Electrical Overstress and Latch-up Investigation

A device that fails with no sign of mechanical damage has usually been taken beyond its ratings, and the question is which rating and for how long. The damage is often invisible, which is why the investigation begins with the circuit rather than with the package.

What Electrical Overstress Is

Overstress is any condition outside the absolute maximum ratings, including a voltage above the rating, a current above the rating, or a rate of change that the device cannot follow.

The failure can be immediate, as with a gate oxide rupture, or cumulative, as with a junction that degrades over many events. The cumulative form is the harder to find because the device passes every test until it does not. Our reliability notes describe how the margin is assessed.

Common Sources in an Assembly

Inductive loads and long cables store energy that returns to the device when the switch opens. The transient can exceed the rating even though the steady state current is well inside it.

A hot plug event charges a capacitor through a low impedance and the inrush can exceed the current rating. A ground that is lifted during a test or a repair is another path that puts the supply across the wrong pins. Our design release notes describe where the protection is recorded.

Latch-up

Latch-up is a parasitic path that turns on and stays on, drawing a current that the device is not designed to carry. It can be triggered by an overvoltage on an input, by a supply that rises out of order, or by a transient.

Once latched, the device draws current until the supply is removed or the device destroys itself. The current path is internal, so the external circuit may show only a rise in supply current.

Evidence and Limits

The evidence is the supply current, the signal levels at the pins and the state of the device before and after the event. All of it should be captured at the moment of failure, not reconstructed afterwards.

The oscilloscope must be set to capture the event, with the bandwidth and the probe arranged for the rates involved. A measurement made with a standard probe at low bandwidth will miss the event that caused the failure. Our probe notes describe how an internal net is measured.

Protection Measures

The measures are a series element, a clamp, a snubber or a change to the sequence. A series resistor or inductor limits the current, a clamp defines the maximum voltage, and a snubber absorbs the energy from an inductive load.

Each measure has a cost in signal integrity, in space or in efficiency, so the choice follows the transient that has to be handled. Our ferrite notes describe a component that is often used in this role.

Reproducing the Event

The event should be reproduced on a sacrificial assembly before the protection is finalised. The conditions to reproduce are the temperature extremes, the supply tolerance and the load, because the rating is a limit at a condition.

The reproduction should be repeated to show whether the failure is deterministic or statistical, since the two lead to different conclusions.

Corrective Action

The action is the protection change plus a verification that the same event no longer damages the device. The verification is the reproduced test with the protection in place, performed at the worst case condition.

Where the protection cannot be added, the rating of the device is the other option, and the two should be compared on cost rather than on preference. Our derating notes describe how the margin is set.

Process Control and Verification

Where the requirement is not written down, the shop supplies its own default, and the default is chosen for the process rather than for the design. The measurements that matter are the repeatable ones: conductor width and spacing, annular ring, finished hole size, plating thickness and surface finish are all verifiable on a coupon that travels with the panel.

Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage. Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch.

Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end.

Checks Before Release

The checks that matter are the ones performed on the product rather than on a sample kept for the purpose, because a coupon that travels with the panel is the only evidence about that panel. Where a requirement can be measured, it should be measured at the point of manufacture and recorded against the board or the lot it applies to.

A parameter that is set once and never re verified drifts, and the drift is usually discovered by a defect rather than by the record. The tooling, the material and the profile form one system, and a change to any of them should be assessed against the other two before it is released.

Where the process window is narrow, the measurement resolution has to be better than the window, or the data cannot distinguish a good part from a marginal one. A record that identifies the operator, the date and the settings is worth more than a record that identifies only the result.

FAQ

Can overstress be detected by a functional test? Not reliably. A device that has been degraded may pass every test until the margin that remains is used up.

Is a clamp enough on its own? It defines the voltage, and the energy still has to go somewhere. The clamp, the series element and the layout path should be designed together.

What does gopcb provide for overstress investigation? We provide supply and signal capture at the moment of failure, transient assessment of inductive and hot plug events, latch-up review with the supply and pin conditions, protection selection with its cost, reproduced testing on a sacrificial assembly at the worst case condition, and verification that the corrective action holds.