Short answer: match the coil to the supply before you swap parts. A coil tagged 24 V, 50/60 Hz is an AC coil; feed it DC and its current-limiting reactance disappears, leaving a few ohms of copper to hold the current. It then draws 5-10 times its sealed current — the worked example below lands at 8× — and the winding cooks in one to three minutes. That is how a bank of three starters dies together minutes after energizing, while a meter on the coil still reads a healthy 25.2 V. A DC coil fed AC is the safer mistake: it hums and pulls in weakly instead of burning. Ten minutes with a multimeter settles it: read the tag, measure coil resistance, measure the supply on both ranges, and watch the armature seal.
Symptoms and what each one points to
Sort the failure pattern before you open anything:
- Several starters burn out within minutes of energizing. The shared control source is the wrong type: AC coils sitting on a DC supply. Check this pattern first, because it kills every identical starter on the panel at once.
- One coil dies over weeks, dark winding, burnt-varnish smell. The armature never seals: dirt, rust or a worn mechanism holds the magnet faces apart.
- Loud steady buzz at 100-120 Hz. A cracked shading ring, dirty pole faces, or the coil running below its voltage floor.
- Chatter, or a starter that drops out when a second one pulls in. Control voltage sagging under load.
- Coil cold, starter dead. Wrong voltage class on the coil, an open coil wire, or a blown control fuse.
First suspect: an AC coil on a DC supply
An AC coil is wound with few turns of heavy wire, so its DC resistance is low — tens of ohms on a small contactor. On AC, inductive reactance rather than resistance limits the current, and that reactance follows the magnetic circuit: highest with the armature seated, lowest with the faces apart. On DC, reactance contributes nothing once the current settles, so the coil runs at Ohm's law on its copper alone.
DC coils are built the other way round: many turns of fine wire, high resistance, current set by V/R whether the armature sits or not. Feed an AC coil from a DC bus and it holds something near its open-armature current forever.
In a modern panel that DC usually starts at a DIN-rail switching supply that also feeds the PLC, mounted a few inches from the starters: see mounting a switching power supply in an enclosure.
Worked example: 20 Ω across a coil, DC behind it
The coil data below are example values for a small contactor. Read the sealed VA and the resistance from your own coil's data line and redo the two sums.
- Three identical fan starters share one control supply; each coil is tagged 24 V, 50/60 Hz.
- Resistance across one coil, control circuit locked out: 20 Ω.
- Sealed current on its rated AC supply: 0.15 A, about 3.6 VA.
- Fed that DC bus, the coil draws I = 24 V ÷ 20 Ω = 1.2 A. That is 1.2 ÷ 0.15 = 8 times its sealed current.
- Heat: P = 24 V × 1.2 A = 28.8 W, against the 3.6 W (24 V × 0.15 A) the winding was built to dissipate — 8 times the design heat.
- At 28.8 W in a small Class F (155 °C) winding, insulation fails in one to three minutes, which matches a set of starters dying together before anyone reaches the panel.
The same coil on its rated AC seals at 0.15 A and dissipates 3.6 W, and it lasts for decades. The supply type is the whole difference.
Second suspect: the armature never seals
On contactor coil data, inrush typically runs 5-10 times the sealed current, because the air gap between armature and magnet lowers the inductance. The example coil pulls in at 1.2 A, 8 times its 0.15 A sealed draw. A coil that never seals keeps drawing that inrush: blocked fully open it cooks as fast as the DC case above, and one that seats late or partly runs hot for weeks before it dies.
You hear this one before you measure it. Energize the starter and listen: a clean click followed by silence is a seal, and a sustained buzz means the faces are not meeting. With the circuit locked out, wipe the pole faces, look for rust, paint or chipping, and move the armature by hand to feel for a sticky mechanism.
A DC coil cannot fail this way. Its current is fixed by resistance, so a blocked armature changes nothing but the pull-in force.
The enclosure contributes here. Dust settling through a vented indoor box packs the pole faces, washdown mist corrodes them, and generator-skid vibration — these starters cool generator sets — works mechanisms loose over a season. Clean the faces at every maintenance window; when the dirt comes back faster than you clean it, that is an enclosure problem, not a coil problem.
The other causes: voltage, shading rings and unknown coils
Voltage. IEC 60947 sets the coil operating band: pick up and hold between 85% and 110% of rated control voltage, which is 20.4 to 26.4 V on this coil. A reading of 25.2 V sits 5% over nameplate, so voltage is not the killer in this pattern. The reading worth logging is under load: a marginal transformer can hold 25 V unloaded and sag to 21 V at the far coil with all three starters sealed in — inside the band, but close to the 20.4 V floor.
Shading rings. A shading ring is a shorted copper loop on the magnet pole face; it holds magnetic force above zero between AC half-cycles. A cracked ring makes the coil buzz at twice supply frequency — 120 Hz on a 60 Hz supply — chipping the faces and cooking the coil. The fix is a new magnet assembly, with the faces cleaned before it goes in.
Unknown coils. Order a coil by its full designation: voltage, AC or DC, and frequency, so the coil you get matches the supply you have. In IEC practice 24 V DC is the standard control voltage, so an order that just says "24 V coil" arrives as a DC coil. A tag reading 50/60 Hz works on either supply frequency, so frequency itself is rarely the fault. Do not guess a dead coil's type from wire color or turns; identify it from the starter frame.
Five checks that settle it
| Check | How | Healthy reading | What it tells you |
|---|---|---|---|
| Coil tag | Read it with power off | "24 V 50/60 Hz" is AC only; "24 V DC" is DC | A tag that only says "24 V" needs the starter frame's coil list |
| Coil resistance | Ohms across the coil terminals | AC coil: tens of ohms. DC coil: hundreds of ohms or more | Low ohms on a DC system means an AC coil on the wrong supply |
| Supply type | True-RMS meter at the coil terminals, DC range then AC range | A DC supply reads about 24 V on DC and near 0 on AC; a transformer secondary reads the reverse | Solid readings on both ranges mean a rectified, unfiltered supply |
| Voltage under load | All starters sealed in, measure at the farthest coil | 20.4 V or more (85% of nameplate) | Below that floor, the transformer or supply is undersized |
| Seal | Energize one starter and listen | One click, then silence | A sustained buzz points to faces apart or a shading ring |
Three of the five checks take under a minute each, and the resistance reading alone separates the top two causes.
The fix: match coils, branch-fuse, size the transformer from VA
Order coils wound for the supply you have: an AC coil on the transformer secondary, a DC coil on the DC bus, or a universal coil — a DC winding with a built-in bridge rectifier — where a supply might be either. Keep starter coils off a DC bus that feeds a PLC; give them their own 24 VAC control transformer and label it at the terminals.
Branch-fuse each coil near its sealed current. The example coil seals at 0.15 A and pulls in at 1.2 A, so a 0.3 A time-delay fuse rides every start and still clears a wrong-supply draw. A time-delay fuse carries 5 times its rating for at least 10 s (UL 248-8), so at 4× it opens in seconds — after the brief pull-in pulse, before the winding cooks. A 4 A group fuse sized for three coils and pilot lights carries a runaway 1.2 A forever; that is how coils burn while the panel fuse holds.
Size the control transformer from the coil's two VA numbers. Continuous rating at or above the sealed total, inrush rating at or above everything that pulls in together: three coils at 3.6 VA sealed and 28.8 VA inrush give 11 VA continuous and 86 VA of simultaneous pull-in, which a 100 VA control transformer covers with margin.
Keep it from coming back
- Label every control source at its terminals and on your door schedule: "24 VAC ONLY" or "24 VDC ONLY". The backplate and DIN rail layout guide shows where the control section and its labeling belong.
- Standardize the plant on one control voltage. IEC practice defaults to 24 V DC; once every panel matches your plant standard, every coil order is unambiguous.
- Put sealed and inrush VA on the panel drawing, so the next transformer or supply is sized from data rather than memory.
- Log a commissioning voltage at the farthest coil with every starter sealed. The floor is 20.4 V (85% of nameplate); anything within 2 V of it is a transformer to upsize at the next shutdown.
When the enclosure is part of the cause
Coils that die of dirt rather than DC point at the box. A starter in a dusty or washdown plant belongs in a gasketed enclosure: our stainless boxes are rated NEMA 4X / IP66 by the manufacturer; not UL listed. Choose 304 stainless for indoor washdown, food and beverage and most outdoor sites, and 316L for salt air, brine, chlorides and chemical areas — custom stainless enclosures covers the grades and the starter layouts we build, delivered in 10 business days.
Heat moves too. The failed coil in the example released 28.8 W into a box sized around 3.6 W per coil, and it aged its neighbors on the way out. Estimating enclosure thermal resistance from the heat load shows how to budget the replacement's watts, and the power supply that trips at 60 C shows what chronic heat does next. For a new panel, size the box from the starter count and clearances using the enclosure sizing guide, or start from sizing an IP65 enclosure for a motor starter and contactor.
Bolt the starter box to a rigid frame rather than the generator skid itself; the vibration that loosens armature mechanisms lives on the skid.
The decision by symptom
| What happened | Cause | Action |
|---|---|---|
| Several starters die within minutes of energizing | AC coils on a DC supply | Fit DC or universal coils, add 0.3 A time-delay branch fuses, label every source |
| One coil dark after weeks of buzz | Armature never seals | Clean the pole faces, free the mechanism, log the seal at each start |
| Loud 100-120 Hz buzz, chipped faces | Cracked shading ring | Replace the magnet assembly; clean the faces first |
| Chatter when a second starter pulls in | Voltage sagging below 20.4 V | Resize the transformer from sealed and inrush VA |
| Coil cold, starter dead | Wrong coil or blown control fuse | Match the coil to the frame and supply, then check the branch fuse |
Run the five checks in order and the cause falls out in one visit. The enclosure upgrade is for plants whose coils keep dying of dirt; the corrected coil order is for everyone else.
Catalog dimensions to compare
Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 12 × 10 × 6 in | 304 stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 14 × 8 in | 304 stainless | 16 x 14 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 20 × 16 × 8 in | 304 stainless | 20 x 16 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Can a DC coil run on an AC supply?
Yes, but poorly. A DC coil has many turns of fine wire, so its high resistance keeps the current safe, and the added AC reactance lowers the current further. You get a loud hum, because a DC coil has no shading ring, plus weak pull-in that can chatter the contacts. It rarely overheats. For a supply that might be either, fit a universal coil: a DC winding with a built-in bridge rectifier that runs on AC or DC at its rated voltage.
How do you tell an AC coil from a DC coil without a tag?
Measure resistance across the coil terminals with the supply locked out. For the same starter frame, the AC coil reads low — about 20 Ω in the worked example — because it is wound with few turns of heavy wire and relies on reactance. The DC coil reads about 150 Ω or more, because resistance is its only current limit. If the reading matches no coil list for the frame, replace the coil rather than guess its type.
Is 25.2 volts too high for a 24 volt coil?
No. IEC 60947 sets the coil operating band at 85% to 110% of rated control voltage — 20.4 to 26.4 V here — so 25.2 V sits 5% over nameplate, comfortably inside. The reading worth chasing is the loaded one: a supply can hold 25.2 V with everything off and sag toward the 20.4 V floor when all three starters seal in. Measure at the farthest coil with every starter energized.
Why did all three starters burn out at the same time?
Because they share one control source, and a wrong-type supply kills every AC coil on it within minutes — supply type varies more than coil quality. One dead coil among healthy twins points instead at that starter: dirty pole faces, a worn mechanism or a cracked shading ring. To shrink the blast radius, feed each panel from its own fused control transformer, so a supply-side error takes out one panel, not the whole row.
What fuse protects a contactor coil?
Branch-fuse each coil near its sealed current, with a time-delay element. The example coil seals at 0.15 A and pulls in at 1.2 A, so a 0.3 A time-delay fuse rides every start and clears a sustained wrong-supply draw in seconds: a time-delay fuse carries 5 times its rating for at least 10 s (UL 248-8). A fast-acting fuse of the same size can nuisance-blow on the pull-in pulse, and a panel-level fuse sized for the group carries the same 1.2 A indefinitely.
