Many battery-powered, wheel-based industrial robots, such as warehouse robots, use continuous rotation servo motors, or hub in-wheel motors specifically, for providing mobility. These types of servo motors are typically DC brushless motors and use some form of position sensing for providing feedback in their closed-loop motor-control systems. Over the years, resolvers, optical encoders, resistive potentiometers, and magnetic position sensors have been commonly used for providing the motor position feedback for these types of motors. However, each has its unique advantages and disadvantages. New RF inductive position sensors coming onto the market today offer a competitive cost and performance alternative to these other technologies.
This article will explain how an electric DC hub motor works and is used in wheel-based mobile robots. It will review the pros and cons of older position-sensor technologies used in servo motor controller feedback circuits.
In addition, it will provide a detailed explanation of RF inductive position sensors, describe how they are used in motor encoder applications, and compare their performance capabilities and cost benefits to the other position-sensing technologies.
How a brushless hub motor works
A hub (in-wheel) motor is typically a brushless compact DC electric servo motor built into a wheel, be it for a mobile robot, or even for an electric bike or automobile.
A traditional electric motor has two major components: a stator and a rotor. As the name suggests, the stator is stationary. It is the hollow outer ring of the motor and has a ring-shaped permanent magnet in it.
Sitting inside the stator is a rotor, which consists of an inner metallic core shaft surrounded by a tightly wound wire coil(s). Again, as the name suggests, the rotor rotates, and the turning shaft is employed to drive wheels or gears.
A hub motor works differently. It still has a stator and a rotor, but their roles are reversed. The rotor is held fixed, and the stator turns. Since the outer ring of the motor turns, it is ideal for simply attaching it to the hub of a wheel, and thus the phrase in-wheel hub motor.
This type of wheel drive is much more efficient than having to use a gear box to redirect a traditional motor’s rotating shaft in another direction to turn the wheels on a mobile robot, or electric vehicle.
In addition, a hub in-wheel motor can be employed on each wheel on the moving robot base to provide the required rotational energy and torque to move the robot at the desired speed and direction.
Robot manufacturers prefer brushless continuous rotation servo hub motors over brushed motors, because the former operate more efficiently and are much more reliable due to the fact that there are no brushes to wear out and/or get contaminated with dust (see photo below).
Brushless hub motors, however, need a method for providing angular position feedback of the rotating stator for the motor controller to provide the correct drive signals at the correct time. Otherwise, the motor will fail to continue to rotate.
Resolvers, optical encoders, Hall-effect latches, and rotary magnetic encoders have filled this need in the past. We’ll now explore the advantages and disadvantages of each in more detail.
Types of position sensor encoders used in servo motors
On one end of the position-sensing spectrum are resolvers and optical encoders. Resolvers support very high RPM rates and provide high resolution, but they are large, heavy, power-consuming, and expensive.
Optical encoders also support very high RPM rates and provide high resolution, but they are likewise expensive and are susceptible to dust and dirt, two items commonly found in warehouses and industrial buildings (see chart, below).
At the opposite end of the spectrum are resistive potentiometers. They are very inexpensive encoders, but are not very accurate, and wear out fast. As a result, they are typically used in low-cost servo motors targeted for cheap toy robots.
On the higher end of the spectrum, closer to the performance of resolvers and optical encoders, are magnetic position sensors.
The most basic type of magnetic position sensor used for position feedback in motors is the Hall-Effect latch. It is a very simple magnetic sensor integrated circuit (IC) device that changes its output state when a magnet passes by it of some particular magnetic polarity and field strength, e.g. a magnetic south pole. The device remains in that state until a magnetic field of equal strength, but of opposite polarity passes by it, e.g. a magnetic north pole.
Hall-Effect latches are very inexpensive and reliable, however they’re not that accurate and typically three or more are needed depending the number of phases in the motor.
The other type of magnetic position sensor is the magnetic encoder IC. It is a much more sophisticated magnetic sensor that measures rotary/angular position. Though more expensive than the Hall-Effect IC, they are still relatively inexpensive compared to resolvers and optical encoders. In addition, unlike the Hall-Effect position feedback circuit, only one magnetic encoder IC is required.
However, though magnetic encoder ICs are relatively inexpensive and small, they can’t support the very high RPM rates and accuracy levels that resolvers and optical encoders can achieve. And like most magnetic sensor ICs, with the exception of ams’ devices, they are susceptible to magnetic stray fields, which are commonly found in industrial facilities. Stray magnetic fields can cause a magnetic sensor to provide inaccurate and reliable position information back to the motor controller.
Introducting the RF inductive position sensor
Fortunately today, manufacturers of mobile wheel-based robots have an alternative to the traditional position sensor technologies used in brushless hub motors. RF inductive position sensors address many of the disadvantages found in the other position-sensing technologies just mentioned, while still providing the very high RPM and resolution performance of resolvers and optical encoders.
In addition, they provide unique features, and support expanded applications that enables lower-cost motor control circuits.
How a RF inductive position sensing system works
RF inductive position sensor ICs are normally mounted on a printed circuit board (PCB) that also includes a transmit coil and two receive coils. Typically, there is a circular hole in the center of the PCB.
Positioned above the PCB is a metal target, or multi-blade vane often resembling a propeller, that is attached to a rotating shaft, and that passes through the PCB’s center hole (see below).
The inductive position sensor IC, via its on-chip oscillator, radiates out over the PCB’s transmit coil an electromagnetic field. The radiating energy is absorbed by the PCB’s receive coils and measured via the inductive sensor’s dual receive circuits.
However, as the metal blades rotate over the PCB operating area, the metal blades shunt, or attenuate, the transmit coil’s radiating electromagnetic field energy over portions of the receive coils. This attenuation is detected by the inductive sensors receive circuits, and results in the inductive sensor IC outputting two sinusoidal outputs, typically 90 degrees out of phase from one another, e.g. Sin/Cos outputs.
A microcontroller receives the inductive position sensor’s Sin/Cos outputs and calculates the rotating shaft’s absolute or incremental angular position based on the phase and amplitude difference between the inductive sensor IC’s Sin/Cos outputs. This angular measurement is used to accurately determine the angular position feedback of the rotating stator, in the case of a hub motor, so that the microcontroller can properly drive the various stationary rotor coils and keep the hub motor spinning.
Unique advantages of RF inductive position sensors
Inductive position sensing offers a number of unique advantages over competing position sensing solutions. First, it provides high linearity, yielding very accurate angle measurements, comparable to resolver or optical encoders. Inductive position sensors for motor control applications, for example, can achieve linearity accuracies down to +/- 0.2o, even at rotational shaft rates that exceed 100,000 RPMs.
Second, unlike on-axis magnetic encoder ICs, inductive position sensors can support off-axis, hollow-shaft, and side-shaft applications. In many motor applications, for example, it is not possible to mount a sensor on the end of a solid metal rotating shaft, such as is the case with a hub drive motor that is directly driving a robot’s wheel. Off-axis angular position sensing gets around this problem.
Similarly, robot manufacturers often need to run wires through robots’ rotating hollow shafts to reduce limb diameters and protect the control/power wires from becoming damaged during operation (see below).
Third, inductive position sensors are immune to magnetic stray fields, a key benefit for industrial robots that typically find themselves in very electromagnetically noisy environments, such as in a factory where there can be thousands of poorly shielded motors running at any particular moment and radiating out electromagnetic fields.
Fourth, inductive position sensors offer the ability to cost effectively support safety standards, such as IEC 61508 and IEC 62061 for industrial robots. Due to their low unit costs, redundant inductive position sensor ICs can be placed on a PCB, and/or alternatively redundant transmit and receive coil pairs can be added to the PCB.
Moreover, smart inductive position sensors can also provide diagnostic information on their functional status, as well as their paired transmit and receive coils status, to the motor controller, thus enabling a robot to achieve higher Safety Integrity Levels (SILs).
Inductive position sensors are also much less expensive than resolvers and optical encoders, and even many rotary magnetic position sensor solutions. An inductive sensor is a fairly simple integrated circuit, comprised mainly of an oscillator, and a pair of differential receivers and associated automatic gain control circuits, along with a few input/output buffers.
Though an inductive sensor IC needs a PCB with transmit and receive coils, along with a target metal vane, both have very low unit costs associated with them. The target magnet required to operate with a magnetic position sensor commonly has a higher unit cost than the inductive sensor’s required PCB and metal vane, combined. And in relationship to resolvers and optical encoders, they typically cost one to two orders of magnitude more than a complete inductive position sensor solution.
Inductive position sensors are also power misers. Because of their limited size and circuitry, they require very little power, typically less than 15ma, worst case, putting them on par with rotary magnetic position sensors, and much lower than resolvers and optical encoders.
To conclude, new RF inductive position sensor solutions on the market today provide many unique end-user advantages over traditional positional sensor technology used in motor control commutation feedback circuits. This is particularly true in the case of battery-powered mobile wheel-based industrial robots.
Inexpensive, low power-consuming inductive position sensor ICs, immune to stray magnetic fields, and dust and dirt commonly found in factories and warehouses, and that have the precision and performance of resolvers and optical encoders, make them the ideal feedback solution for in-wheel hub motors used in mobile robots.