Permanent magnet motors: Introduction (part 1)

Home | Articles | Forum | Glossary | Books


1. Permanent magnet versus electromagnetic excitation

The use of permanent magnets (PMs) in construction of electrical machines brings the following benefits:

• no electrical energy is absorbed by the field excitation system and thus there are no excitation losses which means substantial increase in efficiency,

• higher power density and/or torque density than when using electromagnetic excitation,

• better dynamic performance than motors with electromagnetic excitation (higher magnetic flux density in the air gap),

• simplification of construction and maintenance,

• reduction of prices for some types of machines.

The first PM excitation systems were applied to electrical machines as early as the 19th century, e.g., J. Henry (1831), H. Pixii (1832), W. Ritchie (1833), F. Watkins (1835), T. Davenport (1837), M.H. Jacobi (1839). Of course, the use of very poor quality hard magnetic materials (steel or tungsten steel) soon discouraged their use in favor of electromagnetic excitation systems.

The invention of Alnico in 1932 revived PM excitation systems; however, its application was limited to small and fractional horsepower DC commutator machines. At the present time most PM DC commutator motors with slotted rotors use ferrite magnets. Cost effective and simple DC commutator motors with barium or strontium ferrite PMs mounted on the stator will still be used in the foreseeable future in road vehicles, toys, and household equipment.

Cage induction motors have been the most popular electric motors in the 20th century. Recently, owing to the dynamic progress made in the field of power electronics and control technology, their application to electrical drives has increased. Their rated output power ranges from 70 W to 500 kW, with 75% of them running at 1500 rpm. The main advantages of cage induction motors are their simple construction, simple maintenance, no commutator or slip rings, low price and moderate reliability. The disadvantages are their small air gap, the possibility of cracking the rotor bars due to hot spots at plugging and reversal, and lower efficiency and power factor than synchronous motors.

The use of PM brushless motors has become a more attractive option than induction motors. Rare earth PMs can not only improve the motor's steady state performance but also the power density (output power-to-mass ratio), dynamic performance, and quality. The prices of rare earth magnets are also dropping, which is making these motors more popular. The improvements made in the field of semiconductor drives have meant that the control of brushless motors has become easier and cost effective, with the possibility of operating the motor over a large range of speeds while still maintaining good efficiency.

Servo motor technology has changed in recent years from conventional DC or two-phase AC motor drives to new maintenance-free brushless three-phase vector-controlled AC drives for all motor applications where quick response, light weight and large continuous and peak torques are required.

A PM brushless motor has the magnets mounted on the rotor and the armature winding mounted on the stator. Thus, the armature current is not transmitted through a commutator or slip rings and brushes. These are the major parts which require maintenance. A standard maintenance routine in 90% of motors relates to the sliding contact. In a DC commutator motor the power losses occur mainly in the rotor which limits the heat transfer and consequently the armature winding current density. In PM brushless motors the power losses are practically all in the stator where heat can be easily transferred through the ribbed frame or, in larger machines, water cooling systems can be used. Considerable improvements in dynamics of brushless PM motor drives can be achieved since the rotor has a lower inertia and there is a high air gap magnetic flux density and no-speed dependent current limitation.

The PM brushless motor electromechanical drive has become a more viable option than its induction or reluctance counterpart in motor sizes up to 10 - 15 kW. There have also been successful attempts to build PM brushless motors rated above 1 MW ( Germany and U.S.A.). The high performance rare-earth magnets have successfully replaced ferrite and Alnico magnets in all applications where high power density, improved dynamic performance or higher efficiency are of prime interest. Typical examples where these points are key selection criteria are stepping motors for computer peripheral applications and servo motors for machine tools or robotics.

2. Permanent magnet motor drives

In general, all electromechanical drives can be divided into constant-speed drives, servo drives and variable-speed drives.

A constant-speed drive usually employs a synchronous motor alone which can keep the speed constant without an electronic converter and feedback or any other motor when there is less restriction on the speed variation tolerance.

A servo system is a system consisting of several devices which continuously monitor actual information (speed, position), compare these values to desired outcome and make necessary corrections to minimize the difference. A servo motor drive is a drive with a speed or position feedback for precise control where the response time and the accuracy with which the motor follows the speed and position commands are extremely important.

In a variable-speed drive (VSD) the accuracy and the response time with which the motor follows the speed command are not important, but the main requirement is to change the speed over a wide range.

In all electromechanical drives where the speed and position are controlled, a solid state converter interfaces the power supply and the motor. There are three types of PM motor electromechanical drives:

• DC commutator motor drives

• brushless motor drives (DC and AC synchronous)

• stepping motor drives


FIG. 1. Basic armature waveforms for three phase PM brushless motors: (a) sinusoidally excited, (b) square wave.

Brushless motor drives fall into the two principal classes of sinusoidally excited and square wave (trapezoidally excited) motors. Sinusoidally excited motors are fed with three-phase sinusoidal waveforms (FIG. 1a) and operate on the principle of a rotating magnetic field. They are simply called sinewave motors or PM synchronous motors. All phase windings conduct current at a time.

Square wave motors are also fed with three-phase waveforms shifted by 120 degrees one from another, but these waveshapes are rectangular or trapezoidal (Fig. 1b). Such a shape is produced when the armature current (MMF) is precisely synchronized with the rotor instantaneous position and frequency (speed).

The most direct and popular method of providing the required rotor position information is to use an absolute angular position sensor mounted on the rotor shaft. Only two phase windings out of three conduct current simultaneously.

Such a control scheme or electronic commutation is functionally equivalent to the mechanical commutation in DC motors. This explains why motors with square wave excitation are called DC brushless motors. An alternative name used in power electronics and motion control is self-controlled synchronization.

Although stepping motor electromechanical drives are a kind of synchronous motor drive, they are separately discussed due to their different control strategies and power electronic circuits.

2.1 DC commutator motor drives

The DC commutator motor or DC brush motor is still a simple and low cost solution to variable-speed drive systems when requirements such as freedom from maintenance, operation under adverse conditions or the need to operate groups of machines in synchronism are not supreme. Owing to the action of the mechanical commutator, control of a DC motor drive is comparatively simple and the same basic control system can satisfy the requirements of most applications. For these reasons the DC electromechanical drive very often turns into the cheapest alternative, in spite of the cost of the DC commutator motor. In many industrial applications such as agitators, extruders, kneading machines, printing machines, coating machines, some types of textile machinery, fans, blowers, simple machine tools, etc., the motor is required only to start smoothly and drive the machinery in one direction without braking or reverse running. Such a drive operates only in the first quadrant of the speed-torque characteristic and requires only one controlled converter (in its rectifier mode) as shown in FIG. 2a. At the expense of increased torque ripple and supply harmonics, a half-controlled rather than fully-controlled bridge may be used up to about 100 kW. If the motor is required to drive in both forward and reverse directions, and apply regenerative braking, a single fully controlled converter can still be used but with the possibility of reversing the armature current (FIG. 2a).

Electromechanical drives such as for rolling mills, cranes and mine winders are subject to rapid changes in speed or in load. Similarly, in those textile, paper or plastics machines where rapid control of tension is needed, frequent small speed adjustments may request rapid torque reversals. In these cases a four-quadrant dual converter comprising two semiconductor bridges in anti-parallel, as in FIG. 2b, can be used. One bridge conducts when armature current is required to be positive, and the second bridge when it is required to be negative.

Natural or line commutation between successively conducting power semi conductor switches takes place when the instantaneous values of consecutive phase voltages are equal and cross each other. The phase with a decreasing voltage is suppressed, while that with an increasing voltage takes over the conduction. Natural commutation may not be possible in such situations when the armature inductance is very small, e.g., control of small DC motors. The normal phase control would cause unacceptable torque ripple without substantial smoothing which, in turn, would impair the response of the motor. Power transistors, GTO thyristors or IGBTs can be turned off by appropriate gate control signals, but conventional thyristors need to be reverse biased briefly for successful turn off. This can be accomplished by means of a forced-commutation circuit, usually consisting of capacitors, inductors and, in some designs, auxiliary thyristors. Forced commutation is used mostly for the frequency control of AC motors by variable-frequency inverters and chopper control of DC motors.


FIG. 2. DC commutator motor drives: (a) one quadrant, fully-controlled, single converter drive, (b) four-quadrant, fully-controlled, dual converter drive; Lp are re actors limiting the currents circulating between the rectifying and inverting bridges, La is the armature circuit inductance.


FIG. 3. Chopper controlled DC motor drives: (a) one-directional with added re generative braking leg, (b) four-quadrant chopper controller.

FIG. 3a shows the main components of a one-directional chopper circuit for controlling a PM or separately excited DC motor. Thyristors must be accompanied by some form of turn-off circuits or, alternatively, be replaced by GTO thyristors or IGBTs. When the mean value of the chopper output voltage is reduced below the armature EMF, the direction of the armature current cannot reverse unless T2 and D2 are added. The thyristor T2 is fired after T1 turns off, and vice versa. Now, the reversed armature current flows via T2 and increases when T1 is off. When T1 is fired, the armature current flows via D2 back to the supply. In this way regenerative braking can be achieved.

Full four-quadrant operation can be obtained with a bridge version of the chopper shown in FIG. 3b. Transistors or GTO thyristors allow the chopper to operate at the higher switching frequencies needed for low-inductance motors. By varying the on and off times of appropriate pairs of solid switches, the mean armature voltage, and hence speed, can be controlled in either direction. Typical applications are machine tools, generally with one motor and chopper for each axis, all fed from a common DC supply.

2.2 Synchronous motor drives

In a two-stage conversion a DC intermediate link is inserted between the line and motor-side converter (FIG. 4). For low power PM synchronous motors (in the range of kWs) a simple diode bridge rectifier as a line-side converter (FIG. 4a) is used. The most widely used semiconductor switch at lower power permitting electronic turn-off is power transistor or IGBT.

Motor-side converters (inverters) have load commutation if the load, e.g., PM synchronous motor can provide the necessary reactive power for the solid state converter. FIG. 4b shows a basic power circuit of a load-commutated current source thyristor converter. The intermediate circuit energy is stored in the inductor. The inverter is a simple three-phase thyristor bridge. Load commutation is ensured by over-excitation of the synchronous motor so that it operates at a leading power factor (leading angle is approximately 30 degrees).

This causes a decrease in the output power. The elimination of forced com mutation means fewer components, simpler architecture, and consequently lower converter volume, mass, and losses. A four-quadrant operation is possible without any additional power circuitry. The motor phase EMFs required for load commutation of the inverter are not available at standstill and at very low speeds (less than 10% of the full speed). Under these conditions, the current commutation is provided by the line converter going into an inverter mode and forcing the DC link current to become zero, thus providing turn-off of thyristors in the load inverter.




FIG. 4. Basic power circuits of DC link converters for synchronous motors with: (a) PWM transistor inverter, (b) load-commutated thyristor CSI, (c) forced-commutated GTO VSI (four-quadrant operation), (d) IGBT VSI.

The maximum output frequency of a load-commutated current source inverter (CSI) is limited by the time of commutation, which in turn is deter mined by the load. A CSI is suitable for loads of low impedance.

A voltage source inverter (VSI) is suitable for loads of high impedance. In a VSI the energy of the intermediate circuit is stored in the capacitor. A PWM VSI with GTOs and antiparallel diodes (FIG. 4c) allows a synchronous motor to operate with unity power factor. Synchronous motors with high subtransient inductance can then be used. Four-quadrant operation is possible with a suitable power regeneration line-side converter. Replacement of thyristors by GTOs or IGBTs eliminates inverter commutation circuits and increases the pulse frequency.


FIG. 5. Cycloconverter synchronous motor drive.

The maximum output frequency of the load-commutated CSI is limited to about 400 Hz even if fast thyristors are used. A higher output frequency can be achieved using an IGBT VSI with antiparallel diodes. FIG. 4d illustrates a typical PM brushless motor drive circuit with a three-phase PWM IGBT inverter. In the brushless DC mode, only two of the three phase windings are excited by properly switching the IGBTs of the inverter, resulting in ideal motor current waveforms of rectangular shape. There are six combinations of the stator winding excitation over a fundamental cycle with each combination lasting for a phase period of 60 degree. The corresponding two active solid state switches in each period may perform PWM to regulate the motor current. To reduce current ripple, it is often useful to have one solid state switch doing PWM while keeping the other switch conducting.


FIG. 6. DC brushless motor drive.


FIG. 7. Stepping motor drive.

A cycloconverter is a single stage (AC to AC) line-commutated frequency converter (FIG. 5). Four-quadrant operation is permitted as the power can flow in both directions between the line and the load. A cycloconverter covers narrow output frequency range, from zero to about 50% of the input frequency. Therefore, cycloconverters are usually used to supply gearless electromechanical drives with large power, low speed synchronous motors. For example, synchronous motors for ships propulsion are fed from diesel alternators mostly via cycloconverters. A cycloconverter has an advantage of low torque harmonics of relatively high frequency. Drawbacks include large number of solid state switches, complex control, and poor power factor. Forced commutation can be employed to improve the power factor.

2.3 PM DC brushless motor drives

In PM DC brushless motors, square current waveforms are in synchronism with the rotor position angle. The basic elements of a DC brushless motor drive are: PM motor, output stage (inverter), line-side converter, shaft position sensor (encoder, resolver, Hall elements), gate signal generator, current detector, and controller, e.g., microprocessor or computer with DSP board. A simplified block diagram is shown in FIG. 6.

2.4 Stepping motor drives

A typical stepping motor drive (FIG. 7) consists of an input controller, logic sequencer and driver. The input controller is a logic circuit that produces the required train of pulses. It can be a microprocessor or microcomputer which generates a pulse train to turn the rotor, speed up and slow down. The logic sequencer is a logic circuit that responds to step-command pulses and controls the excitation of windings sequentially. Output signals of a logic sequencer are transmitted to the input terminals of a power drive which turns on and turns off the stepping motor windings. The stepping motor converts electric pulses into discrete angular displacements. The fundamental difference between a stepping motor drive and a switched reluctance motor (SRM) drive is that the first one operates in open loop control, without rotor position feedback.

3. Toward increasing the motor efficiency

Unforeseen consequences can result from problems the contemporary world currently faces, i.e., ...

• fears of depletion and expected scarcities of major non-renewable energy resources over the next several decades

• increase in energy consumption

• pollution of our planet


FIG. 8. World net electric power generation, 1990-2030 (history: 1990-2005, projections: 2005-2030).


FIG. 9. World energy generation by fuel, 2005-2030.


FIG. 10. World energy-related carbon dioxide emissions, 2005-2030.

The world consumption of petroleum is about 84 million barrels per day (1 barrel = 159 l) or about 31 billion barrels (about 5×10^12 l) per year. If current laws and policies remain unchanged throughout the projection period, world marketed energy consumption is projected to grow by 50 percent over the 2005 to 2030 period.

[1. The Organization for Economic Co-operation and Development (OECD) is an international organization of thirty countries that accept the principles of representative democracy and free-market economy.]

About 30% of primary energy is used for generation of electricity. World net electricity generation will increase from 17,300 TWh (17.3 trillion kWh) in 2005 to 24,400 TWh in 2015 and 33,300 TWh in 2030 (FIG. 8). Total non-OECD 1 electricity generation increases by an average of 4.0 % per year, as compared with a projected average annual growth rate in OECD electricity generation of 1.3 % from 2005 to 2030.

The mix of energy sources in the world is illustrated in FIG. 9. The 3.1 % projected annual growth rate for coal-fired electricity generation worldwide is exceeded only by the 3.7-percent growth rate projected for natural-gas-fired generation.

The industrial sector, in developed countries, uses more than 30% of the electrical energy. More than 65% of this electrical energy is consumed by electric motor drives. The number of installed electrical machines can be estimated on the basis of their world production.

The increasing electrical energy demand causes tremendous concern for environmental pollution (FIG. 10). The power plants using fossil and nuclear fuel and road vehicles with combustion engines are main contributors to air pollution, acid rain, and the greenhouse effect . There is no doubt that electric propulsion and energy savings can improve these side effects considerably. For example, the population of Japan is about 50% of that of the U.S. However, carbon emission is four times less (400 million metric ton in Japan versus over 1550 million metric ton in the U.S. in 2005). Mass public transport in Japan based on modern electrical commuter and long distance trains network plays an important part in reduction of carbon emission. It has been estimated that in developed industrialized countries, roughly 10% of electrical energy can be saved by using more efficient control strategies for electromechanical drives. This means that electrical machines have an enormous influence on the reduction of energy consumption. Electrical energy consumption can be saved in one of the following ways:

• good housekeeping

• use of variable-speed drives

• construction of electric motors with better efficiency

Good housekeeping measures are inexpensive, quick and easy to implement.

The simplest way to save energy costs is to switch idling motors off. Motors can be switched off manually or automatically. Devices exist that use either the input current to the motor or limit switches to detect an idling motor. When larger motors are being switched off and on, the high starting current drawn by the motor could cause supply interference and mechanical problems with couplings, gearboxes, belts, etc., which deteriorate from repeated starting.

These problems can be avoided by using electronic solid state converters.

Fan and pump drives employ over 50% of motors used in industry. Most fans and pumps use some form of flow control in an attempt to match supply with demand. Traditionally, mechanical means have been used to restrict the flow, such as a damper on a fan or a throttle valve on a pump. Such methods waste energy by increasing the resistance to flow and by running the fan or pump away from its most efficient point. A much better method is to use a VSD to alter the speed of the motor. For centrifugal fans and pumps the power input is proportional to the cube of the speed, while the flow is proportional to the speed. Hence, a reduction to 80% of maximum speed (flow) will give a potential reduction in power consumption of 50%.

The application of PMs to electrical machines improves their efficiency by eliminating the excitation losses. The air gap magnetic flux density increases, which means greater output power for the same main dimensions.

A 3% increase in motor efficiency can save 2% of energy used. Most energy is consumed by three-phase induction motors rated at below 10 kW.

Consider a small three-phase, four-pole, 1.5-kW, 50-Hz cage induction motor.

The full load efficiency of such a motor is usually 78%. By replacing this motor with a rare-earth PM brushless motor the efficiency can be increased to 89%. This means that the three-phase PM brushless motor draws from the mains only 1685 W instead of 1923 W drawn by the three phase cage induction motor. The power saving is 238 W per motor. If in a country, say, one million such motors are installed, the reduction in power consumption will be 238 MW, or one quite large turbo-alternator can be disconnected from the power system. It also means a reduction in CO2 and NOx emitted into the atmosphere if the energy is generated by thermal power plants.

cont. to part 2 >>

Top of Page PREV.   NEXT Article Index HOME