A ring armature is a rotating DC machine winding wrapped continuously around a closed iron ring, with tap points connecting at equal intervals to commutator segments. It was the first practical alternative to the simple two-pole shuttle armature, and modern DC machines use the drum armature instead because the ring wastes the inside half of every coil. The ring solved early commutation problems by giving multiple parallel paths and smooth EMF — it powered the first commercial generators, including Gramme's 1871 dynamo and Edison's Jumbo, delivering steady DC at 100+ volts decades before drum windings matured.
Ring Armature Interactive Calculator
Vary flux, speed, conductors, poles, and parallel paths to see generated DC EMF and ring-armature copper waste.
Equation Used
The DC machine EMF equation multiplies flux per pole, active conductors, speed, and pole count, then divides by 60 and the number of parallel paths. For a ring armature, the inside half of the winding is treated as inactive copper, so the calculator also shows an approximate drum-armature equivalent at twice the active conductor utilization.
- Only outer ring conductors are active in the magnetic field.
- Flux per pole is uniform and expressed as total useful pole flux.
- Parallel paths follow the commutated DC armature model.
- Ring copper utilization is approximated as 50 percent because the inner half of each turn is inactive.
Operating Principle of the Ring Armature
The ring armature is conceptually simple. You take a doughnut-shaped iron core, wrap insulated copper wire around it as a continuous closed helix, then bring out a tap at every Nth turn to a commutator segment. When the ring rotates inside a magnetic field set up by stationary pole pieces, only the conductors on the outer face of the ring cut flux lines — the conductors threaded through the ring's hole sit inside the iron and see almost no field. Brushes ride on the commutator at the magnetic neutral axis and pick off the EMF from whichever conductors are momentarily under the poles.
The winding forms two parallel paths between the brushes in a 2-pole machine, four paths in a 4-pole. That parallelism is the whole point — it averages the per-conductor EMF, gives smooth DC with low ripple, and lets the machine deliver real current without arcing the brushes. If your tap spacing is uneven, or if the brushes sit off the neutral axis by more than a few degrees, you get circulating currents between paths and the commutator burns. On a typical 1880s Gramme machine the brush position had to be trimmed within ±2° of mechanical neutral or you'd see visible sparking within minutes.
Why did the drum armature replace it? Because the ring wastes copper. Roughly half of every turn — the part threaded through the centre — generates no useful EMF and just adds I²R loss. A drum armature puts every conductor on the outer surface where it cuts flux, doubling the effective copper utilisation. The ring also concentrates heat inside the bore where you can't cool it. On any restoration of a Pacinotti or Gramme machine, you'll find the inner copper darkened and the outer copper still bright — that's a century of waste heat written into the winding.
Key Components
- Ring core: A laminated soft-iron ring, typically 0.35 mm to 0.5 mm laminations stacked and clamped, providing the magnetic path. Lamination thickness controls eddy-current loss — go above 0.5 mm and core loss climbs sharply at any speed over 600 RPM.
- Continuous helical winding: Insulated copper wire, usually cotton-and-shellac on heritage machines, wrapped as one closed helix around the ring. Turn count between taps sets the EMF per segment; on an Edison Jumbo this was around 40 turns per tap for 110 V output at roughly 350 RPM.
- Commutator segments: Hard-drawn copper bars insulated with mica, one bar per winding tap. Mica thickness is critical — 0.8 mm is standard, undercut 1.0 mm to 1.5 mm below the copper surface to prevent brush hangup as the copper wears.
- Field poles: Stationary iron pole pieces wrapped with field coils carrying DC excitation. Pole-face arc typically covers 65% to 70% of the pole pitch — narrower and you lose flux, wider and adjacent poles short magnetically.
- Brushes and brush gear: Carbon or copper-graphite blocks held in spring-loaded holders against the commutator. Brush pressure runs 15 to 25 kPa on heritage DC machines; below 15 kPa you get bouncing and arcing, above 25 kPa the commutator wears in months not decades.
Who Uses the Ring Armature
Ring armatures live almost entirely in heritage and educational territory now. You see them in operating museum dynamos, in physics-department teaching rigs where the geometry needs to be visible, and in restoration projects where a working replica must match the original. Modern industrial DC machines all use drum windings — but if you're rebuilding a pre-1900 generator, you're working with a ring armature whether you like it or not.
- Science museums: The Smithsonian's working replica of an Edison Jumbo dynamo at the National Museum of American History uses a 4-pole ring armature feeding a 110 V DC display load.
- University teaching labs: MIT's Edgerton Center has used cutaway Gramme ring demonstrators to show students how commutation produces DC from a rotating coil.
- Heritage power exhibits: Greenfield Village's Menlo Park reconstruction runs period-correct ring-armature generators on demonstration days, sized for the original Edison incandescent loads.
- Industrial heritage restoration: The Boott Cotton Mills Museum in Lowell maintains a Crocker-Wheeler dynamo that uses a transitional ring-style armature feeding original DC loom motors.
- Film and theatre props: Practical-effects shops building period-accurate laboratory sets — Frankenstein revivals, Tesla biopics — wind functioning ring armatures for on-camera sparking dynamos.
- Maker and education kits: Hands-on physics kits like the Eisco Labs Gramme ring model let students measure EMF directly across two brushes as they spin the armature by hand.
The Formula Behind the Ring Armature
The generated EMF of a ring armature follows the same fundamental equation as any DC machine — total flux per pole, times poles, times rotational speed, times conductors, divided by parallel paths. What matters in practice is how the answer changes across your operating range. At the low end of typical heritage-machine speed (around 200 RPM on a small Gramme dynamo) the EMF barely reaches useful levels and brush contact is marginal because there isn't enough back-EMF to stabilise the arc. At the nominal design point (around 350 RPM on an Edison Jumbo class machine) you hit rated voltage and the commutation is clean. Push past the high end (above 600 RPM on the same geometry) and centrifugal force starts throwing winding wedges, the commutator runs hot, and you lose voltage to armature reaction faster than speed gains it back.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| E | Generated EMF between brushes | volts (V) | volts (V) |
| Φ | Flux per pole | webers (Wb) | lines × 10-8 |
| P | Number of magnetic poles | dimensionless | dimensionless |
| N | Armature rotational speed | RPM | RPM |
| Z | Total active conductors on outer ring face | dimensionless | dimensionless |
| A | Number of parallel paths (= P for ring/lap winding) | dimensionless | dimensionless |
Worked Example: Ring Armature in a heritage telegraph museum's replica Gramme dynamo
A working communications heritage exhibit in Cornwall is hand-winding a replica Gramme ring dynamo to power a period-correct Morse sounder loop at 60 V DC. The build uses a 4-pole field with Φ = 0.012 Wb per pole, a ring carrying Z = 240 active outer-face conductors, and the operator wants to know the EMF across the brushes at the design speed of 400 RPM, plus what happens if they slow to 200 RPM or push to 700 RPM during demonstrations.
Given
- Φ = 0.012 Wb per pole
- P = 4 poles
- Z = 240 conductors
- A = 4 parallel paths
- Nnom = 400 RPM
Solution
Step 1 — at the nominal design speed of 400 RPM, plug straight into the EMF equation:
That sits about 20% below the 60 V target — the museum will need either more flux (re-shimming the field gap to lift Φ to 0.015 Wb) or more conductors. This is the kind of mismatch you only catch with the equation, not by eye.
Step 2 — at the low end of the demonstration range, 200 RPM, EMF scales linearly with speed:
At 24 V the Morse sounder coils will buzz weakly but won't latch the armature cleanly — visitors will see a flickering signal that reads as broken-down equipment, not authentic operation. The sounder needs around 40 V minimum to snap properly.
Step 3 — at the high end, 700 RPM, again linear with speed:
84 V theoretically, but in practice you won't see it. Above roughly 600 RPM on a hand-wound ring this size, armature reaction starts shifting the magnetic neutral axis 5° to 8° forward of the brush position, and you lose 10% to 15% of the predicted EMF to commutation losses and brush sparking. The brushes will also start glowing visibly at the trailing edge, which is the cue to back off the speed.
Result
Nominal generated EMF at 400 RPM is 48 V across the brushes. That tells the operator the build as designed falls short of the 60 V Morse-loop target by about 20%, and the fix is to widen Φ rather than spin faster. Across the operating range, the machine delivers 24 V at 200 RPM (too weak to latch the sounder), 48 V at 400 RPM (workable but under-spec), and a theoretical 84 V at 700 RPM that collapses to roughly 70 V real-world because of armature reaction and commutator sparking — so the sweet spot is actually narrow. If the operator measures less than 48 V at 400 RPM, the most likely causes are: (1) field coil polarity reversed on one pole pair, halving net flux, (2) brush rigging set off-neutral by more than 3°, dumping EMF into circulating current, or (3) inter-segment mica shorts on the commutator caused by copper smearing during initial run-in.
When to Use a Ring Armature and When Not To
The ring armature is almost never the right choice for new construction — but understanding why tells you a lot about why DC machine design evolved the way it did. The comparison that matters is ring versus drum (the modern standard) and ring versus simple shuttle armature (what it replaced).
| Property | Ring Armature | Drum Armature | Shuttle (H) Armature |
|---|---|---|---|
| Copper utilisation | ~50% (inner half of each turn wasted) | ~95% (all conductors active) | ~80% (two coil sides only) |
| EMF ripple at brushes | Low — many parallel paths | Very low — many slots | High — single coil pulsing |
| Typical practical speed range | 150-600 RPM (mechanical limits) | 600-3,000+ RPM | 100-400 RPM |
| Ease of rewinding | Hard — must thread through ring bore | Moderate — slot insertion | Easy — single coil |
| Typical lifespan in service | 20-40 years (heritage examples surviving 100+) | 30-50 years industrial | 5-15 years |
| Cost to build today | High — hand-wound only | Low — slotted lamination standard | Very low — hobbyist scale |
| Best application fit | Museum, education, restoration | All modern DC motors and generators | Toy motors, demonstration only |
Frequently Asked Questions About Ring Armature
The single most common cause is that the conductors threaded through the ring bore are being counted as active in your Z value. They aren't — they sit inside the iron and cut almost no flux. Recount Z using only the outer-face conductors and the prediction will line up.
Second most common: leakage flux. On hand-wound machines the air gap between pole face and ring is rarely uniform around the circumference. A 0.5 mm gap variation around a 100 mm ring drops effective Φ by 8% to 12%. Feeler-gauge the gap at four positions and shim the worst pole.
If the machine is going to be exhibited as historically authentic — running on demonstration days at a museum, filmed for documentary work, or accessioned as a working artefact — rebuild the ring. Visitors and curators can tell the difference instantly when the commutator is visible.
If the machine just needs to function as a power source and authenticity is secondary (back-of-house lab supply, working prop with the armature hidden), a modern drum armature in the same frame gives you 2× the power density and eliminates the inner-bore heat problem. Greenfield Village splits this exactly — front-of-house machines stay original, back-of-house support gear gets modernised internals.
Asymmetric brush sparking on a ring almost always points to unequal resistance between parallel paths. The winding is supposed to form symmetrical paths from brush to brush — if one half-ring has a higher-resistance joint (cold solder at a tap, oxidised commutator riser, broken strand) then current redistributes and one brush carries more than its share.
Diagnostic check: with the machine stopped and field de-energised, measure resistance between adjacent commutator segments going all the way around. Every reading should be within 5% of the others. A single high reading locates the bad tap.
EMF scales linearly with speed in theory, but on a ring the practical ceiling is mechanical, not electrical. Hand-wound rings rely on shellac and binding tape to hold the winding against centrifugal force. Above roughly 1.5× the rated speed the inner-bore conductors start lifting outward — they hit the field poles and short the machine.
The signature failure is a sudden bang followed by smoke from a single sector. Once that happens the ring usually has to be completely rewound. If you need more output, raise field flux or add turns, don't overspeed.
Because the inner conductors do no useful work but still carry full armature current and dissipate I²R as heat. The outer conductors generate EMF (which is energy leaving the machine as electrical output) so a portion of their input mechanical energy becomes electricity rather than heat. The inner ones just heat up.
This is also why ring armatures cannot be rated as continuously as drum armatures of the same copper mass — the bore traps heat with no airflow. Limit continuous current to roughly 60% of what the conductor cross-section would suggest from a current-density table, and add forced air through the bore if you have it.
For visibility and teaching clarity, fewer segments are better — students can actually see the commutation events. 8 to 12 segments on a small classroom Gramme replica works well. Ripple will be visible on a scope, which is pedagogically useful.
For a working museum machine that needs to run lights or motors without flicker, you want at least 24 segments, ideally 36 to 48. Below 24 segments on a 4-pole ring, EMF ripple climbs above 5% peak-to-peak and incandescent lamps will visibly pulse at low speeds. Edison's Jumbo used around 60 segments for exactly this reason.
References & Further Reading
- Wikipedia contributors. Armature (electrical). Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
