A few key figures of merit illuminate aspects of motor performance to simplify the selection process.
There are time constraints within which work must take place. A given amount of torque is mandatory to accelerate and decelerate the load. The work done in accelerating the system depends on the inertial mass of the system.
These values all interact. Experienced motion-systems designers may creatively make trade-offs among them. Fundamentally, they are simple physical dimensions which, when properly interpreted, can define the best technology for specific needs.
It can be illuminating to compare motors of different technologies that have similar sizes and torques. Such a tactic can quickly help show, for example, whether servomotors or stepmotors provide a more promising means of reaching an end.
Clearly, nothing is more economical than open-loop stepping-motor systems for simple motion control. From a mechanical perspective, the original developers of the stepmotor assumed it would operate without feedback as long as the inertia of the load was less than 10 times that of the rotor. Under these conditions, and assuming that torque and friction remain relatively constant, one can operate a stepping motor just by turning two TTL switches off and on. Easy programming has always been one of the stepmotor’s most valued features.
Years of use have revealed numerous limitations in stepping technology. As the field has matured, efforts to broaden stepping motor capability have paid off. Encoder feedback helps improve load handling and makes it possible to microstep these motors. Microstepping brings higher accuracy and smooth performance, eliminating most cogging and harmonic instabilities. These improvements have a cost, but they let today’s stepmotors function as smoothly as servosystems.
Though dc brush motors once were a staple of precision motion applications, they are increasingly replaced by brushless systems. Proper precautions can reduce brush maintenance to a minimum, but there are many medical and electronic uses where particles generated by brushes are completely unacceptable.
Nevertheless, brush servos cost less than brushless equivalents because their technology is mature, and because their magnetic components can be relatively inexpensive. Brush servoamplifiers also use only single-phase power, making them less costly than the three-phase amps required for brushless schemes.
Brushless amplifiers have indeed come down in cost over the last few years, thanks to cheaper power semiconductors and more powerful microprocessors. Manufacturing costs have dropped as product volumes have risen. There have been several process innovations that have cut the price of magnet materials needed in high-performance brushless motors.
In addition, brushless designs have more thermal efficiency than brush motors because the rotor produces no heat. The result is a higher degree of torque produced with less material. Vendor data shows, for example, that a 2-in. diameter brushless servo produces 1.2 hp, an astounding performance benchmark compared to ac induction motors.
Data tells the story
It is interesting to compare the data from several vendors with regard to motor case sizes and torque. This comparison provides a relatively simple way of sizing up performance differences between steppers and servos.
Take the example of a NEMA-23 (2.3-in.-diameter) stepper with two-stack rotor and a NEMA-34 (3.4-in.-diameter) stepper with three-stack rotor, compared with 2 and 3-in.-diameter brushless servos respectively. A point to note is that only a few motors have a torque greater than 650 oz-in. (or about 1.5 hp). The reason is the high cost of the magnetic material and difficulty of processing the rotor. This cost rises rapidly because the volume of material increases exponentially with rotor radius (from πr2L, the volume of a cylinder), so few motors are available in high power ranges.
The first rule that emerges from scrutinizing motor data is that stepping motors are generally uneconomical for applications above about 1 hp, though this will vary somewhat depending on the supplier. Also of note is that stepping motors are rated for torque differently than brushless dc servos. The peak torque performance of the stepper is at relatively low speed and drops rapidly as speed rises.
In contrast, most brushless dc servos produce full torque at all speeds. Thus there is a significant difference in total power. In addition, the intermittent torque capability of a brushless dc system usually exceeds the continuous torque by a factor of two or three.
The difference in torque produced by steppers and brushless servos bears heavily on the suitability of these technologies for a given application. Numerous manufacturing situations see torque loads change as parts change in the production process. These conditions are tough to handle with a stepper because stepper torque peaks in a narrow speed range. If steppers are, in fact, used in such conditions, they must be oversized to a degree that increases the cost of the solution. This phenomenon leads to the general rule that steppers may not be best for circumstances where loads vary.
In contrast, a typical brushless servosystem easily handles varying torque loads or disturbances because of its high intermittent torque qualities. One caveat, though, is that each servomotor vendor uses its own torque rating method. Thus it is not a good idea to cross-reference servomotors from one vendor to another. Misguided cross referencing can result in dramatic oversizing of the motor and amplifier. The better approach is to use actual application data and evaluate the application using the vendor’s own sizing software.
The task of making effective comparisons between servo and stepper technology becomes somewhat involved. One approach is to use the standard horsepower equation to develop the equivalent horsepower at the peak torque and rated speed of candidate systems. The horsepower equation is a speed-dependent measure of power, so it is easy for motor designers to “cheat” a little by designing motors with high base speeds. This makes it appear that a 4,500-rpm brushless dc servo is somewhat better than a 3,000-rpm stepper.
Work done is an objective measure of the power capability available from a given system. It is a valid figure of merit for judging performance. Many applications can be optimized using a brushless servo having higher rated speeds, in combination with gear reduction.
Stepping motors can benefit from the same arrangement. But the highest torques come at low speed and top stepping-motor speeds are 3,000 rpm at best, less in larger motors. This means the ultimate improvement will be limited.
It is useful to compare rotor inertia and T/J, torque-to-inertia ratio. Torque divided by inertia is a unitless parameter of merit, but has great value as an indicator of a system’s ability to accelerate and decelerate a load.
In all systems, inertia mass of the motor and load opposes fast acceleration and deceleration. As inertia rises, the system has greater resistance to acceleration and deceleration. In fact, there are many applications where high inertia mass makes it more difficult to stop the load than to start it.
T/J is an excellent parameter of merit in high-speed systems where loads are typically small. Comparisons show brushless dc servos have superior capabilities in these types of applications, often exhibiting more than twice the acceleration capability of similar-sized steppers. These leads to the general rule that brushless dc servomotors are usually the best choice when conditions call for high speed or fast acceleration or deceleration.
Another interesting comparison is the motor weight and volume. Torque-to-volume and torque-to-weight can be used as unitless figures of merit here. These calculations reveal brushless servomotors have significantly higher torque generated per unit volume, making brushless technology a preferred choice in applications where compact size and light weight is a requirement.
Stepping motors weigh almost twice as much as their brushless servo counterparts for a given torque product. Weight generally is not a major design consideration, but becomes important in cases such as XY actuators and robots. Here, the weight of each motor becomes part of the load in the next supporting axis. The build-up of weight may degrade the performance of the overall mechanism. In robotics, for example, the ability to control the system becomes closely tied to the weight of the mechanics as the number of axes grows.
Comparing weight, torque, speed and inertia
For work product, or effective horsepower, the standard horsepower equation can be used to develop the equivalent horsepower at the peak torque and rated speed of each system. The work done is an objective measure of a system’s power capability, and thus serves as a valid figure of merit for judging performance.
Similarly, rotor inertia and torque-to-inertia ratio provides a way of indicating the ability to accelerate and decelerate a load. The data presented in the table would indicate that brushless servos have more than twice the acceleration capability of steppers. Finally, torque-to-weight and torque-to-volume figures of merit indicate that situations demand compact size and light weight would be best served by servos because these motors generate significantly more torque per unit of volume.
|Motor selection in a nutshell |
• Over 1 hp, use a servo — This is because stepping motors are generally uneconomical for powering applications above this level.
• For constant loads, steppers are better
• Varying loads demand a servo — Systems must be sized for worst-case loads, giving servos the edge here.
• High acceleration requires brushless technology — This rule may also hold for high speeds as well.