EN 
DE 
ZH 
ES Power Module RMA and Warranty Terms Buyers Should Clarify
A failed power module is expensive enough without a dispute over who pays. Learn which warranty clauses, RMA steps, and root-cause rules to clarify with your IGBT or SiC module supplier before you place the order.
A power module can pass incoming inspection, survive validation, and still become a costly dispute eighteen months later, when a returned unit lands on a failure-analysis bench. Who pays for the replacement? Who decides the root cause? How long will the investigation take? Buyers who ask these questions after a failure often discover the answers were never written down. This guide lists the warranty and RMA (Return Material Authorization) terms that procurement and design teams should clarify with any IGBT or SiC module supplier before the purchase order is signed.
Why Warranty and RMA Terms Deserve Attention Before the PO
Power modules sit at the heart of inverters, chargers, drives, and UPS systems, so the commercial stakes of a single failure are high. Four factors make clear terms essential:
- High unit value and downtime cost. One failed module can idle an entire converter or production line.
- Ambiguous failure signatures. Stress from the application and defects from manufacturing can look almost identical on the bench.
- Long time lags. Wear-out and latent defects may surface years after the date code.
- Layered supply chains. Buying through a distributor or an OEM can change who actually owns the warranty claim.
Warranty Terms to Clarify Before You Order
Vague wording such as “standard warranty applies” is a common source of disputes. Ask for definitions, not adjectives. The table below summarizes the clauses most worth confirming in writing.
| Term | What to Clarify | Why It Matters |
|---|---|---|
| Warranty period and start date | Duration, and whether it runs from shipment, invoice, date code, or commissioning | A start date tied to date code can quietly shorten your coverage |
| Remedy | Replacement, repair, credit, or refund, and who chooses | Determines your real recovery after a confirmed claim |
| Covered failure modes | Which defects and wear-out conditions are included | Prevents arguments over what counts as a “manufacturing defect” |
| Exclusions | Overvoltage, improper mounting, gate-drive faults, ESD, unauthorized disassembly | Defines the boundary between supplier and buyer responsibility |
| Storage and handling limits | Allowed storage time, temperature, humidity, and packaging | Long inventory holding can affect eligibility |
| Traceability | Serial or lot coding, and records retained for a claim | Allows a failure to be tied to a production batch |
| Liability cap | Whether consequential or line-down costs are covered | Often the largest gap between expectation and reality |
| Change notification | Notice of material, process, or die changes | A silent change can invalidate your qualification data |
| Claim ownership | Whether the end customer, OEM, or distributor files the claim | Avoids a claim being rejected on standing |
Two points deserve extra attention. First, confirm whether the warranty is tied to a datasheet-conditioned operating envelope. If it is, the supplier will compare your application conditions against rated limits, and you should be ready to supply that data. Second, ask how batch-level problems are handled. A single failure is a return; a pattern across a lot is a quality event, and the contract should say what happens next.
The RMA Process: What a Clear Procedure Looks Like
A well-run RMA process is documented, time-bound, and evidence-based. Look for these steps:
- Request. You submit part number, quantity, date code or serial, failure symptom, and purchase reference.
- Application data. You provide operating voltage, current, switching frequency, junction or case temperature, cooling method, and any waveforms captured near the event.
- Authorization and shipping. The supplier issues an RMA number and packing instructions, including ESD protection. Returned units should not be disassembled or cleaned.
- Incoming verification. The supplier repeats electrical checks and records whether the failure is confirmed.
- Failure analysis. Where warranted, the supplier performs deeper inspection and documents the root cause in a written report.
- Disposition. You receive a decision (replace, credit, or reject) together with supporting evidence and, for confirmed defects, a corrective action summary.
Ask the supplier to state a target turnaround for each stage, who bears shipping and analysis costs if the claim is rejected, and whether no-fault-found units are handled differently. These details are rarely stated in a quote, yet they shape your experience more than the headline warranty period.
Warranty Claim or Application Stress? How Root Cause Gets Decided
Most disagreements come down to one question: did the module fail because of how it was made, or because of how it was used? The table below shows the evidence typically examined.
| Failure Category | Typical Evidence | Usual Direction of the Decision |
|---|---|---|
| Manufacturing defect (voids, bonding, substrate) | X-ray or acoustic imaging, lot correlation, early-life timing | Generally a supplier matter |
| Thermomechanical wear-out (bond-wire lift-off, solder fatigue) | Temperature-swing history compared with the rated cycling capability | Depends on whether the mission profile stayed within rated limits |
| Electrical overstress (voltage overshoot, unprotected short circuit) | Switching waveforms, burn pattern, DC-link and busbar design | Usually an application matter |
| Gate-drive or protection faults | Gate waveforms, desaturation and soft turn-off settings | Usually an application matter |
| Handling, ESD, or mounting damage | Baseplate flatness, screw torque, thermal interface condition | Usually an application matter |
| Insulation or creepage breakdown | Tracking marks, pollution and altitude conditions | Depends on design and environment |
Evidence matters here because laboratory lifetime models only go so far. Semikron engineers presenting at CIPS 2018 noted that the failure mechanisms reproduced in active power cycling tests account for only a fraction of the failures seen in real applications. In practice, that means supplier and buyer both need real application data to reach a fair conclusion. For background on how thermal cycling drives wear-out, see our guide to Coffin-Manson lifetime prediction for power modules. Insulation-related returns often trace back to design rules covered in our article on creepage and clearance for high-voltage modules.
How to Reduce RMA Risk Before It Happens
The strongest warranty position is a failure that never occurs. These habits lower both the odds of a return and the odds of a dispute:
- Capture waveforms early. Record turn-off overshoot, gate voltage, and short-circuit response during validation, and keep them on file.
- Log by date code. Record which lots went into which production batches so any problem can be traced quickly.
- Verify thermal assumptions. Measure junction and case temperature rather than relying on simulation alone; our overview of junction temperature in power modules covers the main methods.
- Match the driver to the device. Gate drive and protection settings should follow the module’s requirements, which is where an IGBT and SiC gate driver selected with the supplier helps.
- Track device-specific reliability topics. For SiC, read about body diode degradation and forward voltage drift and add post-stress checks to your plan.
- Inspect incoming stock. Verify packaging, labeling, and basic electrical parameters before modules enter long-term storage.
Ask About the Quality System Behind the Warranty
A warranty is only as dependable as the process that supports it. Three questions help separate a documented system from a promise:
- Which quality certifications does the supplier hold, and can you see the certificate? For automotive-related programs, IATF 16949 is the automotive quality management system standard and is a common reference point.
- How are changes communicated? JEDEC publishes a standard for notifying customers of semiconductor product and process changes (JESD46, since superseded by J-STD-046). Ask whether the supplier follows a formal change-notification process.
- What is checked before shipment? Look for specific controls such as void inspection and electrical testing, not general statements about quality.
HIITIO’s About Us page describes a quality system with 23 quality control checkpoints, 100% X-ray void inspection, dynamic and static electrical testing, and HTRB testing. It also cites certifications such as IATF 16949 and compliance with IEC 60747, 60749, and 60068 standards, with production in a 10,000-class cleanroom environment. Use this kind of published detail as a starting point, then ask any supplier, including us, to confirm the scope in writing.
Get Clear Warranty and RMA Terms Before You Commit
Reliable modules and clear paperwork go hand in hand. HIITIO Semiconductor offers IGBT modules such as the HCG200FH120A2D (1200 V, 200 A, 62 mm half-bridge), as well as SiC modules such as the HCS200FH120A2C1 (1200 V, 200 A, 62 mm half-bridge, 175 °C maximum junction temperature). Share your application, voltage class, cooling method, and expected volume with our engineers, and we will recommend a suitable module along with the datasheets and documentation you need for your qualification file. Ask our team to provide warranty and RMA terms in writing together with your quotation, so every clause, from start date to failure-analysis reporting, is agreed before the first shipment. Our technical and customer support is available around the clock, and custom solution requests receive a tailored recommendation within 24 hours. Contact HIITIO to start the conversation.
Frequently Asked Questions
What warranty period should I expect on a power module?
There is no universal figure. Warranty length varies by supplier, product class, and application, and the start date matters as much as the duration. Ask for the period and the start-date definition in writing before you order.
What information should I include in an RMA request?
Include the part number, date code or serial number, quantity, failure symptom, and purchase reference. Add operating conditions such as voltage, current, switching frequency, temperature, and cooling method, plus any waveforms recorded near the failure. The more complete the data, the faster the root-cause decision.
Will opening or modifying a module void the warranty?
Often, yes. Disassembly, cleaning, or repair can destroy evidence the supplier needs for analysis, and many warranties exclude units that have been tampered with. Confirm the exact wording and return suspect units in their failed state.
Who pays for shipping and failure analysis?
This is negotiable and differs between suppliers. Typically, the cost position depends on whether the claim is confirmed. Clarify who covers freight, analysis, and no-fault-found cases before the order is placed.
Should I file the claim with the distributor or the manufacturer?
It depends on your purchase contract. Distributors often manage first-line returns, but root-cause analysis may still involve the manufacturer. Ask which party owns the claim and how information is passed between them.
