Integrated motors and drives: Benefits and solutions
Combining a motor with its controlling electronics in one package does more than just eliminate the cost and bulk of the extra wiring-It improves efficiency, reliability, and electromagnetic compatibility. Also, torque curves for the integrated motor system eliminate guesswork. With the indexer, driver, motor, and feedback already combined, the rest of the system integration is simplified: There are fewer interfaces to be completed by the user, and the high power, radio frequency knowledge-intensive interfaces are eliminated.
Optimizing motors and drivers
Motor to driver optimization is easier in an integrated environment. The specifics of a motor may be easily embedded into the controller. This includes feedback type and resolution, motor characteristics, as well as optimal points for field weakening of indirect permanent magnet (PM) motors. Accurate motor modeling parameters used for sophisticated control methods are easily pre-loaded into the controller. The driver capability can be optimized for the motors being driven, saving the cost of over-design, and for some of the losses associated with operating a higher power driver than what is needed for the application.
Motor and driver efficiency improves because the driver is better matched with the motor, and there is elimination of additional wiring between the motor and the driver. The long wiring runs add to wasted power directly and indirectly. The direct cause is due to the runs’ I^2R losses [losses from the flow of current through a resistor]. Due to the high frequency content from the pulse-width modulation (PWM) drive, the effective resistance is not just the normal direct current (dc) resistance of the wire. The high-frequency components of the current waveforms cause the current to flow mostly near the skin of the wires, making their effective resistance even higher-known as the "skin effect." Nonintegrated systems also lose efficiency, due to design considerations, to safely operate the switching elements—field-effect transistors (FETs) or insulated-gate bipolar transistors (IGBTs)—in the presence of strong, reflected voltage waveforms.
When the transmission time from the switches to the load is long compared to the time of the driver switches, it is common for reflections to occur between the driver and the motor. The voltage waveforms will be reflected by the change in impedance between the transmission line—the motor power cable—and the motor and driver. The resulting reflections can produce voltage peaks as high as two or three times the dc power bus link voltage. Safe driver design requires an increased voltage rating for the switching elements to handle these higher voltage levels. Higher voltage switches have a higher cost and higher on-resistance (given the same semiconductor footprint). The higher on-resistance of these higher voltage switches contributes to reduced driver efficiency. Going to larger footprint devices adds to cost as well as requires more drive energy to switch these elements.
The motor’s insulation integrity is improved by eliminating or minimizing the higher voltage spikes which result from line reflections. In regards to line reflections, with typical insulation, the signal propagates at approximately 0.6 times the speed of light-some 20 cm/nanosecond. A 10 cm typical internal wiring would allow the signal to reflect back to the source in approximately 1 ns. If the switch has a 25 ns rise time, only about 4% of the total voltage amplitude would have changed in the reflection period, thus the reflection is held to approximately 4% of the buss voltage. An external cable of length 2.5 m or longer-common in nonintegrated configurations-would allow the full reflection of the drive pulse if impedance matching terminations are not used.
The insulation build (thickness) typically needs to be significantly increased for PWM-driven motors if being driven at significant distances from the driver that has connecting cables. For higher voltage drives, the peaks can triple the drive voltage, resulting in spikes high enough to generate corona. The corona damages the motor structure mechanically, degrading the insulation.
However, the ozone created by the corona can do more severe and widespread damage as well. Ozone damages the insulation and many plastic components within the motor. Rubber components, including seals, may be particularly degraded. Integrating the motor with the controller keeps the wiring from the motor to the driver very short, reducing the propagation time for the signals through these wires to less than the rise-and-fall times of the drivers.
When the signal can reach the load and return to the source in a fraction of the rise time, only that fraction of the voltage change is effectively reflected, keeping the resulting spikes from reflections to a few percent of the switching voltage rather than 300%. This directly tackles the corona problem by significantly reducing the peak, spike voltage present.
Electrical noise emission and susceptibility are also significantly improved by eliminating the long runs between the driver and the motor and between the feedback and the controller. Long wires make great antennas unless they are carefully filtered and shielded. Eliminating the antennas by keeping the lengths short and the loop areas small reduces both E-field and H-field emissions.
Similarly, the encoder or resolver signals are easier to keep clean if they are not running through long runs, especially not in long runs alongside PWM driven motor signals! Often, without these "antennas," the degree of filtering on the driver side also can be significantly reduced, saving bulk and cost, and generally improving efficiency.
Ease of integration and start-up is enhanced with integrated motors and controllers. With reduced electrical noise, reduced wiring, and the controller already cognizant of the motor requirements and feedback configuration, the setup is simplified. Torque-speed curves reflect the whole package, and there is less misunderstanding than when combining multiple devices (including cabling). The power input and communications signals usually are much easier to deal with for the typical integrator than the intricacies of controlling cable reflections, proper shielding and grounding, and the proper use of ferrite beads to break alternating current (AC) ground loops.
Nonintegrated solutions may still have some advantages when significant input/output (I/O) and communications wiring must be brought out to each axis, but the improvement of communications busses and standards is reducing the advantage that once favored a compact control panel with motors far from their control. This was particularly true of separate drivers and controllers, where the wire that is running from the controller to the driver is often more sensitive to noise. The nonintegrated configurations may also be needed for special motors, very harsh environments (such as radiation), or where very small motors are in use.
Integrated motor, controllers, and drivers can significantly simplify the design and integration process of many systems, while improving efficiency, electro-magnetic compatibility, and packaging size and improving overall reliability.
Donald Labriola, PE, is president of QuickSilver Controls Inc. Edited by Emily Guenther, associate content manager, Control Engineering, CFE Media, firstname.lastname@example.org.
- The benefits of motor and drive integration
- How integrated motors and drives improve operations
- Advantages for nonintegrated solutions.
What kind of maintenance is required by implementing an integrated motor and drive solution?