Reversing contactor wiring looks like a two-contactor swap exercise until you consider what happens in the 50 milliseconds when one contactor has not finished opening and the other starts to close. Two phases of a three-phase supply meeting across a closing contact pair is a bolted short circuit — and the interlock chain, not the operator’s reflexes, is what prevents it.
This guide walks the standard forward-reverse circuit: the power wiring, the electrical interlock that makes wrong sequencing impossible, the mechanical interlock that backstops it, the switching dead time, and the weld-detection logic that no interlock can provide. The device-level differences between reversing, latching, and vacuum types are surveyed in the latching, reversing and vacuum contactors guide; here we stay on wiring.
| Rule | What it prevents | How it is implemented |
|---|---|---|
| Electrical interlock | Both coils energized simultaneously via control logic error | Each contactor’s coil circuit passes through the other contactor’s NC auxiliary contact (cross-coupled 21–22 contacts) |
| Mechanical interlock | Both contacts closing even with a welded coil, spring fault, or manipulated circuit | Factory-fitted interlock kit physically bars both armatures from seating at once |
| Switching dead time | Arc overlap and plugging-current shock on fast direction changes | 0.05–0.1 s timer or deliberate two-press sequence (stop, then reverse) |
| Overload protection on the load side | Thermal damage during plugging or stalled reversal | One overload relay downstream of both contactors, or one per contactor per the starter design |
| Weld detection (where required) | Phase-to-phase fault through a welded pole pair | Voltage-sensing relay or auxiliary feedback verifying a contactor has actually opened before the opposite coil is enabled |
The electrical and mechanical interlocks are not redundant — they defend against different failures. The electrical interlock stops logic errors; the mechanical interlock stops physics. Use both on every reversing starter.
Reversal of a three-phase motor requires swapping exactly two of the three phases. The conventional layout:
Wiring discipline that prevents most field faults:
The classic hardwired forward-reverse control circuit, step by step:
Two refinements that cost one relay and prevent the worst field behavior:
In PLC-controlled systems the cross-coupling becomes software interlock, but the hardwired NC feedback of each contactor’s actual state should still be read as an input and gated in logic — software-only interlock with no physical feedback trusts the output card more than the hardware deserves.
A mechanical interlock kit couples the two armatures so both cannot be seated simultaneously — even if a coil welds, a contact pair welds, or someone bypasses the control circuit. On the CJX2-F frames the interlock is a side-mounted bar between the two contactors; on many frames it is factory-fitted as a matched pair.
| Failure scenario | Electrical interlock alone | With mechanical interlock |
|---|---|---|
| Control logic error (both commanded on) | Prevented — NC chain blocks the second coil | Prevented (redundant) |
| KM1 coil circuit faulted such that KM1 stays energized; operator presses reverse | Reverse coil blocked by KM1’s NC — but if that NC contact itself welded, nothing stops KM2 | KM2 physically cannot seat; armatures bar each other |
| KM1 contact welded after a fault | KM2’s NC feedback is still free (coil side) — the electrical interlock does not know the main contacts are welded | Mechanical bar still blocks KM2 — partial protection, and the jam signals the fault |
| Wiring error during maintenance | Depends which wires were touched | Protection independent of wiring |
Read the third row carefully: the mechanical interlock is the only layer that constrains a welded main contact, and even it does not make the situation safe — it turns a potential phase-to-phase explosion into a mechanical jam that the motor will stall against. That is why serious reversing installations add weld detection (next section) instead of relying on the jam.
When a running motor is reversed instantly, two things happen in the transition:
The dead-time options, in rising order of protection:
Both interlocks prevent the unwelded contactor from closing against a healthy one. Neither can detect that a main contact has already welded shut after a fault — and a welded KM1 pole pair across two phases is precisely the phase-to-phase short that the reverse contactor will complete if it closes.
The defense is a per-contactor “actually open” verification before enabling the opposite coil:
Weld detection matters most where reversal is frequent and currents are high (hoists, machine tools, cranes) — exactly the applications where a fault event is statistically likely over the contactor’s life.
| Mistake | Consequence | Correct practice |
|---|---|---|
| Reversing by swapping two motor leads instead of two contactor outputs | One direction wires through the wrong contactor path; interlock protection bypassed for that direction | Do the phase swap at the reverse contactor per the standard layout |
| Skipping the mechanical interlock because “the NC interlock is wired” | No defense against welded contacts or wiring errors | Fit the matched mechanical kit; verify armature blocking during dead checks |
| Using the contactor’s own NC for both hold and interlock via jumpers | One contact fault defeats both functions silently | Separate contacts for holding (NO 13–14) and interlock (NC 21–22) |
| No dead time between directions | Plugging current on every fast reversal; contact life collapse, coupling shock | Timer or stop-first sequence; AC-4-rated selection for reversing-under-load duty |
| Overload relay only in one contactor’s path | One direction runs unprotected | One relay on the common motor path, or matched relays in both paths |
A safe reversing starter is three systems deep: power wiring that swaps two phases cleanly, an electrical interlock that makes simultaneous coil energization impossible in the logic, and a mechanical interlock that makes it impossible in steel. Add a deliberate dead time for the transition, and on high-duty applications add weld detection — because the failure that matters most, a welded pole pair, is invisible to both interlocks. For the basic single-contactor circuits this builds on, start with how to wire a contactor; for the starter architecture around it, see contactor for motor starter; and for the matched interlocked pair, see the CJX2-F400 mechanical interlocking contactor or browse the AC contactor category. For verifying contactor behavior in service, the measurement procedures are in the contactor testing guide.
Two layers: an electrical interlock where each coil circuit passes through the other contactor’s NC auxiliary contact, so both cannot energize together; and a mechanical interlock kit that physically bars both armatures from seating at once, protecting against welded contacts and wiring errors.
Reversing two of three phases flips the rotating field direction. Swapping all three leaves the rotation unchanged, and swapping at the motor instead of the contactor bypasses the interlocked switching path — do it at the reverse contactor.
Two supply phases are connected through the closed contact pairs — a phase-to-phase short circuit with fault current limited only by the supply impedance. This is the event the entire interlock chain exists to prevent.
0.05–0.1 s minimum enforced by timer or by a stop-first operating sequence. Applications that reverse under load also need the contactor selected for AC-4 duty, since plugging current approaches locked-rotor levels.
Not by phase swapping — single-phase reversal means swapping the auxiliary/start winding polarity, which is a different circuit. Use the motor manufacturer’s reversing method; the three-phase cross-coupled scheme does not transfer.
No — one overload relay on the common motor path protects both directions. If relays are fitted per-contactor instead, both paths must include them consistently, otherwise one direction runs unprotected.
