When TVS Protection Fails: Layout, Rating, and Repetition Mistakes

A failed TVS is usually the protection telling you something: the rating was wrong for the waveform, the placement let the inductance win, or the pulse count exceeded the part’s capability. This guide covers the failure modes, the three classic mistakes, and the redesign checklist.

(Last modified date: August 31, 2026)

TVS Failure Diagnosis — Three Signs, Three Causes:
A shorted TVS absorbed more than its rating; a degraded TVS drifted after repeated stress; a cracked package followed a violent event or severe thermal cycling. Match the signature to the waveform and the pulse history before redesigning the protection chain.

Failure Modes Specific to TVS

A TVS fails in three signatures: shorted, degraded, or cracked. A short is the classic end-of-life of a TVS that absorbed more than its rating — the die breaks down and conducts both ways. Degraded means the clamping voltage drifted or the leakage rose after repeated stress. A cracked package follows a single violent event or severe thermal cycling.

Inspect safely: de-energize and discharge before handling, and read the signature before testing, following the failure-mode field guide. The signature points to the likely cause, and the history confirms it. A TVS failure also differs from a rectifier’s in one respect: a shorted TVS is often the protection working as designed — it sacrificed itself to clamp the event — and the diagnosis distinguishes a part that failed from overstress from one that failed because it was undersized for the event. The history and the waveform decide which.

Mistake 1: Wrong Rating or Waveform Basis

The first classic mistake is rating the TVS on the wrong basis: a part selected by voltage class alone, or by a power number without the waveform, misses the event it must survive. The three-voltage selection — stand-off for the rail, breakdown clearing normal transients, clamping below the protected limit — is the starting framework, and the power class is read at the actual pulse width, not at the datasheet’s headline waveform.

A TVS that fails short after a real surge was likely undersized for the event’s width and energy; the waveform comparison is the diagnostic tool. The rating mistake also includes the stand-off margin: a TVS with a stand-off too close to the rail conducts during normal operation and fails from continuous stress. The selection review checks both the rail margin and the waveform basis.

Mistake 2: Bad Placement and Long Traces

The second classic mistake is placement: a TVS near the protected IC but far from the connector lets the transient travel through the board before it clamps, and a long ground return adds inductance that turns the clamp into a spike. The layout rules — close to the entry, short and wide ground, small loop — are part of the protection design, and a TVS in the wrong place fails to protect even when its rating is correct.

The placement mistake shows up as damage beyond the TVS: the protected IC fails while the TVS itself survives, because the clamp happened too late. The placement review also checks the ground return: a long or shared return lets the clamp current disturb the logic it is protecting. The dedicated, short return is part of the layout rule.

Mistake 3: Repetitive Pulses and Thermal Buildup

The third classic mistake is the pulse count: a TVS rated for a single event can fail after thousands of smaller pulses, because each pulse heats the die and the heat accumulates. The datasheet’s pulse-power curve is for a single event at a starting temperature; the repetition rate and the duty cycle decide whether the die cools between pulses. A TVS that degrades or shorts after months of operation was likely eating a pulse count, not a single surge.

The thermal side of the repetition is the same arithmetic as any power part: the average pulse power and the thermal resistance set the temperature rise, and the design that ignores the count overdrives the part. Repetition also ages the TVS’s clamping characteristic — breakdown voltage and leakage drift with accumulated stress, so a degraded TVS clamps at a different level than the datasheet promised, silently reducing the protection margin. Periodic re-testing catches the drift.

Redesign Checklist

Re-measure the event — width, peak, and count — at the real node.
Re-select the rating on the waveform basis, with the three-voltage framework.
Re-place the TVS close to the entry with a short ground return.
Re-check the pulse count against the thermal budget.
Re-test with the event reproduced and record the before-and-after.

The redesign checklist closes with the evidence: the measured event, the re-selected part, the placement photo, the pulse-count analysis, and the re-test waveform are filed together, so the next failure starts from the record rather than from memory. The redesign also runs the protection-chain review: the TVS is one layer, and the fuse, the filter, and the layout are the others — a fix that only replaces the TVS leaves the chain’s other weaknesses in place.

The final validation is the event reproduction at the worst condition, repeated enough times to prove the count is inside the new margin. A TVS protection fix is proven by the waveform it holds and the pulses it survives. Also re-check the protected circuit’s margin: the clamp voltage is verified against the protected part’s absolute maximum with the margin the selection framework requires, so the fix protects the actual component, not the concept of protection. Share the failed part’s signature and event history with the protection experts, and confirm the re-selected part against the actual waveform.

Engineering note: the three-mistake classification follows the TVS selection, layout, and surge-rating methods in their dedicated guides; the signatures are possible causes, and the specific event history of the failed part is confirmed before the redesign.

Frequently Asked Questions

What does a shorted TVS mean?

It absorbed more than its rating — the die broke down and conducts both ways. The event’s width and energy were beyond the part’s capability.

Why did the TVS survive but the IC fail?

Likely placement: the TVS was too far from the entry or its ground return was too long, so the clamp happened too late. The layout is part of the protection.

How does a pulse count kill a TVS?

Each pulse heats the die, and if the repetition rate outpaces the cooling, the heat accumulates until the part degrades or shorts — even though no single pulse exceeded the rating.

What is the redesign flow?

Re-measure the event, re-select the rating on the waveform basis, re-place the TVS, re-check the count, and re-test with the event reproduced.

How do I read the signature safely?

De-energize and discharge first, classify the signature before testing, and confirm the event history before the redesign.