ESD Damage Investigation: Finding Latent Failures

ESD damage is difficult to investigate because the most damaging events are the ones that do not destroy a part. A discharge that passes through a gate oxide leaves a weakened device that passes test, ships, and fails months later. That outcome is called latent failure, and finding its cause means working backwards from a defect signature rather than from a captured event.

Why ESD Damage Is Often Invisible

A destructive discharge leaves an obvious crater and a shorted pin, and the failure analysis is straightforward. A non-destructive discharge leaves damage that is submicroscopic: a pinhole in a thin oxide, a partially melted contact, a shifted threshold. The part still functions within its specification at the test condition, which is why nothing is found at the moment it happens.

The energy involved is small. A human body event of a few hundred volts carries only tens of nanojoules, which is far more than most thin oxides can tolerate but far less than the energy needed to heat a package visibly. The damage is a weakening rather than a break, and it shows up later under electrical or thermal stress.

Latent Failure and Its Signature

Latent failures appear as early-life failures in the field, clustered in time and often in a specific product or lot. The signature is a rise in the failure rate in the first weeks of service without a corresponding manufacturing change, and the failures are usually electrically marginal: leakage above the limit rather than an open or short circuit.

Diagnosing the cluster means looking for a common cause in the handling path of the affected parts. When the failing lot shares a shipping route, an assembly cell or a repair station, that location is the candidate, and the investigation is a comparison of handling conditions rather than a search through the whole plant.

Sources of Charge on an Assembly Line

Charge reaches a part through the human body, through a moving machine part, or through the part itself when its packaging is opened quickly. Operators generate charge by walking, by removing tape, and by sliding boards out of a bag; a board slid across a bench can carry several kilovolts without any sensation. Machinery contributes through belts, rollers and ungrounded covers.

The most damaging source is often the one with the least obvious path: a device at a low voltage relative to its surroundings can be charged and then discharged so quickly that no operator notices. This is why devices with a small body and a large heat sink are among the most vulnerable, since a rapid discharge through a low-resistance path carries high peak current.

<img src="https://www.gopcba.com/wp-content/uploads/2026/08/iStock-1307839840-jpg.webp" alt="Technician testing a wrist strap at an ESD protected workstation” />

Measuring the Workstation

A workstation is verified by measurement, not by appearance. Check the wrist strap with a tester each day: it should present a series resistance near one megohm and a continuous path to ground, and a strap with a broken cord or an intermittent connection will read open. Measure the operator’s system resistance from hand to the common ground point, which should stay well below the specified upper limit.

Measure the bench surface and the floor with a resistance meter using the appropriate electrode and voltage, and record the values against the limits rather than as binary results. Verify soldering irons for tip voltage and tip-to-ground resistance, because an iron with a leaking heating element is a direct path into a sensitive device.

Component Sensitivity Classes

Sensitivity is defined by the model used to test the part. Human body model thresholds are quoted in volts and vary from below 100 volts for the most sensitive devices to several kilovolts for robust ones; charged device model thresholds are lower still and are the reason small, fast parts are a special concern. The class determines the control level required at the workstation.

Working from the class rather than from habit matters because the same handling that is safe for a rugged part will destroy a sensitive one. The list of sensitive part numbers should be maintained and posted where the parts are actually handled, otherwise the protection level is chosen by whoever happens to be working that day.

Handling Rules That Matter

Keep parts in their shielding packaging until the point of use, ground the wrist strap before the bag is opened, and never place an unprotected board on a bench that has not been verified. Use ionizers where insulators cannot be removed, because a neutralized insulator cannot be grounded and will otherwise hold charge indefinitely.

Avoid the routine actions that generate charge: pulling tape from a reel, sliding boards, and handling parts by their leads. Where a station must process a static-generative material, verify the ionizer balance and keep the material away from exposed devices. Training that explains why each rule exists is far more durable than a list of prohibitions.

Investigating a Suspected ESD Event

Start by preserving the evidence and recording the handling history of the affected parts: where they were stored, who opened the packaging, which station assembled them, and what was done to them since. Compare damaged and undamaged units from the same lot, because a difference in handling history is often more informative than the defect itself.

Failure analysis then looks for the electrical signature of an electrical overstress event: leakage paths, altered threshold voltages, and localized damage at an input or output structure. A curve trace on the damaged pins compared with a known-good part is a fast first step, and decapsulation follows when the trace suggests damage inside the die.

Verification and Ongoing Monitoring

Test wrist straps and footwear daily, verify bench and floor surfaces on a documented interval, and check ionizers on the interval the manufacturer specifies. Keep a written record per workstation so a failure can be traced to a date, and treat a workstation that fails a check as out of service until it is corrected rather than logging the result and continuing.

Continuous monitors add value where an operator may forget to connect a strap or where the cord can be pulled loose during work. The monitor alarms immediately rather than at the next audit, which is the difference between a control that exists and a control that works.

Documentation and Training

Write the protection level, the part classes it applies to, and the verification intervals into a single controlled document, and audit against that document. A program that exists only as a set of habits cannot be audited, and an audit that cannot be repeated cannot show whether the process is improving.

Train operators on the mechanisms rather than only the rules, and refresh the training when a new part class enters production. Most ESD failures originate in a routine action that seemed harmless, so understanding why grounding and packaging matter is what keeps the routine compliant.

Related reading: our fabrication notes, board quality and design release notes cover the same ground.

FAQ

What is a latent ESD failure? Damage that weakens a device without destroying it, so the part passes test and fails later under electrical or thermal stress, typically as higher leakage rather than an open or short.

How often should wrist straps be tested? Daily, with the result recorded per workstation. A strap should read near one megohm and maintain a continuous path to ground.

Which line items most often cause ESD damage? Ungrounded soldering irons, unverified bench surfaces, ionizers out of balance and handling inside non-shielding packaging, all of which are found by measurement rather than visual inspection.