A TVS suppressor is the silent guard on a power or signal line. It sits idle for years, then takes a hit from a switching spike or a nearby lightning event and shorts itself to save the rest of the board. The trouble starts when the guard falls and nobody notices, or when a freshly soldered replacement burns out within hours of being installed. Both situations need a careful eye, a multimeter, and a short list of questions about the circuit around the part.
What actually happens inside a TVS when it gives up
A transient voltage suppressor handles surges by entering avalanche breakdown for a few microseconds, dumping the excess energy as heat. The silicon junction can do this many times if each event stays inside the rated peak pulse power. Push past that limit, even once, and the part will end its life in one of three ways: a hard short, an open circuit, or a degraded state where leakage current creeps up while the part still looks alive.
The short failure is the most common and, oddly, the most useful one. A shorted TVS keeps the protected line clamped to near zero volts and forces the upstream fuse, polyfuse, or current-limiting resistor to open. Downstream silicon stays cool. The open failure is rarer but more dangerous, since the line keeps running with no protection at all. The degraded mode is the trickiest of the three, because measurements at room temperature can look normal while reverse leakage at operating voltage has climbed tenfold.
Reading the board before reaching for a meter
A good diagnosis starts with the lights on and the iron cold. Pull the board out, tilt it under angled light, and look at every suppressor in the protection chain. Burned components show discoloration or a faint char ring, often with a recognizable smell of overheated epoxy. A cracked package, a lifted lead, or a pad that has gone matte gray instead of shiny silver all point to one part that absorbed too much energy.
Surface-mount SMA, SMB, and SMC TVS packages tend to crack along the body when they fail violently. The crack can be hairline thin and easy to miss without a magnifier. SOT-23 ESD parts often pop a tiny pit in the top of the molded body, sometimes with a black soot ring around it. Through-hole P6KE and 1.5KE bodies can develop a bulge or a vented hole on the end cap. Solder mask near the part may turn brown, and the copper pour next to the pads can be lightly tinted where the surge dumped heat into the plane.
Photos taken at this stage save time later. Two or three close-up shots under a phone macro lens, with the silkscreen reference designators visible, create a record that survives the rework cycle.
The multimeter pass: short, open, or just leaky
After the visual sweep, power off the board and reach for a digital multimeter. Set it to diode test mode rather than plain resistance, because the small forward voltage applied by diode mode tells you more about junction health than an ohm reading.
A healthy unidirectional TVS reads a forward drop of roughly 0.6 to 0.7 volts in one direction and shows OL (overload) in the other. A reading near zero millivolts in both directions confirms a shorted part. An OL reading in both directions means the junction has gone open and the part offers no clamping at all. Either result is a hard fail and the suppressor needs to come off the board.
The degraded case is harder. The diode test may show normal forward drop, yet the part still leaks under operating bias. To catch this, lift one lead, apply the rated stand-off voltage from a bench supply with a current limit set to a few milliamps, and watch the current. A leakage figure that exceeds the datasheet specification by more than a small margin means the part has aged past its useful life and should be swapped, even if it has not failed outright.
Bidirectional TVS parts read OL in both directions in diode mode when they are healthy. A short still reads near zero in both directions, so the rule for diagnosing a short stays the same regardless of polarity.
Why the same socket keeps eating new diodes
Replacing a TVS only to watch the new one fail within minutes, hours, or days is one of the most frustrating troubleshooting loops in electronics repair. The root cause almost always lies outside the suppressor itself. The original part did its job, took a hit, and shorted. Putting a fresh one in without addressing the source of the stress just feeds another sacrificial component to the same hazard.
Several specific causes drive these repeat failures:
- The stand-off voltage is too low for the actual operating rail, so the part sits in low-level conduction whenever the supply runs at the upper end of its tolerance, heats up, and eventually shorts;
- The peak pulse power rating is undersized for the surge environment, meaning the TVS clamps correctly but cannot dissipate the energy of repeated transients;
- There is no series impedance ahead of the suppressor, so all the surge current funnels through one part instead of being shared with an inductor, ferrite, or current-limiting resistor;
- The ground return path is long, narrow, or shares copper with high-current loads, which raises the clamping voltage seen by the protected silicon and forces the TVS to work harder;
- The protected supply has a real fault upstream, such as a faulty regulator, a stuck switching node, or a bad PSU, that pushes the rail above the breakdown voltage for sustained periods;
- A nearby inductive load (relay coil, solenoid, small motor) has no flyback diode of its own, so every switch-off dumps stored energy into the rail and onto the TVS.
Each of these patterns leaves clues. A TVS that runs hot to the touch under normal conditions is telling you the stand-off voltage is wrong or the rail is too high. A TVS that fails after every thunderstorm but works fine between them is sized for too little peak power. A TVS that fails along with the fuse means the protection is working and the upstream source is the real problem.
Designing the connection so it stops happening
Once the immediate failure is fixed, the connection scheme around the suppressor deserves a second look. Effective protection lives at the entry point of a board, as close to the connector or terminal block as possible, with the shortest practical traces between the line, the TVS, and the ground return. Long traces add inductance, and inductance turns a fast transient into a voltage overshoot that the suppressor never sees in time.
A coordinated protection scheme uses more than one element. A gas discharge tube or MOV at the very front handles the brute energy of a direct surge. A series impedance, often a small inductor or a PTC resistor, follows it. The TVS sits at the back of the chain, where it only has to clamp what slips past the first two stages. Sized this way, the suppressor can ride out years of normal surges without breaking a sweat.
Fusing matters too. A TVS that fails short will pull large continuous current from the protected rail if nothing breaks the circuit. A correctly rated fuse, polyfuse, or breaker on the upstream side turns that short into a clean disconnect rather than a fire hazard. The fuse should clear before the suppressor package itself fails open from sustained heating.
Replacing a suppressor without creating new problems
When the bad part is identified and the cause understood, the rework itself is straightforward. Remove the old TVS with a fine-tip iron for SMD parts or with desoldering braid for through-hole bodies. Clean the pads, check that no copper has lifted, and inspect the surrounding components for collateral damage. A blown TVS sometimes takes a nearby ferrite, resistor, or trace with it, and replacing only the obvious victim leaves a hidden defect on the board.
The replacement part should match the original on three parameters: working stand-off voltage (VWM), breakdown voltage (VBR), and peak pulse power (PPP) rating. Substituting a part with a higher PPP rating is usually fine. Substituting one with a lower stand-off voltage is not, because the new part will sit in partial conduction during normal operation. Orientation counts on unidirectional types, where the cathode band must face the protected line.
Solder the new part with enough heat to wet both pads cleanly. A cold joint on a TVS is worse than no TVS at all, since it presents an unreliable path during exactly the millisecond when the part is supposed to conduct hundreds of amps. Reflow the joint if it looks dull, and inspect it under magnification before powering up.
A final check before powering the board
Before the first power-up, measure resistance from the protected rail to ground with the supply disconnected. A healthy installation reads megohms. Anything below a few hundred kilohms means the new TVS is already partially shorted, the pads have a bridge, or another shorted component is hiding on the same rail. Catching this with a meter is far cheaper than catching it with smoke.
After power-up, touch the TVS body briefly with a finger. It should be no warmer than the surrounding parts. A noticeable temperature rise during steady-state operation is the earliest possible warning that the connection scheme still has a problem worth investigating, well before the next surge arrives and the cycle starts over.
Our Telegram Channel
Our YouTube Channel