Two programmes ship the same charger with the same specification. One ages every unit at full load for a fixed period; the other ages a sample from each shift and keeps the results. Their field failure rates differ by less than the difference in their test costs, and which one is right depends on what the aging is meant to catch.
Charger aging test: three strategies compared
Full burn-in, sampling with a defined plan, and development-only extended testing.
The three differ in what they screen, what they cost per unit and what they require of the process. The variable that separates them is not thoroughness; it is whether the failure they are designed to catch occurs consistently enough for a sample to find it, or rarely enough that every unit has to be exposed.
Aging also has a second output that is often more valuable than the screening itself: the failure data. A strategy that produces units failing on the rack and records what failed gives the design and the process something to act on.
Strategy 1: full burn-in on every unit
Every unit is exposed, and every unit carries the cost.
The unit is loaded at a defined level for a defined period at a stated ambient, and it is measured before and after. This strategy catches weak components, marginal solder joints and thermal issues that would appear in the customer’s first week, and it does so at a cost per unit that is proportional to the rack time. It fits products where a field return is expensive relative to the test, and products where the failure mechanism is distributed rather than clustered.
Strategy 2: sampling with a defined plan
A sample from each build or shift, with the results recorded.
The plan states the sample size, the load, the duration, the ambient and the acceptance criteria, and it is most effective where a defect is associated with a lot, a component or a changeover rather than with individual units. Its weakness is that it cannot screen a single weak unit, so it is a control on the process rather than a filter on the shipment. Where the plan is written, it also produces the trend data that full burn-in rarely provides.
Strategy 3: development-only extended testing
Longer tests on the first builds, and a shortened or absent regime in production.
This is the right shape for a platform whose design has been established: extended testing during development and on the first production build, then a reduced regime that relies on the incoming verification of the critical components and the functional test. It is also the shape that is easiest to under-specify, because the decision to reduce aging in production should follow a finding that the failures are not present, rather than an assumption that they are unlikely.
| Strategy | What it catches | Cost per unit | When it fits |
|---|---|---|---|
| Full burn-in | Weak components, marginal joints, thermal issues | Highest, proportional to rack time | A field return costs more than the test |
| Sampling plan | Lot and shift related defects, process drift | Low | Defects cluster by lot or changeover |
| Development only | Design and first-build weaknesses | Lowest in production | An established platform with stable components |
What the aging plan has to state
Six items, and a plan missing any of them cannot be repeated.
The sample or the population aged; the load during the period; the duration; the ambient temperature; the measurement taken before and after, with its limit; and the action for a unit that fails. Where the plan describes the rack but not the load or the ambient, the results from two builds cannot be compared, and the second build will be judged against a different condition without anyone noticing.
Failure analysis: what to do with the units that fail
The rack failures are the most useful sample the factory will produce.
A unit that fails during aging is a unit that would have failed at the customer, and it is available for examination before the shipment. Opening it, identifying the mechanism and recording it produces the input for the next design or process decision. Where failing units are simply removed and discarded, the rack becomes a filter rather than a control, and the underlying cause moves to the next build unchanged.
Where aging programmes go wrong
Four patterns, and each has a signature.
Aging without a load, which warms the units without exercising them; a duration chosen by the rack capacity rather than by the mechanism; the same ambient used in winter and summer, so the test condition drifts with the season; and failures removed from the rack without a record, so the reported result is the number of units shipped rather than the number that failed. Each is visible in the plan, which is why the plan is worth reading before the purchase order is signed.
FAQ
Is full burn-in always better?
It screens every unit, which is valuable when the failure mechanism is distributed across production. It also adds rack time and cost to every unit, so the decision is a comparison between the cost of the test and the cost of a field return rather than a question of thoroughness. The decision is a comparison between the cost of the test and the cost of a field return rather than a question of thoroughness.
Some factories say aging is not needed for a GaN design. Is that right?
The switching technology does not remove the need to expose thermal and contact weaknesses. What changes is the profile of the failures, and the plan should be set from the mechanisms the design can produce rather than from the technology label. The switching technology does not remove the need to expose thermal and contact weaknesses, so the plan is set from the mechanisms the design can produce.
How long should the aging period be?
Long enough at the chosen load and ambient to expose the mechanism being screened, and short enough for the rack capacity and the schedule. The duration is justified by the failure data from the first builds rather than by a round number. Set the duration from the failure data of the first builds and check it against the rack capacity and the schedule rather than choosing a round number.
What should the aging record contain?
The units or lot aged, the load, the duration, the ambient, the measurement before and after with its limit, and the disposition of any failures. Without the ambient and the load the record cannot be compared with a later build. Record the units or lot aged, the load, the duration, the ambient, the measurement before and after with its limit, and the disposition of any failures.
Next step
Send the configuration, the expected annual quantity and the channel to Wecent Group through the contact page. The reply proposes an aging strategy for the product, states what the plan would contain and what the rack time adds to the unit cost. Related reading: our charger factory, charger manufacturing services and the factory floor. External references: IEC standards, IEC Webstore, UL standards and certification, TUV testing services, Intertek testing services and IQNet certification network.
{“@context”:”https://schema.org”,”@type”:”FAQPage”,”mainEntity”:[{“@type”:”Question”,”name”:”Is full burn-in always better?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”It screens every unit, which is valuable when the failure mechanism is distributed across production. It also adds rack time and cost to every unit, so the decision is a comparison between the cost of the test and the cost of a field return rather than a question of thoroughness.”}},{“@type”:”Question”,”name”:”Some factories say aging is not needed for a GaN design. Is that right?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”The switching technology does not remove the need to expose thermal and contact weaknesses. What changes is the profile of the failures, and the plan should be set from the mechanisms the design can produce rather than from the technology label.”}},{“@type”:”Question”,”name”:”How long should the aging period be?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”Long enough at the chosen load and ambient to expose the mechanism being screened, and short enough for the rack capacity and the schedule. The duration is justified by the failure data from the first builds rather than by a round number.”}},{“@type”:”Question”,”name”:”What should the aging record contain?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”The units or lot aged, the load, the duration, the ambient, the measurement before and after with its limit, and the disposition of any failures. Without the ambient and the load the record cannot be compared with a later build.”}}]}