Relay Chatter and Contact Bounce: Causes and Fixes
What causes relay chatter and contact bounce? Contact bounce is the microsecond-to-millisecond mechanical rebound of a relay's armature and contacts during a single operate or release cycle, inherent to every electromechanical relay built to IEC 61810; chatter is a longer, repeated make-break cycle driven by an unstable coil supply, a worn socket connection, or excessive vibration. Left undiagnosed, either one erodes contact material, welds a contact closed under an inductive load, or floods a PLC input with false transitions that the control program logs as separate events. This article separates the two failure modes and works through coil-voltage tolerance, contact and socket wear, suppression-element mismatch, AC coil ripple, vibration, and how to tell them apart with a meter and a scope.
Contact Bounce vs Relay Chatter: Two Different Problems
Contact bounce happens inside every electromechanical relay, on every operation, whether the relay is healthy or failing. When the armature snaps toward the coil pole face, the moving contact strikes the fixed contact, rebounds off it under spring tension, and re-strikes several times before settling. On a typical miniature plug-in relay the whole event is over in under 5 ms. Chatter looks similar on a scope trace but plays out over seconds or minutes: the armature pulls in, drops out, and pulls in again, cycling because something outside the relay — coil voltage, a loose terminal, mechanical shock — keeps pushing it across the pull-in/drop-out threshold.
Confusing the two wastes time on a panel. Swap a relay for bounce and the replacement bounces identically, because bounce is a property of the mechanism, not a defect. Swap a relay for chatter and it may quiet down for a week, then start again once a socket terminal works loose or the supply sags under the next load cycle. Diagnose the cause before reaching for a new part. See our interface and coupling relay engineering guide for how the relay-socket-suppression assembly fits together as a system.
Why Coil Voltage Instability Drives Chatter
A DC coil relay pulls in reliably above roughly 0.8 x rated coil voltage and drops out somewhere between 0.1 and 0.3 x rated voltage. That band is a real datasheet spec, not an approximation you can skip when troubleshooting. Feed the coil a supply that sags under load — a 24 VDC rail shared with several contactors, or a long, thin cable run to a remote panel section — and coil voltage can dip into that drop-out band the moment a neighboring load switches on, releasing the armature, then recovering as the load drops and pulling it back in. Repeat that a few times a second and the relay chatters continuously.
Formula: Relay Pull-In and Drop-Out Voltage — Source: IEC 61810-1, coil voltage tolerance
Vpull-in ≥ 0.8 × Un, Vdrop-out = 0.1–0.3 × Un
| Symbol | Description | Unit |
|---|---|---|
| U_n | Rated coil voltage | V |
| V_pull-in | Minimum voltage at which the armature fully seats and bounce settles | V |
| V_drop-out | Voltage below which the armature releases and the relay opens | V |
What we see in the field: chatter that starts only when a nearby contactor or valve energizes is a supply-sag symptom, not a relay defect. Check coil voltage at the relay terminals under load, with the suspect contactor cycling, not at the power supply output where the reading still looks correct. See relay coil voltages: 24VDC, 110VAC and 230VAC explained for tolerance bands across common ratings.
Worn Contacts, Socket Wear and Loose Terminals
Bounce itself does mechanical work. Every re-strike arcs slightly under load, and over enough cycles that arcing pits the contact surface, raises contact resistance, and eventually welds the contact shut on a high-inrush load like a contactor coil or a lamp filament. A pitted contact bounces longer and harder than a new one, because the mating surfaces no longer meet flat — so bounce and wear feed each other once the process starts.
Socket wear is a separate, often-missed cause. The socket carries the current rating, not just the relay, and a socket contact that has lost spring tension after years of insertion cycles makes an intermittent connection under vibration or thermal cycling — the panel reads as chatter even with a brand-new relay plugged in. A loose screw terminal on the socket does the same thing on the coil side: contact resistance rises, coil voltage drops below pull-in, the armature releases, resistance changes as the joint cools, and the relay pulls back in. Re-torque the terminal and the chatter often disappears without touching the relay at all. This is worth checking before condemning the relay — see how a plug-in relay and socket system works for socket rating and retention details.
Coil Suppression Mismatch and AC Coil Ripple
A DC coil relay driven by a PLC transistor output needs a correctly oriented freewheel diode across the coil. Fit it backwards and the diode conducts on every pulse instead of clamping the collapse spike — the coil never builds full flux, the armature does not seat, and the contact chatters against the pole face instead of closing cleanly. Omit the diode altogether on an inductive coil switched by a transistor output and the transient can damage the driving output stage, which then drives the coil erratically until it fails outright.
AC coil relays have their own version of the problem. An AC electromagnet's holding force drops to zero at every current zero-crossing, twice per cycle; a shading ring on the pole face maintains enough residual flux to hold the armature seated between crossings. A worn or cracked shading ring lets the armature buzz at twice line frequency — audible mechanical chatter, not an electrical fault at all, and no amount of coil-voltage troubleshooting fixes it. An undersized RC snubber or varistor on either coil type changes the release time enough to shift a marginal design into chatter under certain load conditions; match the suppression element to the coil, not to whatever was in the parts bin. See relay coil suppression: freewheel diode, RC and varistor for how to size each type correctly.
Vibration, Mounting and Mechanical Causes
A relay mounted near a vibrating source — a motor, a pump, a poorly isolated fan — can chatter purely mechanically, with coil voltage steady and the socket tight. The armature spring and the external vibration frequency interact, and if they land close enough together the vibration adds enough energy to momentarily unseat the armature even with the coil fully energized. This depends on the relay's own mechanical resonance, which is why the same vibration source chatters one relay type and leaves an adjacent one, mounted on the same DIN rail, unaffected. Reorienting the relay, adding rail damping, or moving it away from the vibration source resolves it faster than chasing an electrical cause that is not there.
How to Diagnose Chatter and Bounce in the Panel
Start with the coil, not the contact. Clamp a meter across the coil terminals during the suspect operating cycle — a snapshot reading at rest tells you nothing about a fault that only appears under load. If coil voltage sags into the drop-out band when a neighboring load switches, the cause is supply sag or wiring resistance, not the relay. If coil voltage holds steady and the relay still chatters, move to the socket: re-torque or replace screw terminals, check push-in terminals are fully seated, and try a known-good relay in the same socket. If the fault follows the socket, the socket is the problem; if it follows the relay, inspect the coil suppression element and its orientation before assuming the coil itself is bad.
A storage oscilloscope on the contact voltage separates bounce from chatter definitively: bounce shows a burst of transitions lasting a few milliseconds around a single operate event; chatter shows the same burst pattern recurring every few hundred milliseconds to several seconds, tracking an external trigger such as a neighboring load cycling. Where a scope is not available, a chatter relay is usually audible — a buzz or rapid click rather than a single crisp click — and an ammeter on the coil circuit will show current cycling in step with the noise.
| Criteria | Contact Bounce | Relay Chatter | Contact Weld |
|---|---|---|---|
| Duration | Milliseconds, one event | Recurring, seconds to continuous | Permanent, contact stuck closed |
| Root cause | Normal armature mechanics (IEC 61810-1) | Coil voltage sag, loose terminal, vibration, suppression fault | Repeated arcing from bounce or an inrush overload |
| Typical fix | None required if within spec; select an SSR if the load cannot tolerate any bounce | Re-torque terminals, check coil voltage under load, verify suppression element | Replace relay; re-check contact rating against actual load current |
Frequently Asked Questions
Is contact bounce always a fault?
No. Every electromechanical relay bounces briefly on each operation as a normal property of the moving armature. It only becomes a problem for loads sensitive to rapid make-break cycling, such as certain electronic counters or timers, where a solid-state relay is a better fit than an electromechanical one.
Can I fix chatter by replacing just the relay, or does the socket matter too?
The socket carries the same current and voltage rating as the relay and wears with insertion cycles and terminal torque just like any other terminal block. If chatter persists after a relay swap in the same socket, the socket connection — not the relay — is usually the cause.
Why does a DC coil relay chatter when driven from a PLC transistor output?
A reversed or missing freewheel diode across the coil prevents the coil from building full flux on each pulse, so the armature never fully seats. Check diode polarity first; it is the most common wiring mistake behind DC coil chatter.
Does contact bounce affect a PLC input the same way as chatter?
Both can register as multiple false transitions if the PLC scan time is fast enough to catch the individual bounces or chatter cycles. Debounce logic in the PLC program, or a slower scan on that input, filters bounce; chatter still needs the underlying cause fixed because it recurs indefinitely.
How do I test for chatter without an oscilloscope?
Listen for a buzz or rapid clicking instead of a single crisp click, and clamp a multimeter across the coil to watch for voltage that dips and recovers in step with the noise. An intermittent connection at the socket or coil terminal is the most common cause found this way.
Conclusion
Bounce is a mechanical property of every electromechanical relay and rarely needs a fix on its own. Chatter always has an external driver — coil voltage sag, a loose or worn terminal, a suppression-element mismatch, or vibration — and swapping the relay without finding that driver just delays the next failure. Check coil voltage under load first, then the socket, then the suppression element, before assuming the relay itself is bad. For contact ratings and load-type selection once the fault is resolved, see PLC interface relays: isolating and amplifying I/O, or browse interface and control relays for replacement stock.
