Electronic Power for Trains
This page describes the most recent developments of electric train power equipment including the latest IGBT controlled 3-phase Alternating Current (AC) motors and the new permanent magnet motor.
AC and DC Differences - AC Locomotives with DC Drives - The Diode - The Thyristor - SEPEX - DC Choppers - Dynamic Braking - The GTO Thyristor - AC Motors - The Asynchronous Motor - AC Drive - IGBT - Permanent Magnet Motor
AC and DC Differences
To understand the principles of modern traction power control systems, it is worth a look at the basics of DC and AC circuitry. DC is direct current - it travels in one direction only along a conductor. AC is alternating current - so called because it changes direction, flowing first one way along the conductor, then the other. It does this very rapidly. The number of times it changes direction per second is called the frequency and is measured in Hertz (Hz). It used to be called cycles per second, in case you've read of this in historical papers. In a diagrammatic representation, the two types of current appear as shown in the diagram above left.
From a transmission point of view, AC is better than DC because it can be distributed at high voltages over a small size conductor wire, whereas DC needs a large, heavy wire or, on many DC railways, an extra rail. DC also needs more frequent feeder substations than AC - the ratio for a railway averages at about 8 to 1. It varies widely from one application to another but this gives a rough idea. See also Electric Traction Pages Power Supplies.
Over the hundred years or so since the introduction of electric traction on railways, the rule has generally been that AC is used for longer distances and main lines and DC for shorter, suburban or metro lines. DC gets up to 3000 volts, while AC uses 15,000 - 50,000 volts.
Until recently, DC motors have been the preferred type for railways because their characteristics were just right for the job. They were easy to control too. For this reason, even trains powered from AC supplies were usually equipped with DC motors.
AC Locomotives with DC Drives
This diagram (above) shows a simplified schematic for a 25 kV AC electric locomotive used in the UK from the late 1960s. The 25 kV AC is collected by the pantograph and passed to the transformer. The transformer is needed to step down the voltage to a level which can be managed by the traction motors. The level of current applied to the motors is controlled by a "tap changer", which switches in more sections of the transformer to increase the voltage passing through to the motors. It works in the same way as the resistance controllers used in DC traction, where the resistance contactors are controlled by a camshaft operating under the driver's commands.
Before being passed to the motors, the AC has to be changed to DC by passing it through a rectifier. For the last 30 years, rectifiers have used diodes and their derivatives, the continuing development of which has led to the present, state-of-the-art AC traction systems.
A diode is a device with no moving parts, known as a semi-conductor, which allows current to flow through it in one direction only. It will block any current which tries to flow in the opposite direction. Four diodes arranged in a bridge configuration, as shown below, use this property to convert AC into DC or to "rectify" it. It is called a "bridge rectifier". Diodes quickly became popular for railway applications because they represent a low maintenance option. They first appeared in the late 1960s when diode rectifiers were introduced on 25 kV AC electric locomotives.
The thyristor is a development of the diode. It acts like a diode in that it allows current to flow in only one direction but differs from the diode in that it will only permit the current to flow after it has been switched on or "gated". Once it has been gated and the current is flowing, the only way it can be turned off is to send current in the opposite direction. This cancels the original gating command. It's simple to achieve on an AC locomotive because the current switches its direction during each cycle. With this development, controllable rectifiers became possible and tap changers quickly became history. A thyristor controlled version of the 25 kV AC electric locomotive traction system looks like the diagram here on the left.
A tapping is taken off the transformer for each DC motor and each has its own controlling thyristors and diodes. The AC from the transformer is rectified to DC by chopping the cycles, so to speak, so that they appear in the raw as half cycles of AC as shown on the left.
In reality, a smoothing circuit is added to remove most of the "ripple" and provide a more constant power flow as shown in the diagram (left). Meanwhile, the power level for the motor is controlled by varying the point in each rectified cycle at which the thyristors are fired. The later in the cycle the thyristor is gated, the lower the current available to the motor. As the gating is advanced, so the amount of current increases until the thyristors are "on" for the full cycle. This form of control is known as "phase angle control".
In more recent thyristor control systems, the motors themselves are wired differently from the old standard DC arrangement. The armatures and fields are no longer wired in series, they are wired separately - separate excitement, or SEPEX. Each field has its own thyristor, which is used to control the individual fields more precisely.
Since the motors are separately excited, the acceleration sequence is carried out in two stages. In the first stage, the armature is fed current by its thyristors until it reaches the full voltage. This might give about 25% of the locomotive's full speed. In the second stage, the field thyristors are used to weaken the field current, forcing the motor to speed up to compensate. This technique is known as field weakening and was already used in pre-electronic applications.
A big advantage of SEPEX is that wheel slip can be detected and corrected quickly, instead of the traditional method of either letting the wheels spin until the driver noticed or using a wheel slip relay to switch off the circuit and then restart it.
The traditional resistance control of DC motors wastes current because it is drawn from the line (overhead or third rail) and only some is used to accelerate the train to 20-25 mph when, at last, full voltage is applied. The remainder is consumed in the resistances. Immediately thyristors were shown to work for AC traction, everyone began looking for a way to use them on DC systems. The problem was how to switch the thyristor off once it had been fired, in other words, how to get the reverse voltage to operate on an essentially one-way DC circuit. It is done by adding a "resonant circuit" using an inductor and a capacitor to force current to flow in the opposite direction to normal. This has the effect of switching off the thyristor, or "commutating" it. It is shown as part of the complete DC thyristor control circuit diagram (left). It has its own thyristor to switch it on when required.
Two other features of the DC thyristor circuit are the "freewheel diode" and the "line filter". The freewheel diode keeps current circulating through the motor while the thyristor is off, using the motor's own electro magnetic inductance. Without the diode circuit, the current build up for the motor would be slower.
Thyristor control can create a lot of electrical interference - with all that chopping, it's bound to. The "line filter" comprises a capacitor and an inductor and, as its name suggests, it is used to prevent interference from the train's power circuit getting into the supply system.
The thyristor in DC traction applications controls the current applied to the motor by chopping it into segments, small ones at the beginning of the acceleration process, gradually enlarging as speed increases. This chopping of the circuit gave rise to the nickname "chopper control". It is visually represented by the diagram below, where the "ON" time of the thyristor is regulated to control the average voltage in the motor circuit. If the "ON" time is increased, so does the average voltage and the motor speeds up. The system began to appear on UK EMUs during the 1980s.
Trains equipped with thyristor control can readily use dynamic braking, where the motors become generators and feed the resulting current into an on-board resistance (rheostatic braking) or back into the supply system (regenerative braking). The circuits are reconfigured, usually by a "motor/brake switch" operated by a command from the driver, to allow the thyristors to control the current flow as the motors slow down. An advantage of the thyristor control circuitry is its ability to choose either regenerative or rheostatic braking simply by automatically detecting the state of receptivity of the line. So, when the regenerated voltage across the supply connection filter circuit reaches a preset upper limit, a thyristor fires to divert the current to the on-board resistor.
The GTO Thyristor
By the late 1980s, the thyristor had been developed to a stage where it could be turned off by a control circuit as well as turned on by one. This was the "gate turn off" or GTO thyristor. This meant that the thyristor commutating circuit could be eliminated for DC fed power circuits, a saving on several electronic devices for each circuit. Now thyristors could be turned on and off virtually at will and now a single thyristor could be used to control a DC motor.
It is at this point that the conventional DC motor reached its ultimate state in the railway traction industry. Most systems now being built use AC motors.
There are two types of AC motor, synchronous and asynchronous. The synchronous motor has its field coils mounted on the drive shaft and the armature coils in the housing, the inverse of normal practice. The synchronous motor has been used in electric traction - the most well-known application being by the French in their TGV Atlantique train. This used a 25 kV AC supply, rectified to DC and then inverted back to AC for supply to the motor. It was designed before the GTO thyristor had been sufficiently developed for railway use and it used simple thyristors. The advantage for the synchronous motor in this application is that the motor produces the reverse voltages needed to turn off the thyristors. It was a good solution is its day but it was quickly overtaken by the second type of AC motor - the asynchronous motor - when GTO thyristors became available.
The Asynchronous Motor
The asynchronous motor, also called the induction motor, is an AC motor which comprises a rotor and a stator like the DC motor, but the AC motor does not need current to flow through the armature. The current flowing in the field coils forces the rotor to turn. However, it does have to have a three phase supply, i.e. one where AC has three conductors, each conducting at a point one third into the normal cycle period, as visually represented in the diagram on the left.
The two big advantages of the 3-phase design are that, one, the motor has no brushes, since there is no electrical connection between the armature and the fields and, two, the armature can be made of steel laminations, instead of the large number of windings required in other motors. These features make it more robust and cheaper to build than a commutator motor.
Modern electronics has given us the AC drive. It has only become available with modern electronics because the speed of a 3-phase AC motor is determined by the frequency of its supply but, at the same time, the power has to be varied. The frequency used to be difficult to control and that is why, until the advent of modern electronics, AC motors were almost exclusively used in constant speed applications and were therefore unsuitable for railway operation. A modern railway 3-phase traction motor is controlled by feeding in three AC currents which interact to cause the machine to turn. The three phases are most easily provided by an inverter which supplies the three variable voltage, variable frequency (VVVF) motor inputs. The variations of the voltage and frequency are controlled electronically.
The AC motor can be used by either an AC or DC traction supply system. In the case of AC supply (diagram left), the line voltage (say 25kV single phase) is fed into a transformer and a secondary winding is taken off for the rectifier which produces a DC output of say 1500 - 2000 volts depending on the application. This is then passed to the inverter which provides the controlled three phases to the traction motors. The connection between the rectifier and the inverter is called the DC link. This usually also supplies an output for the train's auxiliary circuits.
All the thyristors are GTOs, including those in the rectifier, since they are now used to provide a more efficient output than is possible with the older thyristors. In addition, all the facilities of DC motor control are available, including dynamic braking, but are provided more efficiently and with less moving parts. Applied to a DC traction supply, the 3-phase set-up is even more simple, since it doesn't need a transformer or a rectifier. The DC line voltage is applied to the inverter, which provides the 3-phase motor control.
Control of these systems is complex but it is all carried out by microprocessors. The control of the voltage pulses and the frequency has to be matched with the motor speed. The changes which occur during this process produce a set of characteristic buzzing noises which sound like the "gear changing" of a road vehicle and which can clearly be heard when riding on the motor car of an AC driven EMU.
Having got AC drive using GTO thyristors universally accepted (well, almost) as the modern traction system to have, power electronics engineers have produced a new development. This is the IGBT or Insulated Gate Bipolar Transistor. The transistor was the forerunner of modern electronics, (remember transistor radios?) and it could be turned on or off like a thyristor but it doesn't need the high currents of the thyristor turn off. However it was, until very recently, only capable of handling very small currents measured in thousanths of amps. Now, the modern device, in the form of the IGBT, can handle thousands of amps and it has appeared in traction applications. A lower current version was first used instead of thyristors in auxiliary supply inverters in the early 1990s but a higher rated version has now entered service in the most recent AC traction drives. Its principle benefit is that it can switch a lot faster (three to four times faster) than GTOs. This reduces the current required and therefore the heat generated, giving smaller and lighter units. The faster switching also reduces the complex "gearing" of GTOs and makes for a much smoother and more even sounding acceleration buzz from under the train. With IGBTs, "gear changing" has gone.
Permanent Magnet Motor
The next development in electric motor design is the permanent magnet motor. This is a 3-phase AC synchronous motor with the usual squirrel cage construction replaced by magnets fixed in the rotor. The motor requires a complex control system system but it can be up to 25% smaller than a conventional 3-phase motor for the same power rating. The design also gives lower operating temperatures so that rotor cooling isn't needed and the stator is a sealed unit with integral liquid cooling. By 2011, a number of different types of trains had been equipped with permanent magnet motors, including 25 AGV high speed train sets, trams in France and Prague and EMUs in Euroe and Japan. The reduced size is particularly attractive for low floor vehicles where hub motors can be an effective way of providing traction in a compact bogie. Development of motor design and the associated control systems continues and it is certain that the permanent magnet motor will be seen on more railways in the future. A good description of the motor by Stuart Hillmansen, Felix Schmid and Thomas Schmid is in Railway Gazette International, February 2011.