A power supply burn-in test is useful only when the purchaser can identify the units, the approved procedure and the actual result. “Aged” is not a test report, and hours on a bench do not by themselves establish years of service life.
Define the claim before requesting the process
Ask what the supplier means by ageing or burn-in for the exact order. Is the document describing a production screening step, a sample qualification exercise, a functional check, or an accelerated reliability study? Keep those purposes separate. They can involve different populations, procedures and conclusions.
Texas Instruments’ reliability-testing overview describes accelerated stresses and HTOL in the context of semiconductor devices. It does not establish a JEDEC burn-in requirement, temperature or duration for a complete WEHO power supply. That distinction is a useful reminder: the name of a component-level method cannot substitute for an order-linked finished-product procedure.
Illustrative OEM acceptance station: the exact-source RSP-250-12 is fixed to a metal plate with its genuine guarded terminals connected to a separate fixture. Any display values are illustrative, not a recorded WEHO measurement. This is a source-referenced equipment arrangement, not a photograph of the WEHO production process or evidence of a completed burn-in result.
Anchor the record to the full delivered model
The accepted specimen here is RSP-250-12, with a 12 V/20 A output marking. The retained complete WEHO RSP-250 specification also lists the corresponding nominal output row. The photographed input marking is 100–240V/2.5A, with 50/60Hz and an original power-factor legend. These literal specimen markings are not silently replaced by broader family operating ranges or test-condition values.
Record the purchase order, full model, hardware or document revision where supplied, and the actual serial or lot identifier used in the inspection. A report for a neighbouring voltage variant or an unidentified family sample does not establish coverage of the received units. Keep the source photograph and the accepted order documents linked to the report.
The RSP series name is also not a license to treat every variant as having the same exact output power. Use the accepted full-model row and its conditions. This article does not infer an actual test load from the 20 A marking, and it does not prescribe a universal percentage of that value.
Request a controlled procedure, not an invented recipe
For the relevant process, obtain its identifier and revision, the intended population, the approved input and load conditions, the environmental conditions, the duration or sequence and the acceptance criteria. If those fields are absent, record them as unresolved. Do not fill them with a familiar temperature or time taken from a semiconductor standard, another supply or an internet example.
The procedure should state what is observed before, during and after the process, and how interruptions or abnormal results are handled. The purchaser needs to know whether the result refers to every delivered unit, a sampled group or an engineering prototype. A generic brochure sentence does not establish that distinction.
Any work involving mains, exposed terminals, elevated temperatures or test loads needs an approved safe arrangement and qualified personnel. The documentary checklist below is not an instruction to run the supply beyond accepted limits or to devise a hazardous improvised test.
A different dry end-of-line fixture uses a lower-right, screw-mounted RSP specimen and separate interface equipment. The image illustrates physical connection and source identity only; it supplies no test duration, temperature, pass indication or lifetime claim.
Keep the setup reproducible
Record the actual equipment and its identification, the fixture or mounting arrangement, the input source, the load configuration and the observation points used by the approved procedure. Where calibration or equipment status is relevant to that procedure, retain the corresponding evidence. Describe any deviation rather than making the setup appear compliant through a caption or a model name.
For this specimen, the complete mechanical drawing gives a 215 mm front-to-rear length, 115 mm terminal-edge width and 30 mm body height. It depicts bottom M4 mounting locations and the associated 3 mm engagement limit, with side M4 details carrying a 5 mm limit. The supply is an enclosed, screw-mounted unit, not a DIN-rail product. Those source facts support the illustrative plate mounting; they do not certify a hidden thread engagement or validate a test fixture.
The authentic front has nine guarded screw clamps. With the blue adjustment component at the left, the grouped output positions precede protective earth, N and L. Preserve that actual anatomy. A wire in a ventilation opening, a fabricated terminal or an improvised mounting foot is a physical error, not a documentation issue that can be corrected by prose.
A useful result includes coverage and disposition
Request the actual tested population or identifiers, the procedure revision, the recorded observations and the decision under the applicable criteria. Include interruptions, rejects and deviations, together with their disposition. A clean summary that omits those records cannot tell the purchaser how the process covered the order.
Where a failed unit is repaired or replaced, keep its identity and the required follow-up process traceable rather than resetting the record to a new unexplained pass. This is a proposed procurement record structure, not a statement that WEHO follows a particular rework or retest policy.
Also distinguish a production acceptance record from reliability monitoring. TI’s authored discussion of testing and screening warns, within its semiconductor scope, that periodic process-monitor samples are not acceptance tests on a specific component group and that extra screening does not replace application qualification. It does not give a WEHO finished-supply lifetime guarantee.
Do not convert a screening record into a life promise
A recorded screening step supports only the conclusion justified by its procedure, coverage and results. Predicting useful life in a particular machine requires its own assumptions and evidence, including the relevant use conditions and failure mechanisms. The number of hours in one process is not automatically an acceleration factor or a warranted field lifetime.
Keep manufacturer process evidence, finished-equipment acceptance and long-term service claims as separate records. This lets engineering and purchasing ask a precise question without overstating what the available report proves.
Häufig gestellte Fragen
Does “100% burn-in” identify the test conditions?
No. Ask for the covered unit identifiers and the order-applicable procedure, conditions, observations and criteria. The phrase alone is incomplete.
Can I use a JEDEC HTOL duration as the power-supply burn-in time?
Not from this article. The cited TI overview concerns semiconductor reliability methods, not an approved finished-supply procedure.
Does passing a screen prove a specified number of service years?
No automatic lifetime claim follows. Retain the scope and assumptions of any separate life assessment or warranty.
Do the application images show WEHO’s factory tests?
No. They are independently reviewed source-referenced illustrations. No factory process, measured load, duration or result is depicted as evidence.
Why keep the full model and lot information?
They link the report to the actual delivered population and prevent a family example from being mistaken for the received units.
Can a photograph replace the actual screening record?
No. It can support physical identity and arrangement, not undocumented coverage, criteria or observations.
