Leonard drives 75 YEARS OF DEVELOPMENT

by K. A. YEOMANS, B.Sc, M.Sc.(Eng.), D.I.C., C.Eng., M.I.E.E.

Although the Ward Leonard drive system is familiar to most engineers, its special features are not always appreciated,, When the system was ferst introduced in America 75 years ago, it was the subject of fierce criticism, and its early devel- opment for mine-winding and rolling-mill drives owes much to pioneering activities in Europe. Modern methods enable faster responses to be achieved, but no drive has shown itself more suitable than the Ward Leonard scheme for the purpose for which it was invented.

IT is NOW over 75 years since the firstpape r then president of the ALEE, F. Sprague, convertor, using the basic principle that was presented by H. Ward Leonard whose remarkable pioneering efforts had the maximum efficiency of the motor is which described the drive system that given the world its first commercially achieved when 'voltage varies as speed has made his name famous. Although successful electric municipal-transport and torque as current'. By this means, a his system is now well known, few appre- system2 at Richmond, Va., USA, in 1887. high torque could be achieved for ciate that it is almost as old as the history starting without an excessive power of applied electrical engineering, and that New system demand. Regeneration of power at slow- the idea was first opposed by unusually The new system of connection was down, or when descending a gradient, fierce criticism of its technical and com- proposed as an answer to the current was also claimed as a feature. mercial value. problems in applying electric motors to Ward Leonard first described his traction duties. The series motor had been Merits system in an article in Electrical World successfully developed by Sprague for The merits claimed for the system in November 1891, and this was followed this application as a robust and reliable included the increased efficiency, the by a paper1 presented on the 8th June machine, capable of withstanding the reduced space, the size and cost of control 1892 at the 9th general meeting of the abuses and severe duties encountered in switches and rheostats, and the savings American Institute of Electrical Engi- municipal street-car practice; but the in power-station plant made possible by neers. The occasion was a 3-day meeting heavy currents drawn in starting or the steadier load. It was admitted that the held at Chicago by the then well estab- in crawling heavily laden up steep cost of equipment on the car would be lished institute, whose activities had been gradients caused severe fluctuations in greater, but it was estimated that the total more usually confined to New York. At the power-station loadings. In addition, cost of the transport system would be this meeting, a wide range of papers on power losses in the rheostats used for less. The system had already been tried electric-circuit theory, on power-station starting resulted in apparently un- and found successful for electric lifts and management, on electrotechnical educa- avoidable inefficiencies. Many engineers cranes. tion and on mining and traction applica- had attempted to apply variable-ratio From the beginning, the proposals tions was discussed. At this time, electrical transmissions between motors and axles came under heavy criticism. Time was engineering was developing rapidly, and to achieve better efficiency at starting too short for a full discussion on the paper the d.c. electric motor was finding in- and at low speeds, but none had suc- at that meeting, but, from the chair, creasing use for a variety of machine ceeded. Sprague voiced his reservations on the drives, cranes and electric traction. Ward Leonard's arrangement (Fig. 1) merit of the scheme, and later he left the The latter development owed much to the was proposed as a variable-ratio power chair to comment on its practicability. 144 Electronics & Power April 1968 1 Ward Leonard's drive was proposed in 1892 as a variable-ratio power convertor for 'a new system of electric propulsion1

H W

While admitting it to be 'beautiful Massachusetts Institute of Technology uncertain of the reason for this, theoretically' and to 'accomplish every- at 22 years of age, and a year later he attributing it solely to the fact that the thing as desired from a physical stand- became associated with Thomas A. generator was separately excited. In fact, point', he did not consider it to be a Edison, as a member of a staff of four the usual difficulty in this respect arose practicable economic proposition for engineers selected to introduce the new from the practice of setting the brush axis street-car practice, for which he defended Edison power-station system. He was to commutate at full output with a the switched-field series motor that he had appointed superintendent of the Western highly saturated field. With this adjust- pioneered at Richmond. Electric Light Co. in Chicago at 26 ment, any alteration to main-field excita- years of age. tion to reduce the ouput resulted in Criticism The early experience of Ward Leonard neutral shift and sparking at heavy In New York about four months later, gave him an approach to drive problems currents. It was Ward Leonard's practice the discussion of the Ward Leonard paper that differed from that of his critics. He to set the generator brushes for no-load presented at Chicago was unusually was concerned much more with the neutral, probably to ensure good com- critical, the president (Sprague) again economics of the power-station operation mutation under regenerative conditions. vacating the chair to reiterate his earlier and distribution system than were the In discussion, Ward Leonard often adverse opinion. The opening speaker electric-railway engineers, whose main displayed a large degree of tol- claimed that Ward Leonard had been concentration of effort had been in the erance and patience, which was well unduly optimistic, and that the motor- design and operation of the street cars demonstrated at this meeting by a lengthy generator (m.g.) losses would be such themselves. duologue on commutation with one as to give only a marginal gain in efficiency The earlier electric railways were urban questioner who persisted in his —even at low-speed running—which was street-car systems with short trains of one doubts. Again, however, numerous objec- not worth the additional complication. or two cars, frequent stops and a compact tions were made to the new proposals, Comment was especially made on the distribution system. Electric railways, one member commenting that'. .. if Mr. heavy expense and maintenance charges .however, were being extended in range, Sprague, seven years ago, had objected of the m.g. set. and, in 1894, Ward Leonard presented to everything as the majority have done The criticism in the recorded account his paper—'How shall we operate an tonight, I do not think that we would be of the discussion3 on the paper would electric railway extending 100 miles from in the position with electric roads that we have discouraged most men, but Ward the power station?'5 It was proposed that are in today.' Leonard in his reply seemed still confi- power be transmitted at 20 kV a.c, with dent. His formal vocative to the meeting— variable-power convertors on the loco- New ideas 'Mr. chairman and gentlemen'—was motive to feed the separately excited d.c. It is likely that many of the objections perhaps unusual in that day, for it is traction motors. Unattended transformer were, in fact, raised by those who saw in hardly to be found recorded elsewhere in stations were used to feed the trolley the new scheme competition to their own the transactions of the ABEE. He denied wires at 500V a.c, and the convertor interests. There is little doubt that Ward that his efficiencies were optimistic, and m.g. set was to be driven by a single-phase Leonard himself had reached this opinion, revealed that they were based on practical a.c. motor. for in 1895, after a visit to France and tests. The increased costs of car equip- England,6 he made adverse comment on ment would also be lower than his critics Working model the relative lack of enterprise in trying had quoted, for it was intended to use At the meeting, the features of the new ideas in America and of the in- high-speed lower-cost machines for the Ward Leonard m.g. scheme were de- hibiting effect that the activities of large m.g. sets, the maintenance of which had monstrated by tests on a working model, companies had on engineering develop- been shown, in practice, to be negligible in which the ease of speed control and the ments. The AJDEE president, E. J. compared with that required for traction ability to regenerate and to reverse Houston, commented that '. . . If here motors which were, of necessity, exposed smoothly were shown. and there an inventor does not succeed to adverse working conditions. He again Discussion on the model, particularly in getting his particular system adopted made the point that his, and no other, on the absence of sparking at the gen- by a corporation, he may perhaps scheme would regenerate satisfactorily erator brushes, was lengthy. Most of the naturally feel that such corporations down to low speeds. questioners were puzzled by the excellence stand in the way of its introduction . . . H. Ward Leonard was born in 1861 in of generator commutation at low-voltage But they will be apt to buy anything of Cincinnati, USA.4 He graduated at the outputs, and even Ward Leonard seemed real commercial value. . . . There is not a Electronics & Power April 1968 145 monopoly of one corporation in America.' card, refused at first to believe that the the system on account of the expense The report of the visit to Europe press had been started. The benefit of due to the power loss of the idling m.g. included a description of the Heilmann Ward Leonard's scheme was readily set when the lift was not in use. locomotive built in France. This was not acknowledged, for there was not only a M.G. set losses are less significant when a complete 'new propulsion' system, for saving in power but also the convenience the duty required of the drive is high and the locomotive generator was driven by a of the control; the steady speed when starts are frequent. With heavy steam engine, but speed control was regulation at low speeds enabled a 25% drives used for mine winders and rolling obtained by varying the generator field, increase in production to be achieved. mills, the m.g. set losses are significantly and it must have been pleasing to report The early application of the system to less than the power that would otherwise that, in spite of the criticism of the ALEE, gun-turret control was notable. Turrets be lost by using rheostatic resistance the generator did not 'spark disastrously' had previously been slewed by steam starting. A crucial development came, and the commutator was 'in most cylinders, and both hydraulic systems and however, with the Ilgner flywheel - perfect condition'. series motors with rheostatic control had equaliser development, introduced in However, the Ward Leonard scheme been found to be unsuitable. The ability Germany for mine hoists8 in 1903. Mine was never widely adopted for use on of the new scheme was considered winders were frequently the main load locomotives. The reason perhaps lies remarkable, 23 individual stops and on a local generating plant or a long mainly in the success of the robust series starts being made in 20 s, with a turret feeder, and equaliser systems were essen- motor, which proved itself so satisfactory rotation of less than 1°. Later, in 1898, tial to minimise the effect of the in operation that its limitations in effi- the scheme, which was used on one fluctuating load of the winder on the ciency, in space requirements for rheo- turret of the USS Brooklyn, 'helped sink supply. Similar conditions also occur in stats, in jerkiness in acceleration and in the Spanish fleet' off Santiago de Cuba, reversing rolling-mill applications, and inability to recover power by regenera- and was thereafter, with few exceptions, the Ilgner scheme was first used for this tion became accepted. Also, in the 1890s, adopted by the US Navy and widely purpose in Austria in 1906, and then a.c. motors and systems were undergoing acclaimed in other countries. widely in Europe and in America. rapid development and were potentially The introduction of the Ward Leonard a more attractive alternative. It is likely, Speed control principle in Europe was expedited by its too, that Ward Leonard did not appre- In the discussion of the 1896 paper,7 spectacular success as a drive for the ciate the importance of the series-motor the emphasis was on controllability moving pavement at the Paris exposition load characteristic in balancing the rather than, as previously, on efficiency. of 1900, after all other methods had tractive effort between the axles of a Steinmetz, whose own activities at that failed (Fig. 3). multimotor system. time were concerned with the develop- ment of a.c. machines, pointed out the Improved regulation Applications clear advantage of the smooth start and In the second decade of the 20th In other fields the system had clear the 'perfect' speed control, even at low century, Ward Leonard's drive became advantages, and in a paper7 in 1896 speed; an advantage 'not. . . of economy increasingly accepted, but a difficulty Ward Leonard was able to report or efficiency nor of first cost . . . but of arose when attempts were made to numerous applications (Table 1). securing a method of gradual accelera- improve the speed-regulation perform- The drive had been first devised in tion'. It is not widely appreciated, even ance by compounding the generator. 1891 to drive a 30 hp calico printing press. to this day, that this has been one of the Although the reduced speed performance This type of printing machine had to be more valuable features of the scheme. of a motor fed from a variable-voltage started with compressed cloth between The performance of the Ward Leonard generator is better than that of a motor the rolls and needed to be readily con- drive is dominated by the electrical time with series resistance, the full-speed trolled to stop, start and run at low speed. constant of the generator shunt field and regulation is worse than that of the same All attempts to use electric motors had by the electromechanical time constant motor directly connected to the supply, previously failed, for the starting torque associated with the inertia at the motor for the resistance of the motor armature required was much higher than the shaft. If a voltage is suddenly applied to circuit includes that of the generator. running value, and sufficiently large the generator field and later reduced to Compound-wound motors can be currents could not be satisfactorily zero, the motor speed and acceleration arranged to improve regulation at the top supplied from the limited capacity of will have the form displayed in Fig. 2, speed, but the adjustment is not valid for factory generating plant. In addition, with a smoothly varying acceleration and varying speeds. The obvious alternative low-speed operation was irregular with retardation which cannot be easily of compounding the generator led, at series armature resistance owing to the obtained by rheostatic armature control first, to unexpected difficulties. Although erratic load fluctuations as the machine from a constant-potential supply. The it is possible to compound a generator operated. falling acceleration as top speed was for level-speed regulation in the steady Calico printing presses had previously approached may have been responsible, state, the dynamic performance will be been driven by twin-cylinder steam in part, for the unexpectedly good com- less damped and even completely unstable engines, and it was generally considered mutation of the generator that had been if the armature resistance is overcom- an impossibility to use an electric drive. noted earlier, for the current peak under pensated. The accepted procedure of However, on using a separately excited acceleration would have decayed before reducing speed by inserting series resist- generator to supply the individual press the generator field reached saturation. ance in the generator field circuit makes motor, a perfect drive had been achieved. the situation worse. The effect on the factory's generating Electric lifts Various methods of overcoming the plant was so small that the steam engineer, The inherent characteristics of the difficulty were proposed, and a system on the evidence of his steam-indicator Ward Leonard scheme were of great with a series booster in the main advantage for electric lifts, which were armature circuit with its field excited by Table 1 Applications of Ward Leonard then in a state of rapid development. systems Two companies had become interested in the system at an early stage and had Power used it successfully at lift speeds up to Application (hp) 250ft/min (a medium lift speed), with travelling cranes ~ '. ~ 1-50 regeneration during the slowdown period lifts 5-40 to control the speed almost down to a mine hoists 10-125 standstill. Smooth response of a lift gun turrets of warships 30 drive (particularly limited rate of change billet shifter in rolling mill. 30 acceleration—^ Heilmann locomotive . . 8x50 of acceleration) is an important factor for cloth-printing press . . 25 passenger comfort, but, although the newspaper press . . . 50 critics admitted that starting and stopping 2 Motor-speed and acceleration wave- boring machine .... 5 forms of a Ward Leonard drive in were 'extremely pretty', many condemned response to a stepped field voltage 146 Electronics & Power April 1968 3 Moving pavement at the Paris ex- position of 1900. The2mile closed circuit consisted of two parallel pavements driven at 3 and 6mile/h by180 5hp motors

that of any m.g. scheme, for their on-load efficiency is better, and there is practically no loss at all under no-load conditions. Being static devices, they require less maintenance, and there is no commutator to spark on the 'generator'. The greatest advantage, however, for heavy drives is that the equipment is lighter than rotating machines. Foundation and building costs the armature current proved acceptable it is difficult to accommodate varying are significantly less. for larger drives. The dynamic perform- saturation of the generator field circuit, However, although control is simple ance of this system is always overdamped, and it does not reduce the commonly and rheostatic power loss is avoided, the and there is an advantage that the observed, abrupt change due to brush- simple controlled-rectifier scheme does compounding effect is independent of neutral shift that occurs as the generator not meet the most important claims made any saturation of the main-generator field current passes through zero. An auxiliary by Ward Leonard for his drive. The system. One interesting development used series exciter has been used for com- phase-angle controlled-rectifier scheme tandem main and booster field systems pounding to compensate for this effect,12 adjusts the mean voltage applied to the on a common armature,9 but, in general, but further developments to meet an d.c. motor, but the rectifier current, and the expense of an additional heavy- increasingly stringent demand for higher therefore the supply current, is the current current booster machine was not ac- lift speeds, required in the skyscrapers supplied to the motor. High motor- ceptable for the smaller drives, such as being built in New York, resulted in the starting torques cannot be achieved with those used for electric lifts. development of a feedback-regulating reduced line currents. The rectifier drive In 1920, the compounding problem was exciter—the Rototrol13'14—used notably is thus only suited for use on supply largely overcome by the use of a damping for the 1200ft/min lifts at the Rockefeller systems of comparatively large capacity winding on the generator field. This Center in 1932. if significant voltage fluctuations are to development was first used for lifts10 to This system was followed by several be avoided, and therefore Ward Leonard's meet the requirement of consistent drive other control-generator schemes,15 in- demand for reduced power-station equip- performance, irrespective of load, which cluding the amplidyne, which have found ment is not met. was occasioned by an increasing demand wide use in a variety of regulated Ward Nevertheless, the increased speed of for automatic passenger operation. An Leonard systems. response and the reduced space require- alternative method of increasing the Although the inherent 'lags' of the ment and cost of the latest thyristor effective time constant of the field by Ward Leonard system are an advantage equipments make controlled rectifiers a using partial self excitation of the fieldan d when smoothness of response is desired, better choice for many applications where an alternative stabilising scheme using a they limit performance when fast changes the supply is corrected, or sufficiently 'damping' motor have also found wide use. of speed are needed. Such a requirement 'stiff', to tolerate the poor power factor Variable-voltage generator systems exists, for example, for reversing rolling and high harmonic disturbance. The were also developed during this period to mills; it is largely achieved by using increasing reliability of this equipment is give controlled torque rather than con- tandem motors to reduce the system an important factor, but it must be trolled speed. Electric excavator and oil- inertia and by laminating the generator acknowledged that, in practice, the m.g. well drives are frequently required to field to reduce main-field eddy-current sets used in Ward Leonard drives also accommodate stalled-torque conditions, circuits which add to the 'wound'-field have a good record in this respect and, and series-wound motors had previously time constant. Lamination of the interpole although on some large or unusual in- been used for these systems in conjunction field circuit to enable the commutating stallations the commutators of genera- with controlled generators. However, the field to follow the rapidly changing tors have caused concern, in general stalled (zero-speed) characteristic was not armature current was also found to be the vast majority of sets have given ideal, and it was necessary to provide necessary. Reversal times of a fraction completely satisfactory service, in many heavy-current contactor gear for re- over 1 s are common.16 cases with little more maintenance than versing the fields of the motors. On using that usually given to induction motors of differential series windings on the gen- Static devices the same power. 11 erator, and constantly excited shunt drive In the last 30 years, controlled rectifiers The essential feature introduced by motors, a much improved characteristic have been developed as an alternative to Ward Leonard was that of the electrical was achieved. the Ward Leonard system. They have power convertor. The drive motor is the advantage of overcoming most of the supplied at a variable potential, and the Higher performance points of technical criticism advanced power required is obtained by trans- Although compounding improved the against Ward Leonard's proposals. Their formation from the constant potential of speed regulation over a range of loads, efficiency, without doubt, is better than the supply source. Mechanical torque si Electronics & Power April 1968 147 an intermediate agency, and the variable- potential control of the second conversion in the generator is quite unlike the voltage- modulation control of the rectifier switch or valve device of the electronic-circuit approach.

Universal success Amendments Many have sought to develop alter- native variable-speed motor schemes. Few systems have had the universal success of to the IEE wiring the Ward Leonard scheme, and one may now wonder at the early fierce objections that hindered its initial introduction. regulations Perhaps its fundamental simplicity is such that many were loth to give credit to the inventor for a system that trained engineers must surely feel that they could well have invented for them- selves. The IEE Council has authorised the issue of thefirst set of amendments to the 14th H. Ward Leonard is credited with over edition of the IEE wiring regulations, which took effect as from the 1st March 1968. The amendments are occasioned mainly by the issue of new and revised British 100 inventions, including the double-arm standards, and the main changes they introduce concern classes of excess-current circuit breaker and lighting systems for protection, 250V mineral-insulated cables, cable materials, butyl-rubber-insulated trains and automobiles. He died suddenly cables and glass-fibre-insulated flexible cords. The opportunity has also been at an AIEE banquet in New York on the taken to bring up to date a number of other references to British standards and codes 18th February 1915, having earlier made of practice. his last contribution to an AIEE discus- The regulations with 1968 amendments have been approved by the UK Minister sion, in which he related the story of his of Power and the Secretary of State for Scotland for their purposes, including those first installation of an interlocking stage of the Electricity Supply Regulations, 1937, and the Buildings Standards (Scotland) dimmer in 1885 in Chicago. He was a Regulations, 1963-67. life member and fellow of the AIEE and Copies of the 1968 amendments, which need to be read in conjunction with the text of the 14th edition (1966), may be obtained from the Secretary of the IEE; price served as vice-president and manager. 3s. post free. He received the Gold Medal at the Paris exposition of 1900 and at St. Louis, USA, in 1904; he was awarded the John Scott Main changes introduced by the 1968 amendments to the 14th Medal at Philadelphia Franklin Institute edition of the IEE wiring regulations in 1903. It is said that he was charming Classes of excess-current protection rating. The decision to include this in manner and commanding in ap- The 14th edition introduced into the exemption was taken at too late a stage pearance, and history has shown the regulations for the first time the concept to allow current ratings for these cables great value of his innovations to engineer- of 'close' and 'coarse' excess-current to be included in the 14th edition, but ing progress. protection in relation to the current ratings were obtainable from the IEE rating of cables, operation of the pro- Secretary on application. With the issue tection within 4h at 1 • 5 times the designed of BS 3207: Part 1: 1968, the need for load current being adopted as the this exemption has disappeared, and the definition of 'close' protection. At the present amendments provide current References time of issue of the 14th edition, it was ratings for this class of cable. 1 LEONARD, H. w.: 'A new system of electric appreciated that, while this definition was propulsion', Trans. Amer. Inst. Elect. Engrs., accurately related to the excess-current Cable materials 1892,9, pp. 566-577 2 SPRAGUE, F. J. : 'The solution of the municipal performance of thermoplastics cables, it A partial revision of BS 2899, for rubber rapid transit problem', ibid., 1888, 5, did not accord well with the classifications insulation and sheath of electric cables, pp. 358-398 of fusing factor in BS 88: 1952—'Electric 3 LEONARD, H. w.: Discussion contributions, issued in 1967, has necessitated revision of ibid., 1892, 9, pp. 761-793 fuses'—some Class Q fuses which com- references in the regulations to various 4 LEONARD, H. w.: Obituary, Elect. World, plied with that standard being suitable to types of synthetic rubber compound. 1915, 65, p. 576 5 LEONARD, H. w.: 'How shall we operate an afford 'close' protection, and other Class electric railway extending 100 miles from Q fuses being suitable only for 'coarse' Butyl-rubber-insulated cables the power station?', Trans. Amer. Inst. Elect. The 14th edition included temporary Engrs., 1894, 11, pp. 78-107 protection. This arrangement was incon- 6 LEONARD, H. w.: 'Note on recent electrical venient to designers but was acceptable provisions for the use of butyl-rubber- engineering developments in France and as a temporary expedient, considering insulated cables, pending publication of England', ibid., 1895, 12, pp. 36-51 7 LEONARD, H. w.: 'Volts v. ohms—speed that BS 88: 1952 required fuselinks to be the corresponding British standard. This regulation of electric motors', ibid., 1896,13, marked with the fusing factor. A revision has now been issued as BS 4180: 1967, pp. 377-386 8 PATCHELL, w. H. : 'Application of electricity of BS 88 was in progress at the time, for enabling these temporary provisions to to mines and heavy industries' (Constable, which the classifications P and Qi as be dispensed with. 1913), p. 165 corresponding to the definition of 'close' 9 PHILLIPS, R. s.: 'Electric lifts' (Pitman, 1939), pp. 85-89 protection, and the classifications Q2 and Glass-fibre-insulated flexible cords 10 BOUTON, E. M. : 'Variable voltage control as R as corresponding to 'coarse' protection, BS 4217, issued in 1967, deals with glass- applied to electric elevators', Trans. Amer. Inst. Elect. Engrs., 1924, 43, pp. 199-210 were adopted. This revision has now been fibre-insulated flexiblecord s suitable for 11 MCNEAL, R. w.: 'Variable voltage equipment published as BS 88: Part 1:1967, and has use with lighting fittings and in other for electric power shovels', ibid., 1926, 45, enabled the wiring-regulations definitions pp. 1181-1184 applications where the cord is not subject 12 CALLAWAY, C. R., and HARRINGTON, E. D.: of classes of excess-current protection and to abrasion or undue flexing. The regula- 'A notable high speed elevator installation', the British standard classifications to be tions are now amended to provide for the Gen. Elect. Rev., 1926, 29, pp. 84-96 13 SANTINI, D. : 'Vertical transportation at better aligned. use of this type of cord, which may be Rockefeller center', Elect. J., 1934, 31, operated at temperatures up to 180°C pp. 246-249 14 FORMHALS, w. H.: 'Rototrol—a versatile 250 V mineral-insulated cables and which forms a useful addition to the electric regulator', Westinghouse Engr., May The 14th edition included an exemption range of types of 'high-temperature' cord, 1942, pp. 51-54 especially where operation at tem- 15 TUSTIN, A.: 'Direct current machines for admitting the use of 250 V mineral- control systems' (Spon, 1952) insulated copper-sheathed cables, as a peratures above the former limit (150°C 16 ABLETT, c. A.: 'Electrically driven reversing temporary measure pending an extension for silicone-rubber-insulated cords) is rolling mills', Trans. Instn. Engrs. Shipbld. Scotland, 1910, pp. 1-79 of BS 3207 to provide for that voltage desired. 148 Electronics & Power April 1968