Until the 1980s the conventional (brushed) d.c. machine was the automatic choice where speed or torque control is called for, and large numbers remain in service despite a declining market share that reflects the move to inverter-fed induction motors.
Applications range from steel rolling mills, railway traction, to a very wide range of industrial drives, robotics, printers, and precision servos. The range of power outputs is correspondingly wide, from several megawatts at the top end down to only a few watts, but except for a few of the small low-performance ones, such as those used in toys, all have the same basic structure, as shown in Figure 1.
The motor has two separate circuits. The smaller pair of terminals connect to the field windings, which surround each pole and are normally in series: in the steady state all the input power to the field windings is dissipated as heat – none of it is converted to mechanical output power. The main terminals convey the ‘power’ or ‘work’ current to the brushes which make sliding contact to the armature winding on the rotor.
DC Motor Basics
The supply to the field is separate from that for the armature, hence the description ‘separately excited’. As in any electrical machine it is possible to design a d.c. motor for any desired supply voltage, but for several reasons it is unusual to find rated voltages lower than about 6 V or much higher than 700 V.
The lower limit arises because the brushes give rise to an unavoidable volt-drop of perhaps 0.5–1 V, and it is clearly not good practice to let this ‘wasted’ voltage became a large fraction of the supply voltage. At the other end of the scale, it becomes prohibitively expensive to insulate the commutator segments to withstand higher voltages.
The function and operation of the commutator is discussed later, but it is appropriate to mention here that brushes and commutators are troublesome at very high speeds.
Small d.c. motors, say up to hundreds of watts output, can run at perhaps 12,000 rev/min, but the majority of medium and large motors are usually designed for speeds below 3000 rev/min.
Increasingly, motors are being supplied with power-electronic drives, which draw power from the a.c. mains and convert it to d.c. for the motor. Since the mains voltages tend to be standardised (e.g. 110 V, 220–240 V, or 380–440 V, 50 or 60 Hz), motors are made with rated voltages which match the range of d.c. outputs from the converter. As mentioned above, it is quite normal for a motor of a given power, speed and size to be available in a range of different voltages.
In principle, all that has to be done is to alter the number of turns and the size of wire making up the coils in the machine. A 12 V, 4 A motor, for example, could easily be made to operate from 24 V instead, by winding its coils with twice as many turns of wire having only half the cross-sectional area of the original.
The full speed would be the same at 24 V as the original was at 12 V, and the rated current would be 2 A, rather than 4 A. The input power and output power would be unchanged, and externally there would be no change in appearance, except that the terminals might be a bit smaller.
Traditionally d.c. motors were classified as shunt, series or separately excited. In addition it was common to see motors referred to as ‘compound-wound’. These descriptions date from the period before the advent of power electronics, and a strong association built up, linking one or other ‘type’ of d.c. machine with a particular application.
There is really no fundamental difference between shunt, series or separately excited machines, and the names simply reflect the way in which the field and armature circuits are interconnected. The terms still persist, however, and we will refer to them again later.
But first we must gain an understanding of how the basic machine operates, so that we are equipped to understand what the various historic terms mean, and hence see how modern practice is deployed to achieve the same ends.
We should make clear at this point that whereas in an a.c. machine the number of poles is of prime importance in determining the speed, the pole number in a d.c. machine is of little consequence as far as the user is concerned. It turns out to be more economical to use two or four poles in small or medium size d.c. motors, and more (e.g. ten or twelve or even more) in large ones, but the only difference to the user is that the 2-pole type will have two brushes at 180o, the 4-pole will have four brushes at 90o, and so on. Most of our discussion centres on the 2-pole version in the interests of simplicity, but there is no essential difference as far as operating characteristics are concerned.
Torque is produced by the interaction between the axial current-carrying conductors on the rotor and the radial magnetic flux produced by the stator. The flux or ‘excitation’ can be furnished by permanent magnets (Figure 2(a)) or by means of field windings (Figure 2(b)).
Permanent magnet versions are available in motors with outputs from a few watts up to a few kilowatts, while wound-field machines begin at about 100 W and extend to the largest (MW) outputs.
The advantages of the permanent magnet type are that no electrical supply is required for the field, and the overall size of the motor can be smaller. On the other hand, the strength of the field cannot be varied, so one possible option for control is ruled out.
Ferrite magnets have been used for many years. They are relatively cheap and easy to manufacture but their energy product (a measure of their effectiveness as a source of excitation) is poor.
Rare earth magnets (e.g. neodymium–iron–boron or samarium–cobalt) provide much higher energy products, and yield high torque/volume ratios: they are used in high-performance servo motors, but are relatively expensive and difficult to manufacture and handle.
Nd–Fe–B magnets have the highest energy product but can only be operated at temperatures below about 150oC, which is not sufficient for some high-performance motors.
Although the magnetic field is essential to the operation of the motor, but none of the mechanical output power actually comes from the field system. The excitation acts like a catalyst in a chemical reaction, making the energy conversion possible but not contributing to the output.
The main (power) circuit consists of a set of identical coils wound in slots on the rotor, and known as the armature. Current is fed into and out of the rotor via carbon ‘brushes’ which make sliding contact with the ‘commutator’, which consists of insulated copper segments mounted on a cylindrical former.
The function of the commutator is discussed below, but it is worth stressing here that all the electrical energy which is to be converted into mechanical output has to be fed into the motor through the brushes and commutator.
Given that a high-speed sliding electrical contact is involved, it is not surprising that to ensure trouble-free operation the commutator needs to be kept clean, and the brushes and their associated springs need to be regularly serviced.
Brushes wear away, of course, though if properly set they can last for thousands of hours. All being well, the brush debris (in the form of graphite particles) will be carried out of harm’s way by the ventilating air: any build up of dust on the insulation of the windings of a high-voltage motor risks the danger of short circuits, while debris on the commutator itself is dangerous and can lead to disastrous ‘flashover’ faults.
The axial length of the commutator depends on the current it has to handle. Small motors usually have one brush on each side of the commutator, so the commutator is quite short, but larger heavy-current motors may well have many brushes mounted on a common arm, each with its own brushbox (in which it is free to slide) and with all the brushes on one arm connected in parallel via their flexible copper leads or ‘pigtails’.
The length of the commutator can then be comparable with the ‘active’ length of the armature (i.e. the part carrying the conductors exposed to the radial flux).
Function of Commutator
Many different winding arrangements are used for d.c. armatures, and it is neither helpful nor necessary for us to delve into the nitty-gritty of winding and commutator design. These issues are best left to motor designers and repairers. What we need to do is to focus on what a well-designed commutator-winding actually achieves, and despite the apparent complexity, this can be stated quite simply.
The purpose of the commutator is to ensure that regardless of the position of the rotor, the pattern of current flow in the rotor is always as shown in Figure 3. Current enters the rotor via one brush, flows through all the rotor coils in the directions shown in Figure 3, and leaves via the other brush.
The first point of contact with the armature is via the commutator segment or segments on which the brush is pressing at the time (the brush is usually wider than a single segment), but since the interconnections between the individual coils are made at each commutator segment, the current actually passes through all the coils via all the commutator segments in its path through the armature.
We can see from Figure 3 that all the conductors lying under the N pole carry current in one direction, while all those under the S pole carry current in the opposite direction. All the conductors under the N pole will therefore experience a downward force (which is proportional to the radial flux density B and the armature current I) while all the conductors under the S pole will experience an equal upward force.
A torque is thus produced on the rotor, the magnitude of the torque being proportional to the product of the flux density and the current. In practice the flux density will not be completely uniform under the pole, so the force on some of the armature conductors will be greater than on others. However, it is straightforward to show that the total torque developed is given by
T = KT ɸI ……(equation 1)
Where ɸ is the total flux produced by the field, and KT is constant for a given motor.
In the majority of motors the flux remains constant, so we see that the motor torque is directly proportional to the armature current. This extremely simple result means that if a motor is required to produce constant torque at all speeds, we simply have to keep the armature current constant. Standard drive packages usually include provision for doing this, as will be seen later.
We can also see from equation (1) that the direction of the torque can be reversed by reversing either the armature current (I) or the flux (ɸ). We obviously make use of this when we want the motor to run in reverse, and sometimes when we want regenerative braking.
The alert reader might rightly challenge the claim made above – that the torque will be constant regardless of rotor position. Looking at Figure 3, it should be clear that if the rotor turned just a few degrees, one of the five conductors shown as being under the pole will move out into the region where there is no radial flux, before the next one moves under the pole.
Instead of five conductors producing force, there will then be only four, so won’t the torque be reduced accordingly? The answer to this question is yes, and it is to limit this unwelcome variation of torque that most motors have many more coils than shown in Figure 3.
Smooth torque is of course desirable in most applications in order to avoid vibrations and resonances in the transmission and load, and is essential in machine tool drives where the quality of finish can be marred by uneven cutting if the torque and speed are not steady.
Broadly speaking, the higher the number of coils (and commutator segments) the better, because the ideal armature would be one in which the pattern of current on the rotor corresponded to a ‘current sheet’, rather than a series of discrete packets of current. If the number of coils was infinite, the rotor would look identical at every position, and the torque would therefore be absolutely smooth.
Obviously this is not practicable, but it is closely approximated in most d.c. motors. For practical and economic reasons the number of slots is higher in large motors, which may well have a hundred or more coils, and hence very little ripple in their output torque.
Operation of Commutator – Interpoles
Returning now to the operation of the commutator, and focusing on a particular coil (e.g. the one shown as ab in Figure 3) we note that for half a revolution – while side a is under the N pole and side b is under the S pole, the current needs to be positive in side a and negative in side b in order to produce a positive torque.
For the other half revolution, while side a is under the S pole and side b is under the N pole, the current must flow in the opposite direction through the coil for it to continue to produce positive torque. This reversal of current takes place in each coil as it passes through the interpolar axis, the coil being ‘switchedround’ by the action of the commutator sliding under the brush.
Each time a coil reaches this position it is said to be undergoing commutation, and the relevant coil in Figure 3 has therefore been shown as having no current to indicate that its current is in the process of changing from positive to negative.
The essence of the current-reversal mechanism is revealed by the simplified sketch shown in Figure 4. This diagram shows a single coil fed via the commutator, and brushes with current that always flows in at the top brush.
In the left-hand sketch, coil-side a is under the N pole and carries positive current because it is connected to the shaded commutator segment which in turn is fed from the top brush. Side a is therefore exposed to a flux density directed from left (N) to right (S) in the sketch, and will therefore experience a downward force. This force will remain constant while the coil-side remains under the N pole.
Conversely, side b has negative current but it also lies in a flux density directed from right to left, so it experiences an upward force. There is thus an anti-clockwise torque on the rotor.
When the rotor turns to the position shown in the sketch on the right, the current in both sides is reversed, because side b is now fed with positive current via the unshaded commutator segment. The direction of force on each coil side is reversed, which is exactly what we want in order for the torque to remain clockwise.
Apart from the short period when the coil is outside the influence of the flux, and undergoing commutation (current reversal), the torque is constant. It should be stressed that the discussion above is intended to illustrate the principle involved, and the sketch should not be taken too literally.
In a real multi-coil armature, the commutator arc is much smaller than that shown in Figure 4 and only one of the many coils is reversed at a time, so the torque remains very nearly constant regardless of the position of the rotor.
The main difficulty in achieving good commutation arises because of the self inductance of the armature coils and the associated stored energy. As we have seen earlier, inductive circuits tend to resist change in current, and if the current reversal has not been fully completed by the time the brush slides off the commutator segment in question, there will be a spark at the trailing edge of the brush.
In small motors some sparking is considered tolerable, but in medium and large wound-field motors small additional stator poles known as interpoles (or compoles) are provided to improve commutation and hence minimise sparking. These extra poles are located midway between the main field poles, as shown in Figure 5.
Interpoles are not normally required in permanent magnet motors because the absence of stator iron close to the rotor coils results in much lower armature coil inductance. The purpose of the interpoles is to induce a motional e.m.f. in the coil undergoing commutation, in such a direction as to speed-up the desired reversal of current, and thereby prevent sparking. The e.m.f. needed is proportional to the current (armature current) which has to be commutated, and to the speed of rotation.
The correct e.m.f. is therefore achieved by passing the armature current through the coils on the interpoles, thereby making the flux from the interpoles proportional to the armature current.
The interpole coils therefore consist of a few turns of thick conductor, connected permanently in series with the armature.