Sources
Tramways and Electric Railways in the 19th Century, Cassier’s Electric Railway Number 1899.
Sprague, F.j. (1931), Electric traction in the Space of Three Dimensions, Journal of the Maryland Academy of Sciences.
Sprague, F.J. (1888), The Solution of Rapid Transit, the American Institute of Electrical Engineers.
Burch, E.P. (1911), Electric Traction for Railway Trains, McGraw Hill, New York.
Middleton, W.D. & Middleton, W.D. III, (2009), Frank Julian Sprague Electrical Inventor and Engineer, Indinana University Press.
Dalzell, F, (2010), Engineering Invention, MIT (ISBN: 9780262042567).
Introduction
Today, the electric motor is an essential part of railway technology both for diesel and electrically powered trains. Back in the 1870s, steam locomotives were the only form of traction on the railways and electricity was a novelty science but it eventually became mainstream, gradually being developed for lighting and motor power. Within 20 years, electrically driven tramcars and locomotives had been introduced for urban railways and were being trialled on main line railways.
First, it is worth remembering that electric motors for use on tramcars were developed by a number of different engineers, all of whom contributed ideas and tested them, largely independently. There was a great deal of rivalry in the early electric motor business because people realised the enormous potential of the technology and the enormous profits that it might generate.
The first electric motor that we would recognise today as a workable machine was developed by the Belgian engineer Zenobe Gramme. He discovered, by accident in 1873, that the dynamo that he'd invented produced electric current that another machine of the same design could convert back into rotation. When the armature of the dynamo was turned within a magnetic field to produce electric current and this was connected by a pair of wires to another dynamo, he discovered that the armature of the other dynamo was turning. He thus found that the mechanical design of a dynamo (or generator, as we would call it today) was the same as an electric motor. Other engineers soon took up the concept and improved it. Although the machines were crude, they were the first direct current (DC) electrical machines to be commercially successful and were gradually improved as experience with them was gained.
The Gramme machine had a ‘ring armature’ as shown in Figure 1. The disadvantage of the ring armature was that the arrangement of the armature coils tended to reduce the electorate-magnetic effect of the armature. This was solved by re-arranging the windings into the drum form adopted by Siemens (Figure 2).
Figure 1: A very simplified schematic of a Gramme electric motor with a ring armature. It shows the principal parts. The armature is built round the axle but insulated from it. The commutator (not shown but placed at one end of the armature) connects the armature wiring with the field by means of brushes. In this way the armature is connected to the field ‘in series’, giving us the ‘series wound motor’. Drawing adapted by author from ‘Renewable and Efficient Electric Power Systems’ by Gilbert M. Masters, IEEE Publishing, Wiley, 2013.
Figure 2: A schematic of a drum-wound electric motor. The main difference between this and the Gramme design was that the windings were on the outside of the armature, instead of being wrapped around a ring. Only one armature winding is shown here but there were actually lots of them. The drum principle proved to be a more efficient design and soon became the standard for most electric motors. Drawing adapted by the author from Milne, A.G., (1971), January. IEE Power Division: Chairman's address. Let there be light. In Proceedings of the Institution of Electrical Engineers (Vol. 118, No. 1, pp. 89-98). IET.
Sprague’s Design
The Gramme motor worked but it was not very efficient and an American engineer, Frank J. Sprague, was convinced that it could be improved. During late 1883 and early 1884, Sprague worked on developing a better version of the motor. The DC motor consists of a rotating part, known as the armature, and a static part known as the field. In the early designs, the field was normally connected in parallel with the armature circuit to create the magnetic field that would generate the turning of the armature. This was known as a shunt wound motor and Sprague’s early motors were designed like this. Later, he added a series field to make what we now call a compound wound motor. This worked better in how it controlled the speed of the motor.

Figure 3: Sprague No. 6 motor, showing the two-gear reduction drive and the arrangement of the horseshoe magnet wrapped round the armature. The two legs of the horseshoe carried the field windings. The motor frame is suspended between the axle and the spring on the bogie transom. Drawing: Cassier’s Magazine 1899, modified by the author.
The Wheelbarrow
The Sprague motor was a reasonable success. It was used for driving looms and other, similar, constant speed machinery. Once it started to sell, Sprague also used the design as the basis for his experimental electric traction motors for tramcars. During this development, he contributed another important principle for electric traction. He thought that the motor should be mounted under the vehicle as close to the wheels as possible. Previously, motors had usually been mounted inside the vehicle with a connection to the axle by a chain or belt. Sprague thought that the motor should be close to the axle as possible and should drive it through a pinion and gear arrangement.
Sprague's motor was mounted so that one end of it was supported by the axle while the other was carried by the transom of the truck (bogie) frame. Sprague referred to this as a ‘wheelbarrow’ design. Today, this is known as the nose suspended motor. The design has survived for over 100 years.
Improvements
One common problem for streetcar motors was contamination by dirt and water. The early motors were not enclosed, the designers assuming that they would be sufficiently protected by the car body. However, the fields and the ends of the armatures where the commutators were located were open to the elements and quickly became damaged by water, mud, snow or dust. In an effort to minimise the damage, canvas covers were tried at first but, in March 1891 in the US, Westinghouse, who entered the field of motor manufacture the year before after seeing the successes of other suppliers, produced the first electric traction motor that included most of the requirements that became the standard: series armature winding, machine wound coils and four field coils (Figure 4). Six months earlier, a company called Wenstrom had produced a motor where the armature windings were fitted in slots cut into the core rather than being wound round a drum, another feature to become standard.

Figure 4: A Westinghouse No. 3 motor of 1891 showing the various improvements compared with the Sprague No.6 motor of 1888. These included a hinged cover containing the field windings that enclosed the armature and provided protection as well as giving better performance. Also, the gear drive is now single and the pinion and gear are enclosed in their own oil-filled case. Most of the basics for the modern DC traction motor were now in place. Photo: Cassier’s magazine 1899.
Gears
It was quickly recognised that to provide effective torque on an electric railway car with a motor small enough to fit under the car, the drive connecting the motor to the axle needed to be geared. The ratios chosen were initially quite high; Sprague’s original two-stage gear ratio for the Richmond streetcars was 12:1. The early drives had two pinions and two gears but the system did not wear well. The gear teeth wore out very quickly and they were noisy. The average life of a streetcar motor gear set was about two months. Sometimes gears would seize, causing locked wheels and stalling the vehicle.
Some designers tried to overcome the problem by using gearless motors, where the armature was fitted directly round the axle but these motors were heavier and less efficient than the geared ones. The first gearless motor was designed by Edward Hopkinson for the City & South London Railway in 1890, following a suggestion originally made years before by William Siemens.
By the mid-1890s, refinements in motor and gear design had reached a stage where gears were reliable enough and the gear ratio was normally between 3 and 4 to 1. With this ratio, only one pinion/gear set was required.
Brushes
Although various engineers had built electric cars for use on street railways, none of them were really successful until Sprague equipped the first viable electric street tramway in Richmond, Virginia in 1888, using his compound wound motors, but there were considerable problems. The two most serious were first, that the motors were underpowered at 7.5 h.p. and second, the vulnerability of the brushes. Larger motors were fitted eventually but the brushes remained a problem.
The brushes were crucial to the operation of the motor. They connected the static field to the rotating armature. The problem was that, up to that time, brushes were made of copper or brass. Because they were flexible and were expected to operate in both directions, they wore out very quickly. Then, another engineer in the US, Charles van de Poele, put forward the idea of using carbon brushes in 1890; the problem was on the way to being solved and their use survives to this day. The carbon brush was refined by a Hopkinson patent, also of 1890, who proposed putting the brush in a tube and adding a spring to keep the pressure on the commutator constant.
The Series Motor
Most of the early traction motors were shunt wound - where the field coils were wired in parallel with the armature circuit - apart from Sprague’s early compound wound motors. The power to the motor was usually controlled by varying the field resistance. However, in 1891, Westinghouse produced a series wound motor, where the field was wired in series with the armature and where the whole motor circuit was controlled by a variable resistance that was inserted in series with the motor upon starting and then cut out in steps to increase speed. Again, this remained the standard method of motor control until the introduction of solid state thyristor control in the 1980s.
The End of the Beginning
By the early 1890s, the design of the DC traction motor was largely settled and it remained in widespread use throughout the 20th century. It underwent a few developments in manufacture and improvements in commutator and wiring design but an engineer from 1892 could look at the motor still used under many electric trains today and recognise the machine as almost the same as his.
However, with the introduction of solid state power electronics in the 1970s, the writing was on the wall for the DC motor and the long desired goal of being able to use alternating current (AC) induction motors, with 3-phase power control, was finally within sight. In the railway traction business now, the DC motor is only still seen on older trains.