motor windings. This is the ‘speed limit’ of this specific electromechanical system.

Between these two areas is the advised operating region for the motor. It should also be noted that the same mapping can be used to identify a stalled condition in which the motor does not commutate and, therefore, no back EMF is generated. This condition can easily be managed within the system controller by configuring a minimum threshold during motor stimulus.

Using mapping data for design

Once the mapping has been completed and the desired velocity profile is known then the optimum SLA value can be selected. This will represent the most efficient operating point for a given system. Motor control variables such as current drive, acceleration, and speed can be dynamically adjusted to avoid issues that impair efficiency such as mechanical resonance and excessive drive current. The advantage of the sensorless/back EMF approach is that the feedback from the sensor is not simply binary but can be used to obtain detailed diagnostic information from the motor without additional system complexity. This, in turn, allows subtle changes in SLA to be used for real-time compensation before step loss is encountered. ■