Steve's Railroad Pages

Information on Diesel-Electric
and Electric Locomotives

go to electric
information page

What a Modern Locomotive Is -- The Short Version

This is the really simple version. Modern locomotives have electric motors connected to the drive axles. The electric motors receive electric power either from an on-board power source (e.g., a diesel motor) or from a central power source via a distribution system (e.g., a thrid rail). The link between the electric motor and the source of the electricity is called the transmission. The electrical power lines that criss-cross our towns and cities are called electric "transmission" lines; the link from a diesel motor to the electric motors on a locomotive's axles is called the transmission. That's it!

Why it looks the way it does

Why do modern freight locomotives look the way they do -- a cab at one end, lots of bulky equipment at the other? Why do Amtrak, LIRR, and other passenger locomotives that have been recently designed without regard to any freight predecessors have one cab with, at most, a hostler's position at the other end?

The issue of cabs on locomotives has a number of "histories" that have converged to produce the style seen today. First, many early U.S. diesel locomotives did have two cabs, such as Baldwins built for Jersey Central (while the same locomotives for other roads had only one cab), as did other diesel and electrics such as various boxcabs and the GG1, and of course the AEM7s of today have cabs at each end. What gnaws at ones mind, though, is really about the big freight locomotives (like the SD80MACs of Conrail). So . . .

When the big frieght roads first dieselized, there were questions about MUs and crews. The railroads did not want to put a crew in each cab of an MUed set, so that brought forth such oxymorons as referring to the evolutionarilly critical A-B-B-A FTs from General Motors as "a locomotive." Calling it one locomotive (one with a cab at each end!) meant it needed only one crew (and note that the B-units had no cab, or if any just a hostler's position). As these evolved into F3s, F7s and F9s, and A-units unpaired and mixed with other units, the ubiquity of single-cab units was assured. Roads taking E-unit derivatives (i.e., double-engine units, even those from other builders) such as the Jersey Central Baldwins (DR-6-4-20) did sometimes take two cabs when it was clear that the unit would only be operating as a single unit; alternatively, some cab units came semi-permanently coupled back-to-back with a draw-bar instead of a coupler (e.g., PRR's DR-12-8-1500/2), essentially a single unit with a cab at each end (today, similar issues are resovled in passenger operations through push-pull, with a cab at the other end of the train).

The other key development was of the non-hood units, the "road switchers." General Motors again had the critical development with the GP-7. The unit was logically divided into two hoods, with a somewhat centered cab for ease of bi-directional movement. Under one hood -- the longer one -- was the motive equipment, while under the other -- the shorter one -- was usually a steam generator for passenger service. Local builder ALCO had similar locomotives with the RS-1s, RS-2s, and RS-3s, leading ultimately into its Century line. Passenger operations into the 1970s (e.g., the Long Island) took road switchers long-hood forward to protect the engine crew from the steam generator in case of an accident, particularly a grade crossing accident where a vehicle could have flipped over the anti-climber (where the anti-climber would have been had the units been so equipped) and crushed the steam generator into the cab (shades of an exploding steam locomotive!). Much earlier, many roads had begun turning their road switchers around and running them short hood forward for the better visability (who needed a long hood like a steam engines' boiler, anyway?) -- except for The Norfolk and Western [N&W] and The Southern [SOU], which seemed to delight in taking the most incrediably long hood and designating it front, with the cab so configured; and the Erie Lackawanna, which apparently had some SD-45s configured for either direction. Long-hood forward and bi-directional units sometimes had two control stands, one on each side of the cab facing "forward," but sometimes had just a single stand sitting parallel to the cab side for use with the engineer facing in either direction. As the time came when passenger equipment no longer needed steam, the short hood was cut and shortened, and that gave the configuration of the road switcher of today (note that during the overlap when short hoods were being cut but some passenger steam-generators were still needed in operation, GM's SDP-35, SDP-40, and SDP-45 had the steam generators at the coupler-end of the long hood).

At this point, a second cab could have been added, but only a hand-full of units have ever been built that way (some electrics on mining roads come to mind). For the most part, the big roads run with two or more units anyway, so having east-west pairs is not difficult (turning facilities such as wyes and loop tracks seem to be plentiful). The smaller roads simply don't want the expense. And, of course, it seems, N&W and SOU successor Norfolk Southern would just as soon run them long-hood anyway!

Baisc DC Motor Concepts

The traditional electric motor on a diesel-electric or electric locomotive is a DC motor. Internally, DC motors have two main components: the stator is the stationary outside part of a motor. The armature is the inner part which rotates. To get the armature to rotate, electric motors require two sets of windings, the field winding (on the stator) to develop the magnetic field within which the armature will turn, and the armature winding (which in many DC motors can make transition between series and shunt winding). In series, current passes through both the field and armature windings "in series", that is, one after the other, while in shunt the current is divided and passes either through the field or armature windings. DC motors with series windings develop high starting torque, while motors in shunt develop high speed. (A third category, the compound, combines both series and shunt windings simultaneously, with mixed properties: c.f. a motor that can transition between series and shunt, but not use both at the same time; a fourth category is the permanent-magnet motor, which not surprisingly uses a permanent magnet for the field, and is only used in relatively low-power applications.)

DC motors turn because an electrical field rotates. The field rotates because an electrical current passing into the armature changes polarity, with the armature tugged forward with each change. (N.B., it is the field in the armature that is changing, not the field in the stator.) In order to accomplish this change in polarity, the windings in the armature are connected to the outside world by means of a commutator, a conductive sheaf that allows for the current in the windings to have a change in polarity by breaking and making the connection. The electricity flows into the commutator through conductive brushes (usually carbon). These are sources of friction, heat, and general wear in the DC motor.

Series-wound motors are also called universal motors (see below), universal in the sense that they will run equally well using either AC or DC: simultaneously reversing the polarity of both the stator and the rotor cancel out, thus the motor will always rotate in the same direction regardless of the voltage polarity. Sometimes called "AC motors" instead of "DC series-wound motors" or "universal motors," these are not the motors to which one refers when referring to AC traction motors.

Note: In the motor wiring diagrams, the DC motors do not have separately excited windings. While some first generation diesel-electrics had an auxiliary generator to provide current to separately excited windings, which required manual switching by the engineer, that design is now obsolete. Contemporary U.S. road locomotives do not have separately excited windings; however, there has been other work in this direction, based in computer-control systems, that has included wheel-slip detection and control and wheel-creep systems, in which the windings are under separate control (see below).


When speaking of a vehicle such as a railroad locomotive, transmission is the process by which power is transmitted from one location and used in another. Often this implies some type of changing process, for example the manner in the rotational force of a crankshaft is converted to electrical force in a generator.

In a modern diesel-electric locomotive, this is a multi-staged process that goes from fuel oil to turning wheels. In the days of steam the intermediate process was, well, it was steam. Coal or oil was burned in a firebox. The heat generated heated water, turning it into steam (and continued to heat the steam), which in turn drove a piston in a back-and-forth motion, which -- through the drive gear -- turned the wheels. In diesel-electric locomotives, the fuel is burned in cylinders, driving pistons in a back-and-forth motion, which -- through the crank shaft -- turns an electricity generating device (a generator or alternator or both), which provides electricity for electric motors that are connected to the axles of the trucks, which turn the wheels.

Figure One: Block Diagram of a Diesel-Electric Locomotive
diesel motor
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
electric motors


Transition is the process by which the transmission of a diesel-electric locomoitve is brought from series wiring to parallel wiring. When in series, all current in the locomotive pass through all motors: this produces maximum low-speed force in the motors, i.e., maximum starting torque. When in parallel, current is divided among the motors: this produces maximum high-end efficency, i.e., highest motor speed. This is just as with the wiring internal to DC motors, where having the motor wound in series develops high starting torque, while placing the motor in parallel will develop high speed. Electrically, as current increases through the motors in a circuit with a given total current and voltage, the voltage drop across each motor will decrease: parallel circuits apply the total voltage to each load (i.e., in this case, motor), while series circuits apply the total current to each load.

Not all locomotives can make transition -- yard locomotives are often wired only for series. The motors on a diesel-electric road locomotives are often capable of making multiple transitions, with both trucks and motors on a truck capable of being switched into series or parallel wiring.

See the motor diagram for an example of a DC motor capable of transition and the transition diagram for examples of two- and three-axle transition schemes. (A note on how the construction of the motors is depicted is above). These diagrams are quite general: some locomotives would put like axles on different trucks in series (first axle with first axle, second with second), or have other arrangements.

Electric Locomotives

In straight electric locomotives or M.U. cars (multiple-unit electric), the on-board diesel engine and generator and/or alternator is replaced by Niagara Falls or Indian Point or some other such central power generating source. The power is transmitted to the railroad and delivered to the trains. Electric transmission lines are generally high-voltage AC. This is because of the greater efficiency (i.e., less loss) of AC during transmission over certain distances, and the likewise greater efficiency of high-voltage transmission over low-voltage transmission (see below and also the Formulas and Concepts page under Electric Power Transmission for more extensive notes on AC, DC, and long distance electric transmission). In rail applications (as in most others), the most efficient transmission of electricity from generating stations to the tracks requires transmission lines of high-voltage AC with substations to convert this to line voltage for equipment. The power is delivered to the individual locomotives or cars generally either by a track level third rail (there are some fourth rail systems, too) or through overhead wires known as catenary). While there are some exceptions, AC systems usually use catenary, while DC systems usually use third rail.

Figure Two: Block Diagram of a Electric Locomotive
central generating facility
down arrow indicating logical direction of flow
distribution system
down arrow indicating logical direction of flow
on-board electrical equipment
(e.g., transformer, rectifier, motor-generator, inverter, etc.)
down arrow indicating logical direction of flow
electric motors

Notes on Electric Power Transmission and Distribution Systems

Long-Distance Transmission, Substations, and Local Distribution

Two principal elements of an electric system are the transmission of the electric power from the source of the electricity (i.e., a generating plant) to the local use area and, ultimately, distributing that power to the consumer (e.g., an electric locomotive or a home). Thus, these parts of the system may be divided into the transmission system and the distribution system, with transmission convenying the power long distances (at high voltages), and the distribution system delivering the power locally (at low voltages).

Long-Distance Transmission

As a general rule, in the United States, DC cannot be transmitted as economically as AC in transmission systems; railroads follow the practice of AC transmission systems, with high voltage AC stepped-down to distribution voltages at substations.

Extensive notes on long distance electric transmission, including some formulas, are on the Formulas and Concepts page under Electric Power Transmission.

Local Distribution

Local distribution is almost universally accomplished by third rail or overhead wires. The highest third rail voltage in use in the U.S. today is the 1,000 volt system on San Francisco's BART system. The highest, historically, is reputed to have been an interurban that ran 2400 volts (historically, only for a brief period): it did not really work very well, as arcing and leakage were such critical issues that the system was conveted to a lower voltage.

As for AC, it is generally delivered at higher volatages than third rail through overhead wires. As for AC third rail, no such systems exist (at least that I know of, certainly not in the U.S.). This may not be practicable: AC's advantage comes from high voltage transmissions that can be readilly stepped up or down, with conversion to DC in the distribution system providing easy control of motors without including on-board rectifiers (but more on that following). (Now, with AC traction in common usage, there may be a rationale for that to change, but distribution is still basically low-voltage DC or higher voltage AC.)

Distribution on the North East Corridor

The NEC uses three different combinations:

  • D.C. to New York City: 11kV 25Hz
  • N.Y.C. to New Haven: 11kV 60Hz
  • New Haven to Boston: 25kV 60Hz

Understandably, Amtrak's engineers (the slide-rule type) want everything at 25kV 60Hz, and that is the standard for new track, such as the latest electrificationm, from New Haven to Boston. Sixty hertz has the advantage of being compatible with the commercial grid (25Hz requires frequency converters, which run [reportedly] $40,000,000 each), and 25kV is not an unusual voltage, so equipment is available.

Higher voltages are also more efficient to transmit than lower: in fact, when the voltage is doubled the amperage is halved for the same power level ( P = I * V: increasing voltage results in a linear decrease in current at the same power level). Since transmission losses are a function of amperage only (dissipated power = I2R, where R is the constant line resistance [or impedance, in the case of AC]), 25kV power can be transmitted over twice as far as 11kV power at the same loss levels. One should note, however, that this does not solve the problem of drawing down the current in a section by multiple trains running within it: high train density will still require short segments (remember that this concerns only local distribution via the catenary and its local feeders, not long-distance transmission from powerplants).

Unfortunately, upgrading track from 11kV to 25kV is expensive, because it is necessary to rebuild catenary to accomodate the higher arc distance at the higher voltage, (e.g., better insulation and insulators).


While it is theoretically possible to run an entire rail line, hundreds of miles long, on a single, synchronized AC phase, in practice it is not practical. Instead, lines are generally broken into ten to twenty mile segments, each one running on a different phase (usually 120 degrees apart). The boundry of each segment is called a "phase break." On former Pennsy track, these were marked by a phase-break signal, which looked like a typical position-light signal with all positions in the entire circle lit. To prevent arcing between the sections, an insulating section of catenary is run across the phase break. By segmenting the catenary into sections, it is a simple matter -- with respect to stringing the catenary -- to have not only different phases but also different frequencies, 25 or 60 hertz, or voltatges, 11kV or 25kV, on the two sides of a phase break.

Unlike former equipment, which had to stop and change internal settings, Acela trains are capable of changing both frequency and voltage while at speed. This is done as part of the normal "approaching phase break" message that is sent to trains via ACSES, which includes the frequency and voltage that will be on the other side of the break (typically no change). When the pantograph hits the insulator between phases, the train temporarily cuts the input power, reconfigures the leads to the windings on the primary of the main transformer, and reconnects the input power -- all in less than a second. To the equipment downstream from the transformer, all that is visible is an AC voltage that drops out briefly every five to 15 minutes, as the locomotive hits the various phases. The blip is short enough there should be no noticeable traction-motor stutter or hotel-power disruption.

Figure Three: Regional Examples of Mainline Electric Operations
  • Northeast Corridor
    • 25kv, 60Hz AC via Catenary
      The PRR originally built this as 11kv, 25hz and supplied its own electricity, as this system was not compatible with commercial 60Hz systems (complete change-over not yet completed: see above).
  • Long Island Rail Road
    • 600v DC via 3rd rail
  • Delaware, Lackawanna, and Western (metro New Jersey lines)
    • 3000v DC via Catenary
  • Montreal Suburban
    • 2500v DC via Catenary
  • Other Systems
    • there are also some 50kv, 60hz catenary systems
    • some DC operators generated their own 25Hz AC for distribution to substations


Substations sit between transmission and distribution systems. They are fairly straightforward: a transformer steps down the AC voltage, then, if using a DC distribution system, a rectifier converts the AC to DC. With their transformers to step-down the high voltage transmission voltages to distribution levels, they are located periodically throughout the system. For the railroad, these are much like the ubiquitious substations with their tranformers located throughout suburban neighborhoods.

In electric locomotive (or MU) applications, the use of DC (as on the Metro North [ex New York Central] Hudson and Harlem lines and on the LIRR) requires closely spaced substations to convert AC to DC, stations more closely spaced than might be required for a similar AC distribution system. This is because of the higher line losses at the lower distribution voltages (this is explained elsewhere under Electric Power Transmission).


Because of long-established AC motor issues of low starting torque and of power control (more on that follows), the traction motors themselves have (up until recently) been DC. This has meant that at some point in the process AC has had to be coverted ("rectified") to DC, either at the substation or in the locomotive.

AC, which reverses direction 60 times a second (the U.S. standard), generally resembles a sine wave in the distribution systems. A simple rectifier is an electrical check valve: flow is only permitted in one direction, while retaining the characteristic sine curve (one-half of the curve, just the "positive" half, let's say). This is referred to as half-wave rectification. A more sophisticated approach is to allow the negative alternations to pass also, but in the same direction as the positive alternations (i.e., no direction change in the current). The AC thus becomes a pulsating DC, with all pulsations of the sine wave in one direction from zero. This is referred to as full-wave rectification.

Figure Four: Rectification
graphic of two alternations of AC, half-wave rectification, and full-wave rectification

Substation-Based Rectification:

The rectifiers in modern substations are solid state, sillicon diode based. They are efficient (the voltage drop is a fraction of a volt through the rectifier) and reliable. Earlier systems, such as used on the Long Island Rail Road, used mercury arc rectifiers, only slightly less efficient, but requiring much support (these were known as "ingitron" systems). They were quite large, housed in large structures, and required much cooling. Other systems, such as used by the New York City Transit Authority, used rotary converter based substations - - very large, very maintenance intensive.

Substations are almost always fed with three phase AC, and the three phases overlap coming out of the rectifier, so the DC pulsates only slightly (filtering can remove the pulsations altogether: see figure seven below for an example). From here, the DC is fed to the third rail (or catenary) by way of breakers, current sensors, switchgear, and whatever else.

Locomotive-Based Rectification:

Prior to the advent of a solid-state technology for converting high-power AC to DC, massive locomotives were often the only solution to this issue when using all AC systems: an AC motor turned a DC generator, which in turn supplied DC to the motors on the axles. A rectifier that takes the form of an AC motor turning a generator is called, not surprisingly, a "motor-generator."

Some railroads, such as the Pennsylvania and New Haven, had MUs with ignitrons on them. Both of these roads also had ingitron based locomotives (known as "rectifiers," locomotives such as the EP-5, E-44, and E-33, but not the PRR's GG1: one of the most massive locomotives, it was actually an all-AC unit, with power control through tranformer taps).

In the mid-1960s, high-power solid-state rectifiers became feasible, and smaller, lighter weight electric locomotives -- and AC transmissions on diesel-electric locomotives -- became available. In the straight-electric market, the last GM (GMD [Canada]) motor-generator unit was the SW1200MG (2300v, 60hz), produced from 1963 to 1971 (1971 being well after the move to solid-state rectification, but the unit had gone into production in 1963 and was maintained for an existing customer).

The conversion to AC/DC transmission

An alternator is generally smaller and simpler than a generator of like capacity. This is because generators, like DC motors, are equipped with a commutator and carbon brushes, which are what reverses the electrical current as the armature turns, preventing the current from alternating, keeping the current direct. This simpler, lighter structure means that a diesel-electric locomotive using an alternator instead of a generator should be more economical. With the advent of economical and compact solid-state rectifiers, which could be routinely installed on locomotives (see above), the greater efficiency of the alternator could finally be realized in rail applications, and AC/DC transmission became a reality. In the mid-1960s, all three major manufacturers begin offing AC/DC transmission units, Alco and GM in 1965 and GE in 1966.

Figure Five: Block Diagram of a Diesel-Electric Locomotive with an AC/DC transmission
diesel motor
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
electric motors

The block diagram in figure five illustrates the AC/DC transmission. A diesel motor turns an alternator; the AC produced by the alternator is rectified to DC for the locomitive's DC traction motors.

GM's first applications of AC transmission were in the GP40 and SD40 of 1965 and 1966, respectively. General Electric's first AC transmission were the U28B/U28C offerings of 1966 (earlier production of these models was straight DC). Alco offered the top-end of its Century line with AC/DC transmissions, the C430/C630, in 1966/1965 respectively.

The move away from DC traction motors

One of the most important advances in locomotive technology in recent years is the AC traction motors. AC motors have been around for many years (the kitchen clock that plugs into a recessed electrical socket directly behind it is an example of one). However, AC motors were never able to match the starting torque of the DC and are notoriously difficult to control in varrying load and speed implementations. Unfortunately, while DC motors provide high starting torque they also have critical limitations (as was noted above). These limitations have long made it desirable that a substitute be found.

Like generators, DC motors are equipped with a commutator and carbon brushes, which are subjected to very high current loads. (In a generator, these are what reverses the electrical current as the armature turns, preventing the current from alternating -- keeping the current direct; in the motor, the commutator and brushes reverse the current, creating the moving magnetic force that rotates the armature.) A DC motor that would have high current loads while not in motion or while moving slowly would receive major damage or burn-out if such a high current were to be applied for too long a period of time. At low speeds, the high amperage damage would occur within minutes. Because of this, until recently, all DC locomotives all have minimum continuous speeds (for example, SD40 & 45 at 11 to 12 MPH, SD50 at 10 MPH, GP40 at 12 MPH, some swithcers and regeared road units, such as some CSX GP38s at 7 MPH).

Power in a DC circuit is simply equal to the voltage times the current. This is expressed as power (in watts, "P") = voltage (in volts, "E" [for Electromotive force) times current (in amps, "I"), or

P=E*I, and

Power (in horsepower) = watts * 0.00134102

(going the other way, watts = horsepower * 745.6999)

In DC motors, the power relationship is simple: at a constant voltage Ohms law requires more current to produce more power (watts = voltage * current). This means that in DC high current levels will be needed to produce high power, lacking a good way to vary voltage on the fly. This becomes expensive, having necessitated heavy conductors throughout the system to carry the high current; further, the high current produce a great deal of heat, further limiting DC traction motors.

For example, to compute the current flow in a 1000 horsepower switcher with DC traction motors at 600 volts,

  • 1000 horsepower = 745,699.872 watts
  • 745,699.872 watts / 600 volts = 1242.83312 amps.

Using today's high-horsepower DC units, e.g. a 4400 horsepower, 6 axle unit, where each of the six motors contribute 733 horsepower to the total unit horsepower, one can get up to 5500 amps per motor in a 600 volts system. (Remember that in DC motors that current goes across the commutator and brushes.) Specifically,

  • 4400 horsepower / 6 motors = 733 horsepower/motor
  • 733 horsepower/motor = 546,598 watts/motor

When operating in parallel, with a 600 volt drop across each motor,

  • 546,598 watts/motor / 600 volts = 910.99667 amps/motor

When operating in series, with a 100 volt drop across each of six motors,

  • 546,598 watts/motor / 100 volts = 5465.98 amps/motor

Note that when operating in parallel, 911 amps * 6 motors is 5466 amperes total in the system.

These modern units -- like their AC brethren -- use computer control to reduce (hopefully to eliminate) wheel slip, but even so they can still slip (and stall), and even with arc suppressors and damping material around the brushes, flashovers and destroyed brushes still occur, caused by low speed, wheel slip, rough track, etc., all of which contribute to the woes of a DC traction motor.

Power-wize, contemporary DC traction motor size is getting very close to the practical limits. This is based on such elements as magnetic saturation and the current capacity of the electrical conductors used to build them, coupled with the physical limits of the structure (it would be necessary to use physically larger motors to forestall magnetic saturation: see note below).

In the 1960s, the Southern Pacific and the Rio Grande both acquired diesel-hydraulic locomotives. In the hydraulic transmission, a driveshaft connects the power-plant to the axles, just as in an automobile. In 1961, both roads acquired German-built Krauss-Maffei locomotives, twin-engined 3450 hp, c-c units with a cowl carbody. In 1963, the SP took an additional 15 units with a road-switcher carbody. In 1964 SP acquired the Rio-Grande units. ALCO also made a forray into the diesel-hydralluc experiments, the DH-643, a double-engine, 4300 hp, c-c unit: three units were built, all going to SP in 1964 after testing on the New York Central. In 1970 SP retired its German units, while the ALCOs were scrapped in 1973. The world still had two decades to wait for a better locomotive transmission.

Note: Magnetic saturation is a rather abstract concept that may best be thought of as the limited ability of an object to be magnetized. In the case of a motor the object is usually a piece of iron wound with wires conducting an electrical current. With an applied voltage to the wound wires, a current is caused to flow, and that current flow causes a magnetic field to be created. With more applied voltage, more current and more magnetic field in proportion to the applied voltage. At some point, the iron becomes saturated, increasing the current does not create more magnetic field, and the linear relationship is broken: increasing voltage no longer causes a linear increase in current but instead creates a geometric increase in current -- that is, lots and lots of current, creating lots and lots of heat, burning out the motor. For more on magnetic saturation, including some formulas, see our Formulas and Concepts page under AC Motor Facts.

The move to AC traction motors

It has long been known that AC motors can be more economical than DC motors, just as with their near cousins, alternators and generators. Like alternators, AC motors are not equipped with wear-prone commutators and brushes, eliminating these sources of limitations of the low speed-high throttle position. AC motors would allow locomotives to (1) have more pulling power, (2) avoid stall burns in the traction motors, and (3) have correspondingly lower maintenance requirements.

An early example of AC in a railroad application is the GG1 (designed in 1934), which utilized 12 six-pole motors, 400 volts AC at 25 Hz. Each motor was rated at 385 hp, with the 12 motors mounted in pairs over each of the six driving axles (see our GG1 page for details of the GG1 electricals). In June of 1989, GM began the modern AC traction motor era with its demonstrator, the F69PH-AC, an AC traction version of the F59, followed in 1991 by the SD60MAC. GM delivered its first production unit to Burlington Northern in 1993. GE delivered its first AC-traction unit to CSX in June of 1994.

As a brief technical aside to provide some background and standardize terminology, series-wound DC motors (i.e., motors with commutators and brushes where the field winding and the rotor winding are connected in series) are also called universal motors, universal in the sense that they will run equally well using either AC or DC: simultaneously reversing the polarity of both the stator and the rotor cancel out, thus the motor will always rotate the same direction regardless of the voltage polarity. So a universal motor is in a sense a type of an AC motor in as much as it will operate on AC. The term "universal motor" differentiates it from the more generally thought of AC motor, the AC induction motor, which lacks commutators and brushes. Unfortunately for universal motors, the fact that they do not lack commutators and brushes means that they do lack all of the advantages of what are more typically thought of as AC motors -- the induction motors -- which is the very lack of commutators and brushes! So to say the universal motor "will run equally well using either AC or DC" may be a slight misphrasing: perhaps one should say, "it will run equally badly!" Therefore, the universal motor does not have a role to play in modern electric traction (although universal motors were used in early AC applications in locomotives); rather it is the induction motor that is the "AC motor" to which one refers when speaking of AC traction motors today. That means no brushes to maintain, no flashovers, no commutator to get damaged, no armatures to rewind, and less potential for damage at high power/low rpm situations.

AC locomotives are more expensive due to the control problems inherent in the AC design. An AC motor's speed is traditionally dependant on its design, but it may be controlled by varying the frequency of the input voltage. Being able to vary frequency has been a significant issue in the development of AC motors in high-horsepower traction applications. To deal with the power control problem, both EMD and GE use an AC to DC to AC conversion, control taking place in the DC phase. In an AC traction motor application, the diesel engine drives an alternator, crating AC. This AC is rectified (i.e., converted to DC) and power control takes place in this stage. This is the same place that power control would take place in a conventional AC/DC transmission. At this point, the DC (called the DC link) goes through a solid-state "inverter," which converts the DC back to AC. This AC then powers the motors.

Figure Six, Block Diagram of a Diesel-Electric Locomotive with AC Traction Motors
diesel motor
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
down arrow indicating logical direction of flow
electric motors

Figure Seven, Voltage/Frequency in a Diesel-Electric Locomotive With AC Traction Motors: Output of Alternator to Output of Inverter
AC Power Control Graphic

Control takes place in the stages around the DC Link. The inverter converts the DC back to AC, with the conversion frequency and voltage specifically controlled (this is what then determines the motor's speed). However, this is not simply an inverter, for in modern applicatons of AC motors, with a reliance made on varying the voltage and frequency of the AC to control power more than on simply the brute force approach of the application of current, the inverter must do more than simply convert DC to AC. Complex electronic circuitry in the form of on-board computers now is used to control the inverter. (This has eliminated the need for that classic of the diesel age, the ammeter, in the cab of the AC-motored locomotive, which has been replaced by a tractive effort display.)

The inverter stage is actually a group of inverters, depending on manufacturer either one for each truck (GM) or one for each motor (GE). Each individual inverter consists of six "gated turn-on (GTO) devices," high-power thyristors (that is, "silicon-controlled rectifiers"), three each for the positive and the negative phases of the AC wave in positive/negative pairs. Each positive/negative pair alternate turning-on, chopping the DC into a square wave AC. Each of the three positive/negative pairs turn-on 120 degrees out of phase from each other (turning-on at 0 degrees, 120 degrees, and 240 degrees), producing three-phase AC. While the phase remains constant, the frequency -- how many cycles per second this is repeated -- is varried. Also able to be controlled at this stage is the voltage, how positive and negative the AC becomes. Thus, the frequency and the voltage of the AC arriving at the AC motor is fully controlable, providing the speed control for the locomotive.

Since the frequency and voltage are closely controlled by onboard computer systems, motors cannot run away as they would on a DC locomotive, and the AC motor will not be subject to damaging wheel slip. The use of AC traction motors, coupled with computer controlled wheel creep systems, has allowed AC units to achieve much higher adhesion levels than similar DC units, up to 45% adhesion, versus the 20% range on other units. This has permitted two-for-three and one-for-two replacement of units, with resulting economies in size and maintenance expenses that offset the added initial investment in the purchase of AC units. (Note, however, that there are other issues with such power reductions: for example, a two-for-one reduction on a tradtionally two locomotive run means one locomotive, and if that one locomotive develops problems enroute [not entirely unheard of] there is no backup.)

Computer control technology has also been applied to DC traction motors, including wheel-slip detection and a wheel-creep systems allowing for brief [we're talking fractions of seconds here] applications of power to facilitate very low speed operations. While, this does not fully eliminate problems with high current flow at low speeds in DC motors, these DC wheel-creep systems and wheel-slip detection systems provide dramatically increased adhesion in DC units as well as in AC units and have eliminated many of the operational issues with DC traction: CSX, for example, does not place a minimum continuous speed on its DC-traction GM SD60s and SD70s and GE Dash 8 and Dash 9 locomotives, the same as for all of its AC locomotives [see also note above].)

While the AC traction motor is less complex and has proven itself dependable in long term railroad use (the PRR GG1 used 12 385 hp AC motors ), the purchase of new units with AC traction motors is an expensive undertaking, representing an investment in new technology with maintenance and operational issues not previously encountered, and the new generation of 1000 hp AC traction motors in railroad use represents a new and untested technology, with some railroads still very reluctant to make the transition. (An interesting W3 site on AC motors is

Expanded AC Motor Principals

The Short Version

This isn't expected to make sense, so don't worry. When an AC motor is at rest and an AC voltage is first applied to it, the difference between the aramature speed and the rotating field is 100%. Under these conditions, a high current will flow at the moment the aramature starts to turn. At the moment of starting, the torque is at 0% of the full load torque, but as the speed increases the torque likewise increases. This is in part because, at low speeds, the motor reactance is high, and the current and voltage are very much out phase. This contributes to the low power factor. In an AC motor, maximum power will be generated when the voltage and current are closest to being in phase, so it can be seen that when the voltage and current are out of phase the motor will not be very efficient.

The Long Version

Power in a DC circuit versus Power in an AC circuit.

Power in a DC circuit

As noted above, Power in a DC circuit is simply equal to the voltage times the current. This is expressed as power (in watts, "P") = voltage (in volts, "E" [for Electromotive force) times current (in amps, "I"), or


For example, ten volts times ten amps equals 100 watts. This relationship can be used in reverse to analyze a circuit. A 40 watt bulb on a 12 volt DC circuit must be drawing 3.333 amps. Further, since Ohm's law states that voltage = current tims resistance (E=I*R), it may be seen that the load here is 3.6 ohms. This is all simple and straightforward because this is a DC circuit.

Power in an AC circuit

AC Motors

One of the miricals of the AC motor is that in the AC induction motor, one of the of the two principal components (these two components in the AC motor are the stator and the rotor), the rotor has no visable electrical contacts to the outside world. Instead, it has an electrical field induced into it by the electrical field of the stator -- no commutator, no brushes! (Acutally, some induction motors have brushes and slip rings, but these are used for connecting control and starting equipment to the windings). The induction of the electrical field into the rotor happens because of the characteristic pulsing flow of current in AC. However, this has other affects as well.

  • Reactance

    In an AC circuit, things are different, because in addition to there being a pure resistive load in the circuit there is also reactance in the circuit. Reactance is the unique effect that is displayed in opposition to AC current flow. There are two types of reactance, inductive reactance (that is, a coil), the tendency of the circuit to absorb and store an electrical potential, and capacitive reactance (that is, a capacitor), the tendency of the circuit to absorb and store current. AC circuits can always be quantified in terms of these three forces: resistance, inductive reactance, and capacative reactance. The total oppostion to the AC current flow is called impedance, and it is the vectored sum of the circuit resistance plus the total reactance, inductive and capacitive. Since inductive and capacitive reactance are forces of opposite direction, they counter each other, thus,

    1. Inductive Reactance, Xl, = 2PiFL
    2. Capacitive Reactance, Xc, = 1/(2PiFC)
      • Xl = Inductive Reactance in Ohms,
      • Xc = Capacitive Reactance in Ohms,
      • F = the frequency of the applied AC in Hertz (cycles per second),
      • L = the inductence of the circuit in Henries, and
      • C = the capacitence of the circuit in Farads
    3. Impedance, Z, = (R2 * X2)1/2, where X2 = (Xl - Xc)2
      • Z = Impedance in Ohms,
      • R = Resistance in Ohms, and
      • X = Reactance in Ohms

    AC induction motors are primarilly inductive circuits, so effectively their impedence may be expressed by the formula

    Z = (R2 * Xl2)1/2

  • Power Factor

    In an AC circuit, voltage times current does not equal power; it equals the effective value of voltage and current, which is measured in "voltamperes" (VA). Correcting voltamperes for "power factor" produces the useful or actual power in the circuit, which is measured in watts. So,

    P = VA * pf

    and the value of pf is determined by how much the voltage and current are out of phase. An incandescent light bulb has a power factor of anywhere from 0.95 to 0.99; AC motors may have power factors ranging from .6 to .9; in all of these situations, the current is lagging the voltage -- inductive circuits.

  • Phase Angle

    A purely reactive circuit has a phase angle between the current and voltage of 90 degrees, which results in a power factor of 0.0. The relationship between phase angle and power factor is that power factor equals the cosine of the phase angle. Therefore, power equals the cosine of the phase angle times the voltamperes. In the above example, the cosine of 90 degrees = 0.0. So, at rest, with 90 degree phase angle (purely reactive circuit -- the resistance of the motor's windings is minimal), the useful power of the motor is . . . 0 watts! There are starting strategies, for example, any substantive resistance in the circuit will reduce the phase angle below 90 degrees, thus increasing power factor above 0 and allowing some work to get done. More typically, capacitor-based systems can reduce the phase angle and can be used to start the motor.

    The situation that has developed is that the power developed in an AC motor is related to the magnitude of the voltage, the current, and the internal resistance of the motor (i.e., the simple resistance of the wires), and the frequency of the AC applied to the motor, because the frequency will change the phase angle. (This concept is expanded upon on the Formulas and Concepts page under Power Facts/AC Motor Facts.)

  • Speed Control in AC Motors

    Since an AC motor's speed is based on the frequency of the AC, a change in frequency directly results in a change in speed; however, the change in frequency also changes the reactance of the circuit (because a reduction in the frequency causes a linear reduction in inductive reactance). This in turn changes the impedance of the circuit, the oppositon to the flow of AC current. Speed control may be accomplished in these motors by utilizing solid-state, micro-processor control devices that vary the frequency and voltage of the AC applied to the motor.

    If it were intended to slow an AC motor, the frequency of the applied AC would be reduced: as frequency decreases, circuit reactance decreases, and therefore impedance also decreases. Given a constant voltage, current would increase, potentially to the point where the motor would be damaged. Therefore, a decrease in the frequency must be accompanied by a decrease in voltage sufficient to stabilize the current.

As one last reminder, there is much more detailed information on our Formulas and Concepts page under Power Facts/AC Motor Facts.


revised 10 January 2005 Go to Railroad home page
Made with Cascading Style Sheets Valid CSS! Valid XHTML 1.1! Level Triple-A conformance icon, W3C-WAI Web Content Accessibility Guidelines 1.0 Bobby WorldWide Approved AAA