Why Does a Solid State Relay Fail in the ON State?

A solid state relay (SSR) often fails in the ON state because its output semiconductor has been damaged by a load short circuit, excessive inrush, a voltage surge, an inductive transient, or too much junction heat. The damaged device can become a permanent conductive path, so removing the input command no longer turns the load off.

But "the load stays on" does not automatically prove that the SSR is shorted. A healthy AC SSR has off-state leakage current; a triac or SCR output normally waits for load-current zero before it turns off; input leakage or noise may still be commanding the SSR; and another wire or device may bypass the output. The first job is to separate these conditions safely. The second is to find the stress that caused the event before installing a replacement.

Safety boundary: If continued energization can create dangerous heat, pressure, motion, fire, or process energy, isolate the power under the site's lockout/tagout procedure. Turning the SSR input OFF is not electrical isolation. An ordinary SSR should not be the only device relied upon to remove a hazardous load.

SSR Stuck ON: Quick Diagnosis

Fastest useful distinction: a true failed-short output carries substantial load current independently of the control input. Leakage usually produces a much smaller current, although even a small current can be meaningful to a high-impedance load. Measure behavior against the exact SSR and load specifications; do not diagnose from the brightness of a lamp or from one unloaded voltage reading.

1. What Does "Failed ON" Actually Mean?

An SSR is failed ON when its output continues to conduct enough current to energize the connected load after a valid OFF command. In a true output-short failure, the semiconductor has lost controllability and presents a low-resistance path between its output terminals. The input LED may go dark and the controller may report OFF, but the load circuit remains powered.

OMRON's SSR safety precautions identify output-element short circuit as a principal failure condition and warn that the load may become impossible to switch OFF. The same guidance recommends a contactor, breaker, or other independent load-power cutoff for fail-safe design. This is a system architecture decision: the device and monitoring logic must match the hazard analysis and the required safe state.

The enclosure alone does not reveal the output technology, turn-on mode, leakage, heat-sink requirement, or internal condition. Always identify the complete part number. Photo: Solid state relay by Mike1024, released into the public domain via Wikimedia Commons.

An electromechanical relay creates a visible physical contact gap when it releases. An SSR switches with power semiconductors such as triacs, anti-parallel SCRs, transistors, or MOSFETs. Severe electrical or thermal stress can melt, puncture, or otherwise damage the semiconductor junction and metallization. A conductive channel may remain even though the control side still appears normal.

Do not treat that broad explanation as a forensic conclusion. A cracked case, discoloration, or burned terminal is useful evidence, but a visually clean device can still be shorted. Conversely, a voltage observed across an OFF circuit can be caused by normal leakage and a high-impedance meter. Diagnosis requires the input state, output current, load behavior, wiring path, and manufacturer-approved test method.

2. Output Technology Changes Both Normal Behavior and Failure

The label "solid state relay" describes a function, not one universal output circuit. Selection and fault analysis start with the complete ordering code and circuit diagram.

A conventional AC triac/SCR SSR is especially easy to misapply. Once triggered, it remains latched until load current reaches zero or falls below the device's holding current. OMRON explains why this current-zero behavior prevents an AC-output SSR from switching a DC load OFF. A DC source does not provide a natural current zero every half-cycle.

Zero-cross and random-turn-on are also turn-on functions. They do not mean that an AC SSR turns off immediately when its input changes. Zero-cross models wait for a specified near-zero-voltage region before triggering; random or instantaneous models do not deliberately wait. Both common triac/SCR versions still normally stop conducting at load-current zero.

This distinction matters during troubleshooting. An AC inductive load may keep conducting after the command changes because current lags voltage. A nonlinear or low-power-factor load may create difficult commutation conditions. That is not the same as a semiconductor that remains shorted after the circuit has been safely isolated.

3. The Main Reasons an SSR Fails Short

The output short is usually the result, not the original fault. OMRON's failure-cause guidance lists excessive inrush, load short circuit, inductive reverse voltage, external surge, high ambient temperature, poor heat radiation, and poor connection or soldering among the important causes. A useful failure investigation follows the current, voltage, and heat paths backward.

Overcurrent, inrush, and load short circuits

A steady-state current rating is not an inrush rating. Lamp filaments, transformers, motors, capacitor-input power supplies, infrared heaters, and some heating elements can draw a startup peak far above normal operating current. The duration, phase angle, repetition rate, starting temperature, and load condition determine the semiconductor stress.

An SSR data sheet may state a maximum continuous load current under specific cooling conditions and a separate non-repetitive surge-current value. A non-repetitive value must not be converted into permission for frequent starting pulses. Compare the complete waveform with the manufacturer's load table, surge curve, and repetition limits.

A load short is more severe. A branch circuit breaker may correctly protect the conductor and still operate too slowly to save a semiconductor. A fuse can help only when its voltage rating, interrupting rating, clearing I²t, time-current behavior, and coordination match the exact SSR and prospective fault current. "Fast-acting" on a label is not sufficient engineering evidence.

Voltage surges and inductive energy

Inductive loads store energy. When current is interrupted, a solenoid, valve coil, contactor coil, motor, or transformer can generate a high voltage until that energy finds a path. A line surge, lightning-related event, capacitor-bank switching, or another load on the same supply can also exceed the SSR output's voltage or dv/dt capability.

The suppression method must fit the circuit. A flyback diode is useful for many DC coils but slows current decay and can delay release. A TVS or diode-plus-zener can release energy faster at a higher clamp voltage. An RC snubber or MOV can suit particular AC or DC conditions, but its continuous voltage, energy, repetition, leakage, aging, and safety approvals must be checked. Never copy a component value from another SSR family without calculation and validation.

4. Heat Is Often the Hidden Cause

An SSR produces heat whenever load current flows. For an AC triac/SCR output, a useful first estimate of conduction loss is P ≈ V_ON × I_LOAD. For a MOSFET output, a first estimate is P ≈ I_LOAD² × R_ON. Use the temperature-dependent values and calculation method in the exact data sheet; switching loss, waveform distortion, current sharing, transients, and interface resistance can add more heat.

Heat must pass through the SSR base, interface material, heat sink, and surrounding air. Photo: Relequick solid state relay with heat sink by Relequick relays, unmodified, licensed under CC BY-SA 4.0 via Wikimedia Commons.

The important value is junction temperature, but it is rarely measured directly in a finished panel. Engineers instead model and validate the full thermal path: junction to case, case to interface, interface to heat sink, and heat sink to ambient air. Sensata's SSR overview likewise treats load current, heat sinking, and ambient as connected selection variables.

Littelfuse's SSR installation guidance illustrates why heat-sink choice, a flat unpainted mounting surface, and the specified thermal interface are installation requirements, not optional accessories. Its numerical instructions apply to the named product series; use the manual for the SSR actually installed.

5. Healthy SSR Behavior That Can Look Like Failure

Off-state leakage and snubber current

No practical SSR output is a perfect open circuit. Many AC units include an RC snubber, and the semiconductor itself has specified off-state leakage. A high-impedance meter can show substantial voltage while the available current is very small. A neon indicator, LED lamp, small relay coil, or electronic power supply may glow, hum, pulse, or fail to reset on that current.

OMRON's leakage-current guidance shows that load reset behavior must be evaluated against the SSR's published leakage. The remedy may be a manufacturer-approved bleeder, interface, or lower-leakage output, but any added component changes heat, power consumption, touch current, and circuit behavior. It must be engineered for the exact voltage and load.

Current-zero turn-off and commutation

A triac or SCR does not necessarily stop conducting at the instant the input is removed. It turns off after load current falls below its holding condition. With a resistive AC load, that is near the next current zero. With an inductive, capacitive, rectified, or distorted load, current and voltage are not aligned.

Low power factor and rapid voltage recovery can cause reset or commutation problems in an unsuitable SSR/load combination. OMRON's reset-failure guidance separates input-side leakage/noise, output leakage, and low-power-factor behavior. The example circuits and component values on that page are not universal prescriptions.

Input leakage, noise, and backfeed

A PLC transistor output may pass a small OFF-state current. Long parallel cable runs can capacitively couple noise into the input. Incorrect common wiring, source/sink mismatch, another powered controller, or an indicator circuit can backfeed the SSR. Measure both voltage and current at the physical input terminals. A PLC screen that says OFF proves only the software state, not the electrical state at the device.

External bypass and control logic

Heater panels often contain contactors, manual bypasses, maintenance jumpers, multi-zone feeds, or shared neutrals. A welded contactor, cross-wiring, or alternate feed can energize the load around a healthy SSR. Trace the complete power path from source to load before condemning the SSR.

Important measurement lesson: voltage alone is not state. Confirm control input, output voltage under a suitable load, and actual load current. Compare all three with the exact data sheet and circuit.

6. How to Test an SSR That Will Not Turn OFF

Testing must start with the safe state of the machine, not with the meter. Work on energized mains circuits should be performed only by qualified and authorized personnel using appropriately rated instruments, PPE, procedures, and test points.

1. Make the process safe. Stop production, control heat/motion/pressure, isolate load power, apply lockout/tagout, and discharge stored energy as required. Do not touch output terminals simply because the input indicator is OFF.
2.  
3. Record the installation before changing it. Capture the complete SSR part number, controller output, wiring, load nameplate, heat sink, thermal interface, enclosure location, fuse/breaker, surge components, terminal condition, ambient, and event history.
4. Verify the input is truly OFF. Measure at the SSR input terminals. Investigate PLC leakage, noise, common wiring, wrong polarity, backfeed, and program logic.
5.  
6. Confirm output technology and load match. Identify AC triac/SCR, DC transistor/MOSFET, zero-cross/random mode, input range, output range, minimum load, leakage, and the permitted load category.
7.  
8. Trace bypass paths. Check contactors, manual bypasses, test jumpers, parallel SSRs, shared conductors, and incorrectly connected auxiliary circuits.
9.  
10. Test output behavior by the manufacturer's procedure. A simple resistance or unloaded voltage test is not valid for every SSR. Use a suitable rated test load or controlled test circuit exactly as the manufacturer specifies.
11.  
12. Investigate the stress event. Test the load and cable for a short/insulation fault; capture inrush; inspect suppression and surge protection; review fault-current coordination; and evaluate the stabilized thermal condition.
13.  
14. Correct the cause before replacement. Select the correct SSR, heat sink, interface, protection, suppression, wiring, and independent cutoff. Commission under realistic load, cycle rate, and worst expected enclosure temperature.
15.  

OMRON's output-check method cautions that an ordinary tester may not supply the voltage and current needed to evaluate a triac/thyristor output. Its example load belongs to that manufacturer procedure; it is not a universal field-test value. Follow the manual for the exact series.

Preserve the failed unit if root-cause analysis matters. Label its orientation and terminals; avoid opening or powering it in a way that destroys evidence. Record whether protective devices operated and which alarm occurred first. A single burned output cannot reveal whether the initiating event was a load fault, surge, thermal runaway, or installation error.

7. Prevent Recurrence: Design the Complete SSR System

Replacing a failed SSR with the same headline current rating is not corrective action. Prevention requires control of load stress, heat, transients, fault energy, installation quality, and the consequence of a failed-ON state.

Choose protection from coordination data

A correctly coordinated fast semiconductor fuse may limit fault energy enough to protect the SSR or contain the fault as the manufacturer defines. It does not eliminate the need for feeder protection, thermal design, surge suppression, or a safe-state circuit. Never select a fuse by nominal load current alone; include normal inrush, fault current, voltage, interruption capability, temperature, and upstream/downstream selectivity.

Make fail-ON observable

Where the consequence justifies it, compare the controller's OFF command with measured load current or temperature. A mismatch can trigger the independent cutoff and a latched fault requiring investigation. Monitoring hardware can also detect an open load, failed heater, fan loss, or abnormal thermal rise. Fault detection must be tested, not merely present in the schematic.

Separate process control from hazard removal

An SSR is excellent for frequent, silent switching, but the same semiconductor output can fail short. A separate contactor, breaker, disconnect, thermal cutoff, or engineered safety function may be required to remove energy. The choice, diagnostic coverage, reset policy, and proof-test interval come from the machine/process risk assessment and applicable standards.

IEC 60947-4-3:2020 provides a relevant scope for semiconductor controllers and contactors for AC non-motor loads. It is not a substitute for the end equipment's product standard, installation rules, or hazard analysis.

8. Load-Specific SSR Selection Priorities

A generic request such as "40 A SSR" omits the information that determines survival. Even continuous current is conditional on ambient, heat sink, mounting, and waveform. For QIANJI product-range review, start with the verified Solid State Relay category, then compare the complete candidate data sheet with the application rather than selecting from the printed current alone.

9. RFQ Checklist After an SSR Fail-ON Event

Send enough evidence for the supplier to reproduce the selection decision. A failure investigation is much stronger when the failed part and the application data travel together.

• SSR identification: manufacturer, full part number, input/output circuit, zero-cross or random turn-on, lot/date code, data-sheet revision, quantity in service, and failure count.
• Load: type, nameplate, AC/DC voltage and frequency, continuous current, measured or published inrush waveform, power factor or L/R, duty cycle, switching rate, and normal/abnormal operating sequence.
• Supply and fault level: nominal range, transient history, grounding, prospective short-circuit current, upstream protection, shared high-energy loads, and SPD arrangement.
• Thermal installation: heat-sink model, thermal interface, mounting method/torque, panel material, spacing, enclosure ambient, airflow, fan status, neighboring heat sources, and stabilized measurements.
• Protection: exact fuse/breaker part numbers, ratings and coordination evidence; MOV, TVS, diode, RC snubber, or SPD part numbers and locations.
• Control input: PLC/output-module model, source/sink wiring, ON and OFF voltage/current measured at the SSR, cable routing, switching frequency, and any input indicator or interface.
• Failure evidence: timestamp, process step, alarms, protective-device operation, photos, thermal image if available, waveform/event log, load/cable test, and isolated SSR test result.
• Safe-state requirement: consequence of continued energization, independent cutoff, feedback, reset policy, proof test, applicable standard, and required certifications.

QIANJI can review a candidate only against the evidence provided. Submit the load nameplate, waveform or inrush data, wiring diagram, protection, heat-sink layout, ambient, and failure record through the verified QIANJI inquiry page. Final approval must be based on the current model data sheet and validation in the customer's installed system.

Frequently Asked Questions

Why does an SSR fail ON instead of OFF?

Power-semiconductor damage can leave a conductive path across the output. Load short circuit, excess inrush, surge voltage, inductive transient, or excessive junction temperature can cause this damage. It is a common failure direction for many industrial SSRs, but the exact failure must still be confirmed.

How can I tell whether an SSR is truly shorted?

First isolate power safely and verify that the input is electrically OFF. Trace external bypass paths. Then use the exact manufacturer's test procedure with a suitable rated load or controlled circuit. An unloaded high-impedance voltage reading is not conclusive because a healthy AC SSR can leak current.

Can a healthy SSR pass current when it is OFF?

Yes. Semiconductor leakage and an internal snubber can pass a specified off-state current. That current may make a sensitive LED, small coil, or electronic input react. Compare the SSR leakage rating with the load's OFF, release, holding, and reset requirements.

Why will an AC SSR not turn off a DC load?

A common AC SSR uses a triac or SCR output that relies on load-current zero to unlatch. DC current does not naturally cross zero, so the output can remain conducting after input removal. Use a correctly rated DC transistor/MOSFET SSR or another DC switching device.

Does buying a higher-current SSR prevent failure?

Not by itself. A larger nominal current does not correct the wrong AC/DC output, insufficient voltage or surge rating, a load short, poor thermal mounting, excessive ambient, unsuitable turn-on mode, weak fuse coordination, or missing transient suppression.

Can a fuse protect an SSR?

A correctly coordinated protective device can limit fault energy, but a generic fuse does not prove semiconductor protection. Use the SSR manufacturer's coordination data or a qualified protection study covering fuse type, I²t, voltage, interrupting rating, available fault current, inrush, and upstream/downstream devices.

Should an SSR always be paired with a contactor?

Not for every harmless load, but an independent means of removing power is needed when the risk assessment says a fail-ON condition cannot be tolerated. The contactor, breaker, disconnect, or safety function must itself be correctly selected, monitored where necessary, and integrated with safe reset behavior.

Final Recommendation

Treat an SSR that will not turn OFF as evidence, not merely as a replaceable bad part. Make the process safe, then distinguish a true shorted output from normal leakage, current-zero behavior, input triggering, and external bypass wiring. If the output is damaged, trace the initiating energy: load short, inrush, surge, inductive transient, excess junction temperature, wrong output type, or installation defect.

Do not install the next SSR until the actual load waveform, heat path, protection coordination, suppression, wiring, and safe-state response have been documented and corrected. Where continued energization can harm people, equipment, or the process, provide an independent means to remove load power and validate the full fault response.

Technical References

• OMRON: Solid State Relay Safety Precautions - output short-circuit failure, heat, stored voltage, and independent load cutoff.
• OMRON: Main Causes of SSR Failure and Countermeasures - inrush, inductive voltage, surge, short circuit, heat, and connection causes.
• OMRON: Solid State Relay Technical Guide - troubleshooting and selection reference.
• OMRON: SSR Further Information - load current, surge, heat generation, heat sinking, and mounting context.
• OMRON: How to Check SSR Output Operation - suitable-load testing boundary for triac/thyristor outputs.
• OMRON: SSR Leakage Current and Load Reset - off-state leakage compatibility.
• OMRON: SSR Reset-Failure Causes - input leakage/noise, output leakage, and low-power-factor commutation.
• Littelfuse: SRP1 SSR Installation Instructions - product-specific heat-sink and mounting guidance.
• IEC 60947-4-3:2020 - semiconductor controllers and contactors for AC non-motor loads within the standard's stated scope.