TVS Diode Selection Guide: Working Voltage, Clamping, and Pulse Rating

A TVS with the right voltage printed in its name can still let the protected circuit see too much voltage. A higher-power replacement may survive the pulse while the downstream IC fails. Those outcomes are consistent when the original selection confused working voltage with clamping voltage.

This guide approaches selection through those failure checks. Start with what the circuit must tolerate, then work backward through the source, protection path, pulse, and installation. The objective is a supported protection window, not just a component that survives by itself.

TVS diode selection guide: the essential checks

Select a TVS by matching its working voltage to the highest normal node voltage, then checking clamping at the actual pulse current against the protected circuit’s transient limits. Verify waveform, temperature derating, polarity, leakage, capacitance, and layout. A peak-power label or breakdown voltage alone cannot establish protection of the assembled system.

If the basic mechanism is unfamiliar, begin with how a TVS diode provides surge protection. Here the focus is diagnosing whether the proposed choice meets both normal-operation and transient requirements.

Failure check one: does normal operation already approach conduction?

List the highest normal voltage at the protected node, including supply tolerance, regulation, expected ripple, and valid operating modes. Distinguish those states from faults that the protection system is intended to interrupt or absorb. A nominal supply label is not the complete normal-voltage range.

Reverse working or standoff voltage identifies a specified blocking operating point, usually with a leakage condition. Breakdown voltage is measured at a stated test current. Clamping voltage is associated with a larger pulse current and waveform. These three entries describe different points on the device behavior.

TI’s How to select a Surge Diode explains those selection parameters and the source/pulse context. Use the terms as the selected manufacturer defines them. A name ending in “12” does not promise that a protected pin will never exceed 12 V.

Also check leakage at temperature against the circuit budget. A rail can tolerate a small bleed current that would substantially disturb a high-impedance sensor node.

Failure check two: is there a usable voltage window?

A candidate must remain acceptable at the highest normal voltage and keep the transient at the receiving circuit within its supported limits. Include tolerance and the actual pulse current. The protected component’s absolute maximum is a stress boundary, not a target for repetitive operation or a guarantee of normal function there.

If the two constraints cannot be satisfied with one TVS, do not force the choice by lowering standoff until leakage becomes unacceptable or raising it until clamping becomes ineffective. Revisit source impedance, current limiting, filtering, active protection, or the protected circuit’s voltage capability.

For a source-specific example, the Vishay SMBJ family datasheet lists the following values for SMBJ12A. They are not a recommendation for every nominal 12 V system.

One device row shows three different voltages

Parameter	Referenced SMBJ12A entry	Meaning for review
Working standoff voltage	12 V	Compare highest normal voltage and leakage
Breakdown range	13.3–14.7 V at 1 mA	Low-current breakdown test, not full-surge clamping
Maximum clamping voltage	19.9 V at 30.2 A	Specified pulse-current condition; waveform notes apply

Failure check three: did you calculate the current the TVS receives?

A surge source has a waveform and impedance. The open-circuit surge voltage is not the same as the voltage across a conducting TVS. The protection device and source interact to determine current. Include wiring, deliberate series impedance, and other connected paths where relevant.

As an illustrative static approximation, assume a 100 V source pulse, 10 Ω source resistance, and an assumed 20 V clamp. Current would be approximately (100 − 20) / 10 = 8 A. This is a hypothetical operating-point estimate, not a standards generator model or a measured TVS result.

The assumed clamp itself depends on current, so a more realistic calculation iterates with suitable device data or a justified model. Check the pulse shape and thermal conditions rather than multiplying nominal source voltage by a guessed current.

Different fault sources need different accounts. Cable inductance interrupted by a switch, an external surge generator, and a charged capacitor can deliver energy differently. Name the source before borrowing a calculation.

Failure check four: does the pulse rating describe this event?

Peak pulse power is conditional on a waveform, duration, temperature, and mounting arrangement. A rating for one waveform cannot be transferred unchanged to a longer pulse or sustained overvoltage. Read the pulse-power-versus-duration and derating information provided for the exact part.

The referenced SMBJ source specifies its peak-power entry with a 10/1000 µs waveform and supporting notes. The short headline does not describe continuous dissipation. Keep those conditions with any comparison to another family.

For a real event, consider instantaneous power and its time integral, alongside permitted current and thermal behavior. Two pulses with equal energy can still have different peak stress and heat distribution. Repeated pulses also need an appropriate cooling interval and repetition assessment.

A sustained abnormal supply is a separate system problem. A shunt device may continue drawing current until another element disconnects the source or a component fails. Review the interruption or current-limiting scheme instead of assuming a larger TVS will regulate that fault indefinitely.

Consider the consequence of a damaged protection component as well. A shorted shunt can draw sustained source current; an open path can leave the load without the intended clamp. Do not assume one guaranteed failure mode unless the selected product documentation supports it. Coordinate the surrounding protection and inspect the assembly after an event according to defined acceptance criteria.

Failure check five: did polarity, capacitance or the package change the job?

A unidirectional TVS commonly uses reverse breakdown for one polarity and forward conduction for the other. A bidirectional type has a different two-polarity characteristic. Match the intended signal or rail and negative excursions; do not choose solely from a shared positive voltage entry.

For data interfaces, capacitance and signal integrity can rule out a power TVS even when the voltage seems plausible. The USB ESD protection guide covers that separate constraint. A surge-power headline and an ESD rating also represent different test environments.

Package and internal configuration need their own checks. The SMB fast recovery diode guide concerns a rectifier in a familiar package family. Sharing an SMB outline does not make that device a TVS, nor does it establish the same thermal or transient performance.

Failure check six: is the clamp effective at the receiving pins?

Trace the complete surge-current path through the TVS and its return. Connection impedance can add voltage, especially when current changes quickly. The voltage at the TVS body need not equal the voltage seen at a protected IC several interconnects away.

TI’s 2025 Analog Design Journal discussion of transient protection layout shows how distributed inductance influences clamping. Short, appropriate paths and return connections are part of the protection design, rather than an assembly detail to postpone.

For a hypothetical inductive term, 10 nH carrying a current change of 1 A/ns corresponds to 10 V from L × di/dt. This ideal calculation illustrates why a fast event is sensitive to small interconnect inductance. It is not a prediction of every PCB or surge waveform.

Evaluate the protection location relative to the event entry, the receiving device, and the intended return. There is no useful universal placement instruction without that path.

Verification: check the system, not only the diode

Begin with correct identity, polarity, footprint, and unpowered inspection. Then use the specified application test conditions and appropriately rated equipment. Observe voltage at the protected terminals, relevant current, functional behavior, and any post-test leakage or damage criteria.

Measurement bandwidth, common-mode capability, and probe connection affect the result. The Tektronix floating-measurement guidance explains why earth-referenced instruments require care around floating power paths. Never remove protective earth to obtain a convenient reference.

A diode surviving one bench pulse does not establish compliance with an entire immunity requirement. Preserve the generator, waveform, coupling method, source impedance, operating mode, polarity, repetition, and acceptance criteria with the result. When formal testing is required, use the applicable method and edition.

A release checklist that exposes the remaining uncertainty

Keep the normal-voltage range, receiving-circuit limits, source model, actual current estimate, candidate clamp data, pulse conditions, temperature basis, polarity, leakage, capacitance, and layout together. Attach the exact manufacturer part number and document revision.

Use TVS versus Zener diodes if the circuit instead needs a reference or steady shunt function. A TVS selection record should explain the transient it is intended to handle and what limits the remaining fault energy.

If a required hot clamp curve or pulse condition is unavailable, obtain clarification or choose a candidate with adequate evidence. Do not fill the gap with a generic family claim. The approved outcome is a protection path that meets defined circuit conditions and has been verified at the receiving node.

FAQ

Should I choose TVS working voltage from the nominal supply?

Use the highest valid normal voltage at the protected node, including relevant tolerance, ripple, and operating modes. Then check the selected part’s leakage and voltage conditions. Nominal supply voltage alone is incomplete. The candidate must also clamp the actual transient within the receiving circuit’s supported limits, so working voltage is only the first side of the selection window rather than the complete choice.

Is breakdown voltage the voltage my circuit sees during a surge?

Not generally. Breakdown voltage is specified at a stated test current, while surge clamping is characterized at a different pulse current and waveform. The installed circuit also includes connection impedance and a return path. Use appropriate clamping data and source conditions, then check the voltage at the receiving terminals. A breakdown entry alone cannot establish the peak voltage that reaches a protected component.

Does a higher peak-power TVS guarantee better protection?

No. Higher pulse capability may help the TVS survive an event, but protection also depends on working voltage, clamping at actual current, pulse conditions, temperature, polarity, and layout. A device can survive while the downstream circuit receives excessive voltage. Compare the complete protection requirement and validate the installed system. Preserve waveform and derating conditions rather than ranking candidates by a power headline alone.