In automation control, various motors are frequently encountered for control. In conveyor belts, elevators, lifting trolleys, and other high-power motors, most of them are variable frequency motors. It is also very common for various brands of PLC + frequency converter to drive and control variable frequency motors.
However, users often encounter various problems:
Why do variable frequency motors need encoders?
Is it okay to not install an encoder?
If an encoder is installed on a variable frequency motor, can it be used for asynchronous servo control? Can it then be used for positioning control?
Why do some variable frequency motor controllers not only have one encoder, but also a dual encoder closed loop?
Some people say, "Variable frequency motors are not good for positioning or synchronization; to achieve synchronous control, you need to switch to a synchronous servo motor."
The encoder signal of a variable frequency motor is often interfered with and is easily damaged. How should you choose an encoder?
This article will first discuss why encoders are necessary for variable frequency motors.
one
Basic concept: Variable frequency motor drives do not have a position loop. The encoder on a variable frequency motor is a "speed encoder," used for accurate calculation of the motor's back electromotive force (EMF). The motor's back EMF is directly proportional to the motor's rotor speed.
Due to the widespread use of servo motors, many control approaches are now being compared and evaluated against them, even though frequency converter control predates servo control. Servo motor control is a closed-loop control system consisting of position, speed, and torque loops, which is reflected in the design principles of permanent magnet synchronous motors. The phase of the drive current is synchronized with the rotor position, and the servo motor's drive inherently makes the position loop a "natural" closed loop. In contrast, frequency converter motors are asynchronous, sometimes called asynchronous motors. Even with feedback from the encoder at the rear of the motor, they only have a speed loop and lack a "position loop" in the motor drive; therefore, this encoder is called a "speed encoder."
As a speed encoder, the main purpose of a variable frequency motor encoder is to calculate the back electromotive force of the motor rotor in order to achieve precise drive control corresponding to the current back electromotive force of the motor.
When the driving current starts the motor rotor to rotate, according to the laws of electromagnetism, when the magnetic field changes, a nearby conductor will generate an induced electromotive force (EMF). The direction of this EMF conforms to Faraday's law and Lenz's law, and is exactly opposite to the voltage originally applied across the coil. This voltage is the back EMF.
According to the law of conservation of energy:
The electrical energy delivered by the motor driver = mechanical energy (driven current and back electromotive force are balanced) + losses (motor current impedance heat loss, mechanical resistance, resistance box heat loss, etc.).
When a motor starts and accelerates, the rotational potential energy generated by the drive current must be greater than the back electromotive force (the vector is positive), but it cannot be too large, as excessive current will be lost as heat in the motor and the resistance box. The feedback from the speed encoder provides the inverter with the back electromotive force to calculate so that the drive rotational potential energy is exactly greater than the back electromotive force.
Each motor has its own characteristic constant, and the back electromotive force is directly proportional to the motor rotor speed and this characteristic constant.
Back electromotive force = characteristic constant × rotor speed
For variable frequency motors equipped with encoders, the encoder signal is fed back to the variable frequency drive, which calculates the current back electromotive force of the motor and provides a reasonable control current.
When the encoder feeds back the signal to the inverter and calculates that the motor speed is too low, far below the rotational speed that the motor should reach under the corresponding designed drive current, this is called motor drive "stall". Stall of the variable frequency motor means that the back electromotive force is too low, and the electrical energy is used for heat loss (the back electromotive force is low, and the voltage is distributed to the impedance). At this time, the motor coil impedance is low, the current increases and the motor gets hot, or the inverter current is too high, which may burn out the motor or inverter. In this case, stall protection is required to stop the motor drive.
To address the potential stalling of variable frequency motors, an early approach was to design both the motor and inverter to be significantly larger, with sufficient margin to handle high current heat loss and prevent damage to the motor or inverter components. This also required a large resistor distribution box to balance excess energy during instantaneous startup. This resulted in a wasteful design that resulted in a large motor and inefficient inverter. Furthermore, a significant amount of energy was wasted on heat loss during motor acceleration.
Adding an encoder to a variable frequency motor can improve the energy efficiency of both the motor and the frequency converter during startup, reducing the possibility of damage to both.
An analogy that experienced drivers know is that vehicles consume the most fuel when starting from low speeds, and even more when accelerating uphill. The same principle applies to electric motors; most of their energy loss occurs during startup and acceleration. For variable frequency motors to truly achieve energy savings, it's best to add encoder feedback. This allows for precise current control during startup, reducing energy waste from heat and protecting both the motor and the inverter from damage.
Therefore, if the encoder for a variable frequency motor is selected and installed properly, the benefits of installing an additional encoder far outweigh the price of an encoder, due to the increased efficiency of the motor and the inverter, the reduction of damage and failures, and the real realization of the energy-saving effect of the variable frequency motor.
two
In vector control mode, encoder feedback can improve the execution of acceleration-torque control.
Vector refers to directional control. When the potential energy of a motor drive remains positive relative to the back electromotive force, it results in acceleration.
When the potential energy driven by the motor remains negative relative to the back electromotive force, it is considered deceleration.
Vector control is a fine-grained control of the acceleration and deceleration of a motor, especially in terms of the accuracy of the motor's acceleration at low speeds during startup and deceleration to stop (low-speed range).
In terms of Newton's second law: F = kma; F = force; m = mass; k = inertia constant; a = acceleration.
Acceleration (deceleration) corresponds to motor torque, and vector control corresponds to the execution force of motor torque control. To achieve the accuracy of vector control, precise feedback of rotor acceleration is required, and it is best to use an encoder as the feedback sensor for acceleration calculation.
Some motors use Hall effect sensors for speed and acceleration feedback; others use sensorless solutions that utilize the motor's own coils to sample and calculate back electromotive force. However, both Hall effect sensor and sensorless solutions have poor feedback accuracy at low speeds. This means that vector control lacks precision during motor startup at low speeds and during deceleration and stopping, resulting in coarse control.
Sensor feedback with a speed encoder is typically 1024 PPR pulse feedback, which is more accurate than Hall effect sensors or the back electromotive force of the motor coil itself without sensors. It is especially effective for energy saving during low-speed startup and for precise positioning during deceleration and stopping.
Returning to the question at the beginning of this article, is it okay to not install an encoder on a variable frequency motor?
Of course, it's possible. However, it lacks the feedback accuracy of speed and acceleration at low speeds, and therefore the drive control accuracy is also lost at low speeds. Furthermore, motor energy consumption is highest during motor start-up and acceleration (low speed), and more than 50% of motor and inverter failures occur during this period.
three
Asynchronous servo control mode requires dual encoders in a closed loop – this addresses the response delay of asynchronous motor acceleration and deceleration and the accuracy issues of the reducer.
As discussed earlier, asynchronous motors differ from synchronous motors in that they lack a position closed loop in the drive mechanism; instead, they rely on the integral of speed over time to obtain the position. We know that servo control refers to a three-loop closed-loop control consisting of a position loop, a speed loop, and a torque loop. The position and speed loops should ideally be independent. Although calculations like position change/time = speed, and speed x time = position exist, such calculations are feasible for synchronous motors but not for asynchronous motors—due to differences in error and execution response delay. Because it's asynchronous control, assuming the error from speed feedback to execution response is controlled within one-thousandth per second (which is already considered good accuracy for asynchronous control), the cumulative position integral error over one thousand seconds could potentially reach a thousand times the error per second! (17 minutes). Therefore, some asynchronous motor controllers use the encoder installed on the motor to directly perform position closed-loop calculations, instead of using the speed loop integral to obtain the motor position. However, this encounters another problem—the actual required process end is at the mechanical transmission end, and the errors and delays in the mechanical transmission and reducer affect the position control execution force.
Precision reducer problems
The function of a speed reducer is based on the lever principle. At the input end of the speed reducer (fulcrum), a smaller force (using a smaller power motor) travels a greater distance (before reduction, the motor speed is higher, and the number of revolutions is greater). Through the speed reducer's reduction ratio, the output end of the speed reducer rotates a smaller distance (after reduction, the output speed is lower, and the number of revolutions is less), yet obtains amplified output force. This is the lever principle. However, using a speed reducer is not simply about lever reduction ratios. It also brings problems such as loss of mechanical precision, mechanical wear, mechanical friction and other resistance efficiency losses, reduced precision and lifespan of transmission gears, and loss of input-output time response of the speed reducer.
The widespread use of synchronous servo motors is currently concentrated in lower-power applications. The design focus of gearboxes for small servo motors is primarily on precision, with less emphasis on other crucial parameters such as torque efficiency, material properties, and mechanical wear. In contrast, asynchronous motors are often used for higher power outputs. Gearbox manufacturers prioritize material properties, mechanical wear, and output torque efficiency, making it difficult to achieve the same level of precision at the end effector as smaller servo motor gearboxes. Unless used in robotic arms, where gearboxes must simultaneously guarantee precision, torque, and wear resistance of the materials, robot gearboxes are highly sought after. Currently, the robot gearbox market is largely monopolized by two Japanese companies, precisely because of the difficulty in creating gearboxes that deliver high torque, wear resistance, and high precision. Most other asynchronous servo motors cannot be equipped with such expensive and monopolized precision gearboxes, forcing them to sacrifice end-effector mechanical position accuracy.
If the control accuracy requirements of the asynchronous servo "position loop" are also required, the solution is to add a "position encoder" at the mechanical end. This encoder at the mechanical transmission end is also called the "second encoder" or "load-side process shaft encoder." For example, it can be installed at the output end of the reducer, as shown in the figure below.
Thus, there is a high-speed speed encoder at the tail of the motor shaft and a low-speed position encoder at the mechanical end of the motor reduction drive. Based on the application characteristics of the encoders, an incremental pulse encoder is selected for the speed encoder, with the pulse frequency corresponding to the speed; an absolute encoder (multi-turn range) is selected for the position encoder. The absolute encoder's encoding is unique for each position, eliminating the need for a counter and preventing concerns about interference and error accumulation, which perfectly corresponds to the position loop control.
This leads to the question posed at the beginning of this article: what exactly is dual encoder closed-loop control?
Because asynchronous motor drives lack a position loop, and due to the errors of the reducer and the position response delay, asynchronous servo control (or position loop positioning control for variable frequency motors) requires two encoders: a speed encoder and a position encoder, which are separate (unlike synchronous motors). Because of the accumulated errors, the speed encoder cannot use time integration to obtain the position loop; and because of the transmission response delay, the position loop encoder cannot use time differentiation to obtain the speed loop, resulting in insufficient control accuracy.
For variable frequency motors, it is futile to use only one encoder as both a speed closed loop and a position closed loop. With only one encoder, it can only serve one function: either a speed closed loop or a position closed loop, not both.
Two encoders need to be installed in two locations: a speed encoder on the rear of the variable frequency motor shaft and a position encoder installed at the end of the mechanical transmission.
This control method is a "dual encoder closed-loop" mode for variable frequency motors and asynchronous servo control.
A typical example is the variable blade control of wind power generation. The following diagram shows the dual encoder closed-loop control principle of the KEB driver P6, which was displayed at the China International Industry Fair.
Four
Position control using PLC, frequency converter, and encoder: variable frequency motors can also be well positioned.
Before servo motors became widely adopted, position-based closed-loop positioning using ordinary motors, encoders, and PLCs was already widely used. With the addition of frequency converters, this application became even more prevalent. Numerous examples of frequency converter + encoder + PLC applications exist, including oxygen lance positioning, feeding trolley positioning, and converter tilting positioning in steel mills; gantry crane lifting, double hook synchronization, and trolley travel synchronization correction in port machinery; and mold opening and closing positioning in injection molding machines. Compared to synchronous servo motors, the main difference lies in the increased demands for speed response and accuracy.
Automated positioning control of port machinery
In principle, a motor has four outputs: force, speed, angular position, and wasted resources: useless spatial electromagnetic field, back electromotive force, and heat loss. Asynchronous motors (variable frequency motors) were initially designed as force output devices (power, torque), but later, variable frequency controllers were added to adjust speed. If a mechanical end-position encoder is added to a PLC, the speed (deceleration control) and braking positioning can be adjusted via PLC commands to the inverter. The position closed loop follows the "outer loop." The quality of the position closed loop depends on the "time delay" (response) between the encoder signal feedback to the PLC and the command output to the inverter, and the difference between the inverter's execution capability (accuracy) after receiving the command and performing deceleration, braking, and positioning. For the first point, if the inverter has a built-in PG card with a dual-encoder closed-loop scheme, this time delay can be reduced. If the encoder signal goes to the PLC, and the PLC command goes to the inverter, then a signal transmission scheme with the fastest possible speed must be designed to reduce this time delay. Regarding the second point, the inverter's deceleration and braking execution capability, this requires an expert familiar with the inverter manual to advise on which execution capability is better. The solution depends on how to reduce time delay (response) and how to improve the deceleration and braking performance (accuracy) of the frequency converter.
The dual-encoder closed-loop scheme, when combined with vector frequency conversion control, allows the speed encoder signal installed on the variable frequency motor to enter the vector frequency converter, improving the control execution capability of the frequency converter for acceleration and deceleration; the position encoder signal installed at the end of the mechanical transmission enters the PLC, which can improve the position control accuracy and position arrival response.
Returning to the question raised at the beginning of this article, some people say that "variable frequency motors are not good at positioning and should be replaced with synchronous servo motors." That's because they only used an encoder and naturally compared it with a synchronous servo motor. However, variable frequency motors are asynchronously driven, and the position loop is not on the motor itself, but on the "outer loop."
In fact, even if the synchronous servo motor is replaced, the dual encoder closed-loop solution is still applicable to the control of high-power synchronous servo motors. If a reducer with a large torque output is used, the accuracy loss and response issues of the reducer will still exist. It is also necessary to add a second encoder (position encoder) at the low-speed load end.
A servo control system with a second encoder added to the low-speed end was showcased at the exhibition.
Several mounting positions of encoders in motion control, and the function of absolute encoders.
five
The problem with multi-motor synchronous control is position synchronization, not speed synchronization.
In the discussion above, we learned that variable frequency motors are asynchronous motors, and their speed and position loops are controlled independently, requiring feedback from two encoders at different positions. The "synchronization" problem of multiple motors requires clarifying the user's requirements: is it speed synchronization or position synchronization? These are two different control strategies, and their "synchronization" effects are different. In reality, most multi-motor synchronization problems require position synchronization, that is, synchronization of motor angular displacement. When calculating each time segment and the accumulated time segment, the angular displacement of multiple motors is always controlled within a certain deviation range; this is position synchronization. This means that for variable frequency motors to be synchronized, a second encoder, namely a position encoder, should be installed to provide the synchronization feedback signal. Examples include conveyor belt synchronization, left-right synchronization of elevators, and synchronization of the movements of variable frequency motors and servo motors.
We will discuss the synchronous control of multiple motors in a future article.
six
Selection Considerations for Speed Encoders and Position Encoders
Special considerations for selecting encoders due to the unique characteristics of variable frequency motors:
1. High motor power often results in unbalanced three-phase voltage during startup, which causes an instantaneous surge of AC induced current that creates a biased magnetic field, leading to signal interference and damage to the encoder.
2. The motor is large and has strong driving force, resulting in large axial movement during startup and mechanical impulse damage to the encoder shaft.
3. In the control scheme of PLC + frequency converter + encoder, the transmission distance of encoder signal into PLC is relatively long.
Speed encoders are incremental pulse encoders, and a resolution of 1024 PPR is generally sufficient.
Speed encoders should be incremental encoders with inverted signals. Due to the three-phase imbalance at motor startup, a centrifugal induced magnetic field is generated in space, causing biased bidirectional AC current coupling interference and impact on the encoder. Unipolar DC open-collector output modes, with energy accumulation when reverse current is cut off, are less able to withstand long-term bidirectional AC induced current impacts and are more prone to damage. Furthermore, the signal is easily interfered with; therefore, encoders with open-collector output signals are not recommended for variable frequency motors (NPN or PNP types are not recommended).
For variable frequency motor encoders, it is recommended to choose bipolar output mode encoders (A+A-B+B-Z+Z-), which have a 0V output channel corresponding to bidirectional induced current to dissipate the induced impact energy. This prevents impact energy buildup and reduces the likelihood of damage to the encoder circuitry. Simultaneously, the encoder should have a wide operating voltage range, the power supply should have reverse polarity protection, and the signal lines should have short-circuit protection to cope with the bidirectional AC coupling impact during the start-up of the variable frequency motor.
An incremental pulse signal encoder outputs a bidirectional A+A-B+B- signal of 5-24V.
Furthermore, bipolar signals with inverted phase are common-mode to external induction on twisted-pair cables, which can be eliminated at the receiving end using differential balancing. This results in strong signal anti-interference and facilitates long-distance signal transmission. The encoder signal transmission distance parameter should be selected to be at least greater than 50 meters. This is not because the actual field distance might only be a few meters, but because encoders used in variable frequency motors require higher anti-interference capabilities during signal transmission than those used in servo motors. The parameters of the selected encoder should be checked.
For information on selecting encoder signal transmission cables, please refer to my previous article.
Discussion of Ten Issues Regarding Encoder Signal Cables and Transmission Interference Immunity
The position encoder should be an absolute multi-turn encoder. Variable frequency motors often have high power, and the on-site interference environment is complex. The incremental pulse signal relies on a counter, which is easily interfered with during counting, leading to accumulated errors. This includes battery-powered or pseudo-absolute multi-turn counter types using Wiegand pulse counters, which are also easily interfered with during count, resulting in skipped turns and incorrect position. As mentioned earlier, in the dual-encoder closed-loop control of wind turbine blades, many imported brand electronic turn-counting pseudo-multi-turn absolute encoders have been replaced due to numerous failures. Therefore, for variable frequency motors (asynchronous servo control), a mechanical gearbox-type absolute multi-turn encoder should be selected. If the encoder signal goes directly to the inverter, it is generally sent as an SSI signal (the inverter must have a dual-encoder closed-loop solution). A typical example is the variable frequency drive control of wind turbine blades. If it goes to a PLC, it should be sent as a bus signal or Ethernet signal, such as CANopen, PROFIBUS-DP, or Profinet.
Due to its synchronous timestamp feature, the bus-type Ethernet EtherCat signal can be fed into both PLCs and motor drivers, making it ideal for multi-motor synchronous control, especially for synchronous control of different types of motors, such as the synchronous operation of variable frequency motors and servo motors, which can improve the efficiency of multi-motor synchronous linkage.
