DC machine construction and EMF/torque equations

Open up any DC motor or DC generator and you find the same five parts: a yoke, poles, an armature, a commutator, and brushes. The machine does not care which way you run it — feed it electricity and the shaft turns (motor); turn the shaft and electricity comes out (generator). What makes this possible is a clever piece of rotating hardware called the commutator, which takes the alternating EMF generated in each spinning armature conductor and rectifies it into a steady DC quantity at the terminals. This article builds the construction picture first, then derives the two equations that govern everything a DC machine does: the EMF equation $E = \dfrac{P\Phi NZ}{60A}$ , which says how much voltage the armature generates, and the torque equation $T = \dfrac{P\Phi ZI_a}{2\pi A}$ , which says how much turning force it produces. Both come from the same conductors, the same flux, and the same winding geometry — so once you know one, the other is a short step away.

The core insight

A DC machine's construction — poles producing flux $\Phi$ , armature conductors $Z$ cutting or carrying current through it, commutator rectifying the result — fixes both its EMF ( $E=\tfrac{P\Phi NZ}{60A}$ ) and its torque ( $T=\tfrac{P\Phi ZI_a}{2\pi A}$ ) through the shared geometric constant $\tfrac{PZ}{A}$ .

The core idea

The five parts, and what each does. A DC machine splits cleanly into a stationary side and a rotating side.

Yoke — the outer cylindrical frame. It gives the machine mechanical strength and also completes part of the magnetic circuit, carrying flux from one pole to the next.
Poles and pole shoes — bolted to the inside of the yoke, wound with the field winding. Current in this winding sets up the main magnetic flux per pole, $\Phi$ , which the pole shoes spread out evenly over the air gap facing the armature.
Armature — the rotating, slotted iron core mounted on the shaft. Its slots hold the armature conductors, and being iron (not air) it gives the flux a low-reluctance path to cross the air gap.
Commutator — a cylinder of insulated copper segments fixed to the shaft, each segment wired to one point of the armature winding. As the shaft turns, a given segment slides under first one brush polarity, then the other, exactly in step with the conductor beneath it reversing which pole it faces — so it always delivers the same-sign EMF to the external circuit even though the conductor's own EMF alternates.
Brushes — stationary carbon contacts pressing on the commutator's surface, forming the sliding electrical junction between the rotating winding and the fixed external terminals.

Why the equations follow from this picture. Every armature conductor sweeps through the flux $\Phi$ set up by the poles as the armature turns at speed $N$ ; by Faraday's law each one generates an EMF, and the commutator/brush arrangement collects these into $A$ parallel paths, all $Z$ conductors contributing. Sum the individual EMFs over one revolution and you get the EMF equation below. Run the same machine as a motor instead — force current $I_a$ through those same conductors sitting in the same field — and each one feels a force by the motor effect; summing these forces over the armature radius gives the torque equation.

E = \frac{P\Phi NZ}{60A}

where: $P$ is the number of poles, $\Phi$ is the flux per pole in weber, $N$ is the armature speed in rpm, $Z$ is the total number of armature conductors, and $A$ is the number of parallel paths in the winding (simple lap winding: $A=P$ ; simple wave winding: $A=2$ ).

T = \frac{P\Phi ZI_a}{2\pi A}

where: the symbols are as above, with $I_a$ the total armature current. Both equations carry the same combination $\tfrac{PZ}{A}$ , a pure machine constant fixed by construction; calling it $k$ , the pair compresses to $E=k\Phi N$ and $T=k'\Phi I_a$ — EMF tracks flux and speed, torque tracks flux and current, and neither equation can be borrowed for the other's job because a $60$ (minutes-to-seconds) and a $2\pi$ (revolutions-to-radians) are doing genuinely different unit conversions.

Figure

Cutaway of a DC machine showing yoke poles armature commutator and brushes labelled around the cross section

Derivation ADeriving the EMF equation from one conductor to the whole armature

The goal is to build $E=\dfrac{P\Phi NZ}{60A}$ from scratch, starting with the EMF a single conductor generates as it sweeps through one pole's flux, then scaling up to all $Z$ conductors arranged in $A$ parallel paths, keeping every unit conversion explicit so the $60$ in the denominator is not a mystery.

Step 1. Find the EMF of one conductor over one revolution. In one revolution, a conductor sweeps past all $P$ poles, cutting a total flux of $P\Phi$ (each pole contributes $\Phi$ webers as the conductor passes under it). If the armature turns at $N$ revolutions per minute, one revolution takes $\tfrac{60}{N}$ seconds. By Faraday's law, the average EMF induced in that one conductor is flux cut divided by time taken: $e = \dfrac{P\Phi}{60/N} = \dfrac{P\Phi N}{60}$ .

Step 2. Scale from one conductor to one parallel path. The armature winding is not one long conductor but $A$ separate parallel paths, each made of several conductors connected in series. Since all $Z$ conductors are shared equally among the $A$ paths, each path contains $Z/A$ conductors in series, and their individual EMFs (each equal to $e$ from Step 1, since every conductor cuts the same flux pattern) add directly.

Step 3. Assemble the total terminal EMF. The EMF appearing at the brushes is the EMF of one path (since all $A$ paths are identical and simply supply current in parallel — they don't add voltage, only current capacity): $E = \left(\dfrac{Z}{A}\right)\times e = \dfrac{Z}{A}\times\dfrac{P\Phi N}{60} = \dfrac{P\Phi NZ}{60A}$ .

Step 4. Sanity-check the roles of $60$ and $A$ . The $60$ converts the given speed $N$ from rpm into a per-second rate — it has nothing to do with poles or windings, which is why it never moves into the torque equation (torque uses angle, not time, so it needs $2\pi$ instead). The $A$ in the denominator correctly reduces $E$ for more parallel paths, because more paths mean fewer conductors in series per path, hence less voltage per path — exactly matching how a lap winding ( $A=P$ , many parallel paths, more current capacity, lower vots per unit conductor count) differs from a wave winding ( $A=2$ , fewer paths, higher voltage).

Worked example

Problem. A 6-pole DC generator has a wave-wound armature ( $A=2$ ) with 720 conductors and a flux per pole of 0.015 Wb. Find the EMF generated when the armature runs at 900 rpm, and then find the torque it would develop as a motor drawing an armature current of 40 A.

Step 1. List the given values: $P=6$ , $\Phi=0.015$ Wb, $Z=720$ , $A=2$ , $N=900$ rpm, and (for the second part) $I_a=40$ A.

Step 2. Apply the EMF equation. $E = \dfrac{P\Phi NZ}{60A} = \dfrac{6\times0.015\times900\times720}{60\times2}$ .

Step 3. Work the numerator: $6\times0.015=0.09$ ; $0.09\times900=81$ ; $81\times720=58320$ .

Step 4. Work the denominator and divide: $60\times2=120$ , so $E = 58320/120 = 486$ V.

Step 5. Apply the torque equation with the same machine constants and $I_a=40$ A: $T = \dfrac{P\Phi ZI_a}{2\pi A} = \dfrac{6\times0.015\times720\times40}{2\pi\times2}$ .

Step 6. Work the numerator: $6\times0.015=0.09$ ; $0.09\times720=64.8$ ; $64.8\times40=2592$ .

Step 7. Work the denominator: $2\pi\times2=4\pi\approx12.57$ , so $T = 2592/12.57 \approx 206.2$ N·m.

Answer: $E = 486$ V as a generator; $T \approx 206.2$ N·m as a motor at the same $I_a$ , same machine.

Competitive stretch

Notice that both results came from the same $\tfrac{PZ}{A} = \tfrac{6\times720}{2} = 2160$ machine constant — once you compute it, $E = 2160\Phi N/60$ and $T = 2160\Phi I_a/2\pi$ are just two different multipliers on the same $\Phi\times(\text{conductor factor})$ , a shortcut worth using when a question asks for both quantities on one machine.

Common traps

Swapping the $60$ and $2\pi$ between the two equations — they convert different things (rpm to rev/s versus revolutions to radians) and are not interchangeable.
Using the wrong $A$ for the winding type — lap winding gives $A=P$ , wave winding gives $A=2$ ; picking the wrong one scales the answer by a factor of roughly $P/2$ .
Treating generator and motor as structurally different machines — construction and both equations are identical; only the direction of energy flow changes.
Forgetting that each individual armature conductor generates an AC EMF (it faces alternating N and S poles as it rotates) and that it is specifically the commutator that rectifies this into the DC seen at the brushes.

Ask the tutor: want to work through how these two equations combine with the back-EMF relation $E_b=V-I_aR_a$ to describe a running DC motor?

Why does the EMF equation use N in rpm but the torque equation does not mention speed at all?

EMF is generated by conductors cutting flux, and how fast they cut it depends on how fast the armature turns — hence $N$ appears, and the $60$ converts that rpm figure into a per-second rate to match the volt-per-second definition of EMF. Torque, in contrast, is a force-times-radius effect that exists the instant current flows through a conductor sitting in a field, regardless of whether the armature is spinning fast, slow, or momentarily stalled. That is why $T=\dfrac{P\Phi ZI_a}{2\pi A}$ has no $N$ : a stalled motor at full field and full armature current still develops full torque, even though it generates zero back-EMF.

If lap and wave windings use the same conductors and flux, why do they give such different voltages?

Because $A$ changes how those conductors are grouped. A lap winding sets $A=P$ , meaning many short parallel paths, each with relatively few conductors in series — good for high current, lower voltage per path. A wave winding sets $A=2$ regardless of $P$ , meaning just two long parallel paths, each stringing together far more conductors in series — good for higher voltage at lower current. Since $E\propto 1/A$ for fixed $Z$ , $\Phi$ , $N$ , the wave winding of the same machine gives a noticeably higher EMF than the lap winding would, which is exactly the trade-off winding designers exploit.

Is the torque given by T = P phi Z Ia / 2 pi A the same as the useful shaft torque you'd measure with a dynamometer?

No, not exactly — this is the electromagnetic (developed) torque, produced purely by the interaction of armature current with the field flux. The torque actually available at the shaft is smaller, because some of it is used up overcoming friction and windage (air resistance) losses inside the machine itself. For the construction and EMF/torque-equation scope of this article, only the developed torque $T=\dfrac{P\Phi ZI_a}{2\pi A}$ is required; the distinction between developed and shaft torque belongs to the machine-losses and efficiency topics that follow.

DC machine construction and EMF/torque equations

The core idea

Worked example

Competitive stretch

Common traps

Before this

Where this leads