How to Dual-Source OBC and DC/DC Power Semiconductors

Dual-source OBC and DC/DC semiconductors by qualifying a second device against the complete power-stage, control, thermal, EMC and manufacturing requirements—not by matching headline datasheet values. The safest approach is to preserve the vehicle interface and qualify the alternate inside the charger or converter assembly, with formal change control and production evidence.

Where is the real single-source risk in an OBC or DC/DC converter?

A bill of material can show two approved distributors while still depending on one semiconductor manufacturer, wafer fab, package family or final-test site. That is not true dual sourcing. Review the complete chain for each high-power MOSFET, IGBT, diode, gate-driver IC, controller, isolated interface and protection device used in the OBC or DC/DC converter.

Risk is highest where the device is custom-marked, allocated, built on a single wafer process, assembled at one site, or subject to long replenishment times. Also check whether the primary supplier controls a unique package, die attach process or firmware-dependent controller. A second distributor may improve logistics but does not remove manufacturer concentration.

Assess exposure against the vehicle programme, not only the current purchase order. Include launch timing, ramp volume, service demand, regional production, engineering-change lead time and the consequences of a line stoppage. A device that is adequate for prototypes may be unsuitable for sustained fleet or OEM production.

What should buyers map before selecting an alternate device?

Start with the installed design baseline. Record electrical ratings, switching frequency, gate-drive voltage, current-sense method, protection thresholds, package footprint, creepage requirements, thermal interface, control-loop assumptions and diagnostic behaviour. For an OBC, include the PFC and isolated DC/DC stages separately; for a DC/DC converter, include both the high-voltage switching stage and low-voltage rectification or synchronous-switching stage.

Then classify each candidate by compatibility. A drop-in candidate shares the relevant pinout, package, thermal path and operating conditions. A functional equivalent may need passive-value, gate-drive or firmware changes while retaining the vehicle electrical interface. A process alternative may preserve performance but require a new PCB, magnetics or enclosure. Only the first category normally supports a credible “no vehicle electrical redesign” claim.

Request controlled technical information before sampling: full electrical specifications, safe-operating-area data, switching waveforms, package drawing, qualification status, lot traceability, manufacturing locations, PCN policy and production-capacity information. A supplier’s automotive qualification is useful evidence, but it does not replace application validation. AEC-Q100 generally addresses ICs and AEC-Q101 discrete semiconductors; the converter assembly still requires system-level evidence.[4][6]

How can an alternate be qualified without changing the vehicle electrical design?

Freeze the vehicle-side requirements first. The alternate should be evaluated inside the existing OBC or DC/DC assembly, with the same HV and LV interfaces, communication behaviour, mounting envelope and vehicle control strategy. If a small internal change is unavoidable, define it as an assembly-level change and verify that it does not alter the vehicle harness, software interface, charging protocol or mechanical installation.

Characterise the device at voltage, current, temperature and switching-frequency corners. Measure conduction loss, switching loss, reverse-recovery or output-capacitance behaviour, gate-charge demand, control-loop response, current sharing and protection timing. Compare waveforms rather than relying on typical values. Differences in parasitics can change overshoot, ringing, EMI and thermal stress even when nominal ratings appear similar.

Test abnormal conditions as well as normal operation: input transients, output short circuit, overload, open-load operation, thermal derating, start-stop cycling, brownout, isolation stress and repeated fault recovery. For OBCs, include charging-state transitions, PFC operation, DC-link behaviour and interaction with the vehicle battery-management system. For DC/DC converters, include wake-up, standby, reverse current and load-step behaviour.

Which evidence should the OEM require from both semiconductor sources?

Require the same evidence package from the primary and alternate. This should include specification revision control, material declarations, manufacturing-site details, lot traceability, process-change controls, reliability data, failure-analysis support and a documented response path for quality incidents. Ask the supplier to identify wafer fabrication, assembly and final-test locations rather than providing only a corporate headquarters address.

Confirm the applicable qualification route for each device. Discrete power semiconductors may be qualified under AEC-Q101, while integrated controllers and gate drivers may use AEC-Q100. Qualification grade, temperature range and test coverage must be checked against the programme requirements; a certificate or supplier statement alone does not establish suitability for the complete converter.

Run supplier audits or equivalent capability reviews covering capacity planning, incoming material control, process control, test coverage, change management and business continuity. Review whether both sources can support the same package, marking, moisture handling and inspection requirements. If the alternate uses a different assembly site, qualify that site explicitly rather than assuming the device remains unchanged.

How should the qualification plan be structured?

Use a staged plan with clear exit criteria. Begin with document screening and cross-reference review, then perform bench characterisation on engineering samples. Proceed to converter-level testing, thermal and EMC evaluation, reliability testing, manufacturing trials and production approval. Each stage should define sample quantities, test conditions, acceptance limits, failure disposition and the authority to release the next stage.

Do not limit validation to the semiconductor laboratory. Build pilot units using the intended PCB process, solder profile, thermal interface material, assembly torque and inspection controls. Check automated optical inspection, X-ray, electrical end-of-line testing and functional diagnostics. A candidate that works on a hand-built board may create a manufacturing defect or test escape in volume.

Where the device includes configuration memory, integrated protection or digital control, verify firmware and diagnostic compatibility. Register maps, timing, fault latching and reset behaviour can differ between controllers. Maintain separate approved part numbers internally where necessary, even if the vehicle-level component designation remains unchanged. This preserves traceability and prevents uncontrolled mixing.

Which sourcing model gives meaningful continuity protection?

After qualification, select an operating model rather than simply listing two suppliers. Options include split allocation, primary-plus-qualified-backup, regional allocation or periodic production rotation. Split allocation provides the strongest evidence that both sources remain capable, but it may increase planning complexity and require tighter lot and configuration control.

Contractual controls should cover capacity visibility, allocation escalation, last-time-buy communication, product-change notification, manufacturing-site changes, counterfeit prevention and failure-analysis response. Define minimum notice periods in the supplier quality agreement, but also establish practical escalation contacts and a decision process for shifting demand.

How can the vehicle programme avoid an unexpected redesign?

Place the alternate qualification requirement in the platform specification before schematic release. The power-stage owner should reserve board, thermal and control margins that accommodate the approved device set. This does not mean designing for every possible component; it means preventing a narrow design that only one supplier can support.

Keep the vehicle electrical interface independent from internal semiconductor selection. The OBC or DC/DC supplier should own the alternate-device matrix, internal drawings, software compatibility assessment, validation report and production-control plan. The OEM should approve the qualification basis and the conditions under which the alternate may be released.

Maintain a live risk register covering lifecycle status, supply concentration, capacity, geopolitical exposure, quality incidents and open technical actions. Review it at design gates and during regular production meetings. Requalification triggers should include a die revision, wafer fab change, package change, assembly-site change, controller firmware change or material substitution—not only a new commercial part number.

Finally, verify the recovery plan with a controlled exercise. Confirm how many approved units can be built with each source, how quickly the ERP and manufacturing instructions can be changed, how traceability will be maintained and which tests are repeated after an emergency allocation shift. Continuity exists only when the alternate can be released under documented production conditions.