A series wound motor or generator is a DC machine in which the field winding carries the full armature current because the two windings sit electrically in series. Frank Sprague built the first practical series wound traction motor in 1887 for the Richmond Union Passenger Railway, and the topology has dominated heavy-start applications ever since. Field flux rises with load current, so torque climbs as the square of current at low speeds. That gives huge starting torque — the reason streetcars, hoists, starter motors, and DC arc welders still use this configuration today.
Series Wound Motor Interactive Calculator
Vary the initial and final series current to see how field flux and torque rise in a series wound DC motor.
Equation Used
For an unsaturated series wound DC motor, the field winding and armature carry the same current. Since flux is approximately proportional to current, torque is proportional to current squared.
- Magnetic circuit is below saturation.
- Field current equals armature current.
- Torque is evaluated in the low-speed DC motor region.
Inside the Series Wound Motor or Generator
The series wound machine is the simplest DC topology that produces useful torque under heavy load. You wind a few turns of thick copper around the field poles, run that winding in series with the brushes and armature, and the same current that produces torque in the armature also produces the magnetic field. Double the load current and, below saturation, you roughly double the field flux as well — torque scales with the product of the two, so it climbs as the square of current. That is why a 1.5 kW car starter motor can crank a cold V8 at 200 lb-ft for two seconds without melting.
The penalty is speed regulation. With no load on the shaft, armature current drops, field flux collapses, and back EMF has to make up the supply voltage somehow — so speed runs away. A genuinely unloaded series motor will spin itself apart. You never belt-couple a series motor or run one without a permanent mechanical load. On a 1500 V DC traction motor like the ones GE built for the Pennsylvania Railroad GG1, internal windage and gear losses are enough to keep speed bounded, but on smaller machines you need a direct geared load.
The commutator and brush gear are where these motors fail. Armature reaction — the magnetic distortion caused by armature current pushing the neutral plane off-centre — is severe in a series machine because field and armature current rise together. If the brushes sit even 2 to 3 mechanical degrees off the shifted neutral, you get visible ring fire around the commutator, copper transfer onto the brush face, and bar burning within hours. Interpoles (commutating poles) wound in series with the armature fix this by injecting a counter-flux at the brush zone, and any series motor above about 5 kW will have them.
Key Components
- Series Field Winding: A small number of turns — typically 4 to 40 per pole — of heavy gauge copper or rectangular bar, sized to carry full armature current without exceeding 75 °C rise. Resistance is kept low, often 0.05 to 0.5 Ω total, so I²R loss in the field stays under 3% of input power.
- Armature with Commutator: Lap or wave wound armature with a copper-segment commutator. Bar-to-bar voltage must stay below 30 V on industrial machines and below 18 V on traction motors to prevent flashover. Mica insulation between bars is undercut 0.8 to 1.2 mm so brush wear does not bridge segments.
- Carbon Brushes and Brush Gear: Electrographitic grade brushes for traction (e.g. Morgan EG8) running 35 to 45 N spring pressure per brush. Brushes must be set on the shifted magnetic neutral within ±2°, otherwise commutation degrades and you get sparking at the trailing edge.
- Interpoles (Commutating Poles): Small auxiliary poles between main poles, wound in series with the armature. They cancel armature reaction at the brush zone. The interpole air gap is set with shim stock to ±0.05 mm — get it wrong and commutation collapses under heavy load.
- Yoke and Main Pole Cores: Solid cast steel yoke for slow-changing flux, or laminated for variable-load duty. Main pole faces are laminated 0.5 mm sheet to limit eddy losses from armature-reaction flux pulsation. Pole face air gap typically 2 to 4 mm depending on saturation point chosen.
Where the Series Wound Motor or Generator Is Used
You see series wound machines anywhere a load demands huge torque from a dead stop and tolerates wide speed variation. The topology dominates traction, hoisting, cranking, and any handheld AC tool that needs both portability and grunt — a universal motor is just a series wound DC machine designed to also run on AC.
- Rail Traction: GE 752 series wound DC traction motor used in the EMD GP38-2 locomotive, rated 1,000 hp continuous per axle at 740 V DC.
- Automotive: Bosch 0001-series engine starter motors on Mercedes OM642 diesel engines, drawing 800 to 1,200 A at 12 V to crank at 200 RPM.
- Material Handling: Demag DRC overhead crane hoist motors on steel-mill ladle cranes, sized 30 to 150 kW with series field for full-load lift from rest.
- Power Tools: Milwaukee 6177-20 14-inch abrasive cut-off saw using a 15 A universal motor — series wound to deliver 4.0 hp peak from a portable AC supply.
- DC Welding: Lincoln Electric SAM-400 engine-driven welder using a series-cumulative compound generator with a strong series field to give the drooping volt-ampere curve stick welding needs.
- Heritage Transit: GE 247 motors on the San Francisco PCC streetcars at the Market Street Railway, 50 hp at 600 V DC per axle.
The Formula Behind the Series Wound Motor or Generator
The defining relationship for a series wound machine is how torque scales with armature current. Below magnetic saturation, field flux φ is roughly proportional to Ia, so torque scales as the square of current — that is the regime where the motor delivers its enormous starting torque. As current climbs and the iron saturates, flux flattens out and torque returns to a linear function of current. At the low end of the typical operating range — say 25% of rated current — you are deep in the I² regime and torque feels almost gentle. At rated current you are at or just past the knee of saturation, where the design sweet spot sits. Push past 200% of rated and you are firmly linear, and the limiter is now commutator heating, not flux.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Shaft torque produced by the motor | N·m | lb-ft |
| Kt | Machine constant (depends on pole count, turns per coil, parallel paths) | N·m / (Wb·A) | lb-ft / (Wb·A) |
| φ | Air-gap flux per pole | Wb | Wb |
| Ia | Armature current (also equal to field current in a series machine) | A | A |
| kf | Field flux constant below saturation, φ / Ia | Wb/A | Wb/A |
Worked Example: Series Wound Motor or Generator in a quarry haul truck retrofit
A surface aggregate operation in Quebec is retrofitting an old Euclid R-32 rear-dump haul truck with a refurbished GE 776 series wound DC traction motor on each rear wheel. Each motor is rated 150 hp continuous, Kt × kf = 0.018 N·m/A² below saturation, saturation knee around 350 A, rated current 480 A. The fleet engineer needs to know shaft torque at three operating points: pulling away from the loader at 240 A, cruising up the haul road at rated 480 A, and stalling against the brake at 700 A.
Given
- Kt × kf = 0.018 N·m/A²
- Ia,low = 240 A
- Ia,nom = 480 A
- Ia,high = 700 A
- Saturation knee = 350 A
Solution
Step 1 — at the low end of the typical operating range, 240 A, the machine is well below the 350 A saturation knee, so use the squared form directly:
That is roughly 765 lb-ft per motor at half-current. The truck pulls away from the loader with controlled, shovel-it-on torque — no jerk, no wheel slip on damp shot rock.
Step 2 — at nominal 480 A you are past the saturation knee, so the pure I² formula overstates torque. Above the knee, flux flattens. A reasonable engineering correction is to compute torque using the knee current squared plus a linear term above it:
About 2,230 lb-ft per motor — the design sweet spot. The truck climbs an 8% grade loaded at steady road speed without overheating the field winding.
Step 3 — at 700 A stall, the machine is deep into saturation. Linear extrapolation from the knee:
Roughly 3,250 lb-ft. Pure I² would have predicted 8,820 N·m — twice the real value — which is exactly the trap engineers fall into when they ignore saturation. At 700 A you have about 2 minutes before commutator bar temperature climbs past 105 °C and the solder in the riser connections starts to creep.
Result
Nominal shaft torque at rated 480 A is approximately 3,024 N·m (2,230 lb-ft) per motor, which is enough to hold the truck on an 8% grade with a full 32-ton payload. The range tells the real story — at 240 A you get 1,037 N·m of gentle pull-away torque, at 480 A you sit at the design sweet spot, and at 700 A you reach 4,410 N·m but only briefly before thermal limits bite. If you measure significantly less than 3,024 N·m at rated current, suspect three things first: brushes shifted off the magnetic neutral and sparking away usable EMF (look for trailing-edge fire on the commutator), an interpole shim that has compressed and widened the air gap past 0.05 mm tolerance which kills commutation under load, or a series field with one shorted turn — easy to spot because the field resistance reads 5 to 10% low and the motor pulls more current for the same shaft torque.
Series Wound Motor or Generator vs Alternatives
Series wound is one of three classical DC field arrangements, and the choice between them comes down to what the load looks like at start versus at speed. Here is how series stacks up against the shunt and compound alternatives on the engineering dimensions that actually drive selection.
| Property | Series Wound DC Motor | Shunt Wound DC Motor | Compound Wound DC Motor |
|---|---|---|---|
| Starting torque (% of rated) | 300-500% | 150-200% | 250-350% |
| Speed regulation (no-load to full-load) | Very poor, 30-100% drop, runaway risk on no-load | Excellent, 2-5% drop | Good, 10-20% drop |
| Typical application fit | Traction, hoists, starters, cranes, universal power tools | Machine tool spindles, conveyors, fans needing constant speed | Rolling mills, plunger pumps, shears with shock loads |
| Commutation difficulty under heavy load | Severe, interpoles essential above 5 kW | Moderate, interpoles common above 50 kW | Severe, similar to series, interpoles essential |
| Cost (relative, same frame size) | 1.0× baseline | 1.05-1.10× | 1.15-1.25× |
| Typical brush life | 1,500-4,000 hours under traction duty | 4,000-8,000 hours steady duty | 2,000-5,000 hours mixed duty |
| Behaviour on AC supply | Runs as universal motor with reduced efficiency | Will not run on AC | Will not run on AC |
Frequently Asked Questions About Series Wound Motor or Generator
That is the classic series-motor runaway and it is not a fault — it is the topology behaving correctly. With no shaft load, armature current drops to whatever bearing and windage friction demands, which can be 1 to 2% of rated. Field flux drops in proportion, so the back EMF needed to balance the supply voltage can only come from speed climbing — and there is nothing to stop it climbing until the armature comes apart at the banding wire.
Rule of thumb: never bench-test a series motor without either a permanent mechanical load (a friction brake, a fan, a generator) or a current-limited supply set to perhaps 20% of rated voltage. On traction machines this is also why field weakening is done by shunting the field, never by opening it.
If the hoist runs on existing DC bus voltage from a substation rectifier — overhead cranes in steel mills, mine hoists, legacy port equipment — series DC still wins on cost, ruggedness, and the ability to handle plug braking and dynamic braking with a simple grid resistor. The torque-current curve is exactly what a hoist wants.
For a new greenfield install with AC mains available, an inverter-fed squirrel-cage induction motor with vector control gives you the same starting torque, no brushes to dress, no commutator to skim, and built-in regenerative braking. The crossover point in our experience sits around 75 kW and 5,000 starts/year — below that, AC+VFD usually wins; above that, the ruggedness and overload capacity of series DC starts paying back.
Hot-engine cranking demands roughly 30 to 50% more breakaway torque because oil viscosity at the rings is lower but ring-to-bore friction at piston-skirt contact rises, and the pinion gear is also fighting hot, expanded ring gear teeth with less backlash. The starter is being asked for torque it could give cold, but heat-soak from the manifold has raised armature winding temperature, which raises copper resistance — a 60 °C rise adds about 23% to winding resistance. Less current flows for the same battery voltage, and because torque scales as I², a 10% current drop costs 19% torque.
Diagnostic: measure cranking current cold and hot. If hot current is 15% or more below cold, the starter is heat-soaked. Either heat-shield it or fit a higher-spec unit (Bosch HD or Denso reduction-gear type) that runs cooler under sustained crank.
Two effects. First, on AC the field flux pulses at twice line frequency and the average flux is lower than the steady DC equivalent because the iron does not fully magnetise in each half-cycle — lower flux means higher speed for the same back EMF. Second, the impedance of the field winding adds an inductive voltage drop on AC that does not exist on DC, so for the same RMS supply voltage, less voltage actually drives torque.
The net result is that a universal motor runs typically 5 to 15% faster on AC than on equivalent DC, with slightly less torque. This is why the nameplate on a Milwaukee or Makita tool quotes a no-load speed at 120 V AC, 60 Hz specifically — not at 120 V DC.
Series-cumulative compound generators used in welders depend on a precise ratio of series-field ampere-turns to shunt-field ampere-turns to produce the drooping V-I curve stick welding needs. If the series field has been rewound with even slightly different wire gauge or turns count from original, the droop characteristic changes and arc stability with it. A 5% change in series-field turns can shift the short-circuit current by 15 to 20%.
Check the original drawing for series-field turns per pole and compare to what is actually on the poles now. The other common cause is a worn equalizer connection between brush rigging — on twin-armature welders these must be intact or the two halves fight each other and arc voltage hunts at 1 to 3 Hz.
At the instant of switching, the motor is stationary, back EMF is zero, and the only thing limiting current is the combined armature-plus-field resistance — typically 0.1 to 0.5 Ω on a medium machine. On a 250 V DC supply that means inrush of 500 to 2,500 A for the first few cycles until the rotor accelerates and back EMF builds. This is normal and is exactly what produces the huge starting torque.
Don't size overcurrent protection on rated current — use a Class D or Class K time-delay device rated 250 to 400% of FLA, or better, use a stepped starting resistor (the original solution — see any 1920s streetcar controller) or a current-limited chopper drive. The thermal element should still trip on sustained overload, just not on the legitimate 8× inrush spike.
References & Further Reading
- Wikipedia contributors. DC motor. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
