Armature DC Motor Guide: Laminations, Windings, and Core Assembly Choices

A converter line runs three 24 V permanent-magnet DC gearmotors at 1,800 rpm. At roughly 4,500 operating hours, one unit starts sparking visibly at the brush boxes and loses about 6% of its speed under the same load. The windings test clean. The fault sits in the armature: a grooved commutator, a rotor stack running a few hundredths of a millimetre eccentric since a rebuild, and brushes bedding on barely 60% of their face.

What follows is a practical view of the armature side of a DC motor: the magnetic circuit, the winding, the commutation limit, and how the core is actually bought, from sample laminations to series delivery.

How the Armature Produces Torque, and Where Commutation Bites

Two relationships explain most bench measurements on a brushed DC motor, and both point straight back to the armature.

• Torque: T ≈ k · Φ · Ia, field flux multiplied by armature current.
• Back EMF: E ≈ k · Φ · ω, which rises with speed and sets the running current at a given supply voltage.
• Current path: supply, brush, commutator segment, armature coil, return brush.
• Commutation: the commutator reverses coil current as each coil crosses the magnetic neutral, so coil current alternates even though the supply is DC.
• Armature reaction: the armature's own magnetomotive force tilts and weakens the main field and shifts the neutral axis. At high current it appears as sparking on the trailing brush edge and a small loss of torque per ampere.

Magnetic circuit

Rotor and stator laminations are stacked from non-oriented electrical steel, normally 0.35 mm or 0.50 mm. Thinner gauge trims eddy-current loss but lowers stacking factor and raises cost per kilogram, so it is rarely a free improvement.

Winding

Copper fill, wire gauge and end-turn length set the I²R loss. Insulation class F or H sets how much thermal margin you are allowed to use before the winding is the limiting part.

Commutation

Brush grade, spring pressure, commutator eccentricity and undercut depth decide sparking. Runout beyond roughly 0.02 mm usually shows up as brush noise long before it shows up as a failure, which is why the fault above was traceable at 4,500 hours rather than at 500.

Armature Voltage or Field Control: Decide Before You Buy

Speed range is a control decision, but it constrains the armature design. Armature voltage control, used below base speed, delivers roughly constant torque and wide turndown — the usual choice for conveyors, feed rolls and positioning axes. Field weakening, used above base speed, delivers roughly constant power, but commutation margin falls as flux drops and the field supply has to stay stable under load.

PWM drives at 20 kHz and above keep current ripple, and the extra armature heating that comes with it, small; below that, check brush and core losses at the actual carrier frequency rather than assuming the datasheet number holds. Series-wound motors cannot be field-weakened the same way, because field and armature share the same current. Use series-rated drives or resistance control, and accept that the speed regulation will be softer.

Three Ways to Build the Rotor Stack, and What Each One Costs

The trade-off is not abstract. An interlocked stack saves tooling money, but the notches create local saturation and metal-to-metal contact. An adhesive-bonded stack keeps the insulation coating intact and usually measures the lowest core loss, at the price of fixtures and cure time. Welded stacks sit in between and are chosen when the mechanical integrity of a long stack matters more than a couple of watts. Where interlock geometry is the deciding factor, the behaviour of interlocked laminated cores in precision assemblies is worth reviewing before the drawing is frozen.

If the design is still open, buying rotor and stator laminations to drawing is the fastest way to compare stack builds before committing to a die.

DC Motor Stator and Rotor LaminationsCustom stator and rotor laminations support design-stage stack comparisons before tooling commitment, helping evaluate magnetic loss and assembly fit.View Product →

Losses You Will Not Find on the Datasheet

A datasheet gives one efficiency figure at one operating point. The split between loss mechanisms is what decides whether the armature runs warm at half load or overheats at 80% of rating.

Indicative split in a small brushed DC motor at rated load; actual shares shift with flux density, speed and brush grade.

Copper loss scales with the square of current, so a 10% overload adds roughly 21% more heat in the winding. That is why brush and winding failures rarely occur at the current printed on the label. Core loss is fixed by flux density, steel gauge and the number of interlaminar shorts, which means a stack with scratched coating heats up even at no load — a useful diagnostic when an assembled motor idles hot.

Where Brushed DC Armatures Still Win

Brushed DC is not legacy everywhere. It holds on where a two-wire connection, high starting torque and a low-cost drive outweigh periodic brush maintenance.

Indicative split of typical brushed DC armature demand, not a market study.

Each segment stresses the armature differently. Automotive auxiliaries punish thermal cycling and vibration, medical and mobility units punish acoustic noise and commutator film stability, and power tools punish peak current. A stack that suits one rarely suits all three without changes to steel grade or build method.

Buying the Core: Tolerances, Tests and the Risks You Carry

Where purchase risk actually accumulates is upstream of the drawing. Five failures account for most of the disputes we see:

• Steel grade substituted for a cheaper non-oriented grade with higher loss.
• Weld spatter or restruck interlocks that flatten the stack and short the laminations.
• Coating damage from bulk handling before the stack is bonded.
• Material supplied unannealed, which lifts core loss without changing a single dimension.
• Lots mixed inside one stack, showing up as unit-to-unit scatter in current draw.

For volume programmes, buying a finished stack rather than loose sheets moves those risks to the supplier and removes the stack-length variation from your assembly line.

DC Motor Stator and Rotor Core AssembliesFinished laminated core assemblies suit volume motor programmes by shifting stacking risk to the supplier and reducing stack-length variation on the assembly line.View Product →

From Sample to Series: A Sequence That Holds

Samples cut by laser also let you test a steel grade change before committing to a die. A 0.35 mm grade that measures lower core loss may or may not justify the extra cost per kilogram, and a stack trial answers that question in a week rather than a quarter.

Laser-Cut Stator and Rotor Lamination SamplesCustom laser-cut lamination samples support early motor design checks and steel-grade trials before die commitment and volume production.View Product →

Past the sample stage, the lamination and core manufacturing route — steel slitting, stamping or laser cutting, annealing where specified, stacking by interlock, adhesive or weld, and final inspection — is where delivery dates are won or lost. It is worth walking that flow once with the supplier before the first purchase order, not after.

Maintenance and Compliance Notes

• Inspect brushes at half the expected wear life, not at failure.
• A uniform chocolate-coloured commutator film is healthy; patchy film points to contamination or the wrong brush grade.
• Keep spring pressure in specification. Low pressure glazes the commutator, high pressure wears brushes quickly.
• Clean with a solvent compatible with the winding insulation; carbon dust conducts.
• Store stacked cores dry and wrapped, because interlaminar rust behaves exactly like an electrical short.

On the compliance side, motor cores for automotive programmes generally sit inside an IATF 16949 quality system, general industrial cores under ISO 9001, and welded frames and bases under ISO 3834, with environmental management following ISO 14001. Those certificates matter less as paper and more as the reason first-article reports, material certificates and change notifications arrive without being chased.

None of this makes the armature complicated. It makes it specific: steel grade, stack build, burr limit, commutator runout. Settle those four and the 4,500-hour sparking problem stops being a mystery.