Protection Diodes: TVS, Reverse Polarity, Flyback and Input Clamps

Protection diodes can clamp a transient, block unwanted current or provide a path for an inductive load when a switch turns off. Those are different jobs. A TVS across a supply does not automatically provide series reverse-polarity isolation, and a relay flyback diode is not a universal surge protector. Start with the event and current path, then choose a diode whose supported voltage, current and energy behavior fit the circuit.

Last updated: October 11, 2026

Define the problem before choosing the diode

Ask what happens at the protected node: a short positive spike, reversed supply, current continuing from a coil, or an input driven beyond its normal range. Include polarity, source impedance, duration, repetition and the circuit’s permitted exposure. A high peak voltage from a limited source and a sustained wrong adapter can require very different responses.

Next, draw where the protection current goes. Does it return to the connector, recirculate through a coil, enter a supply rail or stop at a series blocking element? The path determines both useful behavior and possible side effects. A clamp can protect one node while disturbing another if its return or receiving rail is not designed for the current.

Toshiba’s TVS fundamentals distinguishes transient-protection functions and electrical parameters. Use that context to classify the problem, rather than selecting every protection position from a familiar diode number or an impressive pulse-voltage label.

Four roles of protection diodes

A shunt transient clamp becomes conductive during the defined overvoltage and diverts current. A series blocking diode carries useful forward current and rejects the supported reverse condition. A flyback diode gives an inductive current a recirculation path. A steering clamp routes an input excursion toward a defined rail or clamp node.

All four affect the surrounding circuit during normal operation as well as during an event. Forward drop can reduce available supply voltage; leakage can shift a sensitive measurement; capacitance can load an interface; a coil recirculation path can change release time. Protection is therefore a functional design problem rather than an extra part appended after the signal and power circuits are complete.

Role	Typical connection	Important limitation
TVS transient clamp	Across the protected node and return	Clamp and pulse survival depend on current and waveform
Reverse-polarity blocking	In a defined series path	Forward loss; no automatic positive-surge clamp
Flyback path	Across the intended inductive current loop	Decay time and transistor stress change with clamp voltage
Input steering clamp	From input toward a rail or reference	Diverted current must be limited and accepted

TVS selection needs a normal-voltage and surge-voltage window

A TVS should remain acceptably inactive over the valid normal node range and limit the defined transient sufficiently for the protected circuit. Stand-off voltage, breakdown voltage and clamping voltage describe different conditions. The part-name voltage is not necessarily the voltage reached during a substantial surge.

Consider a hypothetical rail with 5.5 V maximum normal exposure and a downstream transient limit of 8 V. A 5.0 V stand-off candidate cannot be approved from the nominal 5 V label alone. If another candidate supports normal operation but clamps to 10 V at the relevant current, it also fails the assumed protection requirement. A larger pulse-wattage rating does not fix that voltage mismatch.

TI’s surge-diode selection guide explains the relationship between source, clamp and protected-node requirements. Use the supported clamp at the application current and temperature, then allow for the installed connection effects. If no suitable voltage window exists, change the protection arrangement or circuit requirements instead of assuming a different package will solve it.

Pulse watts do not establish protection against sustained faults

TVS ratings are linked to a specified waveform and initial conditions. For example, the referenced ST SMAJ datasheet lists a 400 W peak capability for its stated 10/1000 µs condition. It does not authorize dissipating 400 W continuously or surviving any pulse with a lower peak power.

Energy depends on voltage and current integrated over time. Repetition matters because the junction may still be hot when another event arrives. A wrong supply that continues delivering power can exceed a transient protector’s supported duty even if the initial response clamps the voltage correctly.

Determine which element limits or interrupts a sustained fault and how it coordinates with the clamp. A fuse, current-limited source or controlled disconnect has a different function. onsemi’s TVS characteristics note provides additional context for the pulse parameters. Validate the defined event and fault response separately rather than describing a circuit as protected without identifying either condition.

Reverse-polarity blocking introduces a forward-loss trade-off

An ordinary series diode can prevent current in the supported reverse direction while passing useful load current in the intended orientation. Its forward voltage reduces the downstream supply and produces heat. The diode must also support the relevant reverse exposure, load waveform and installed temperature.

For an illustrative constant 2 A load and assumed 0.45 V forward drop at that condition, conduction loss is 0.9 W. That result describes the series element’s forward term, not the whole supply’s loss. Check whether the remaining load voltage is adequate at minimum input, and whether the heat path supports operation in the enclosure.

An ideal-diode controller and MOSFET arrangement can offer a different balance. TI’s ideal-diode application note explains relevant architectures. Reverse current blocking, reverse supply protection and bidirectional disconnection are not automatically identical functions; the number and orientation of devices and the controller’s supported behavior determine what the circuit actually does.

A flyback diode redirects inductive current

When a switch interrupts an inductive load, current needs a path while stored energy is released. A correctly oriented diode across the intended loop can conduct that current after turn-off. In normal coil energization it must remain reverse biased. Incorrect orientation can instead create an unwanted conducting path across the supply.

A low-voltage recirculation path can slow current decay. That may affect relay or solenoid release time, so replacing a higher-voltage clamp with a plain diode is not always functionally equivalent. A different clamp arrangement can speed decay while allowing a larger voltage at the transistor. Evaluate both stress and the load’s required mechanical response.

For a hypothetical 40 mH inductance carrying 0.15 A, stored energy is ½LI² = 0.45 mJ. At 20 such releases per second, the initial stored-energy rate would be 9 mW. This does not mean every millijoule is dissipated in the diode: winding resistance, clamp behavior and the complete decay path determine the distribution.

Input clamps need current limitation and a receiving path

A steering diode can divert current when an input moves beyond a rail-related level. A resistor or another current-limiting mechanism must keep the resulting injection within supported limits. The receiving rail must also be able to accept it. An unpowered or non-sinking rail can rise and partially power connected circuitry.

With hypothetical 5.3 V input, a clamp node held at 3.6 V and 1.7 kohm series resistance, the estimated current is 1 mA. The actual clamp voltage depends on the diode and current; it is not automatically rail plus a fixed 0.3 V or 0.7 V. Receiver limits, temperature, resistor tolerance and supply-off states still require assessment.

The small signal diode guide explains loading and leakage effects in more detail. Select a simple clamp only when its normal signal behavior and fault path meet the node’s requirements. Limiting current does not by itself prove compatibility with every integrated circuit input.

ESD protection must preserve the normal interface

Electrostatic discharge is a transient problem with relevant waveform and current-path requirements. A device’s speed adjective or an ordinary diode-mode reading does not establish suitable ESD protection. The supported clamp and survival conditions matter, as do the voltage and current that actually reach the protected pin.

For high-speed data lines, capacitance, topology and leakage can affect normal operation. A rail protector with adequate surge capability may load the signal too much. A low-capacitance interface protector may address a different pulse duty from a large supply surge. Match the exact protection device and network to the intended interface.

The USB ESD protection guide addresses those signal requirements. Do not use a high pulse-wattage family as proof of universal data-line compatibility. Verify both the ordinary interface behavior and the response to the defined disturbance with an appropriate installed test arrangement.

Polarity and layout can defeat an otherwise sensible choice

For a simple unidirectional TVS across a positive DC rail, the cathode is commonly connected to the positive node and the anode to the protection return. The device then operates reverse biased during valid positive exposure. Bidirectional devices have a different negative-voltage response; choose it deliberately for the circuit’s valid range and disturbance polarity.

Package marking, array connections and ordering suffixes must be read from the exact manufacturer drawing. Do not transfer a polarity convention from an unrelated family. Follow the current path on the board after assembly, particularly where several protection junctions share a common connection.

Connection inductance adds voltage when current changes rapidly. A distant TVS or a return routed through a sensitive reference can therefore leave unexpected stress at the load. Measure the protected node, not only the suppressor terminals. The unidirectional TVS explanation helps connect physical orientation with the intended positive and negative response.

Check the defined event, then the post-event condition

Use test equipment and procedures appropriate to the source voltage, waveform and available energy. Record normal operation, relevant supply sequencing and the specific disturbance. Check protected-node exposure, current paths and the response of any limiter or disconnect. A circuit can appear to survive while its actual transient voltage exceeded the intended requirement.

After the event sequence, inspect function and relevant leakage or damage indicators. Protection components can fail in different ways; neither an assumed short-circuit failure nor a normal low-current meter reading should replace examination of the actual circuit. Investigate the source of a failed diode before installing the same part again.

The useful selection record identifies one event, one current path and the exact component behavior that addresses it. Choose protection diodes from those conditions, then coordinate them with current limitation, load behavior and layout. That produces a more reliable design than treating surge clamping, reverse blocking and flyback recirculation as the same protective feature.

Continue through this engineering topic

For the complete selection, design, verification, and troubleshooting path, explore the TVS ESD and Surge Protection Hub.

• Practical guide: TVS vs Zener Diode: Two Similar Parts With Two Very Different Jobs
• Practical guide: Good-Ark TVS Diodes in Plain Language: The Shock Absorbers of Your Circuit
• Practical guide: What Is a TVS Diode? Surge Protection Explained Simply
• Ask Good-Ark technical support about this application