Problems found during Design Validation Test (DVT) are inconvenient. The same problems found after tooling, certification, and production fixtures are approved are expensive. A connector that loosens during vibration testing may require a small design change in DVT; discovered at PVT, it can stop the line and invalidate tooling or compliance work.
A Design Validation Test proves that a production-intent electronics design meets its requirements before the factory validates the manufacturing process. This DVT checklist gives PCB and electronics teams a practical sequence for planning samples, stress tests, board-level testing, certification, usability, sourcing, and the final handoff to PVT.
WHAT DVT IS SUPPOSED TO PROVE
DVT proves that a locked, production-intent design meets measurable product requirements under normal use, foreseeable misuse, and specified environmental conditions. It sits after Engineering Validation Test (EVT) and before Production Validation Test (PVT). The required output is not simply a working sample; it is an approved test report showing what was tested, how it was tested, and whether every acceptance criterion passed.
The broader EVT, DVT, and PVT process separates three different questions: does the engineering concept work, does the final design meet its specifications, and can the factory reproduce it consistently? DVT owns the middle question. The PCB, enclosure, firmware, connectors, power supply, test points, and production-intent components should therefore be stable enough to test as one complete system.
THE DVT CHECKLIST
1. Confirm Every PRD Requirement Is Testable
Turn every product requirement into a measurable test with an unambiguous pass/fail result.
Start with the Product Requirements Document (PRD) and build a traceability matrix. Each requirement needs a test method, equipment, sample quantity, test condition, and acceptance limit. “The board must operate from −20°C to 70°C” is testable. “The board must be reliable” is not. Replace vague language with limits for temperature, voltage, current, ingress protection, radio performance, battery life, mechanical load, response time, and expected service life.
For a connected sensor, for example, one requirement might state that the device must maintain a wireless link at a defined distance while operating at the minimum supply voltage and maximum temperature. The matching DVT test must record the firmware version, antenna configuration, chamber temperature, power condition, test duration, and allowed packet-loss rate. This traceability prevents teams from completing many tests while still missing the requirement that matters.
2. Lock the Design Validation Test Plan
Approve the complete DVT test plan before units are assigned to laboratories or test owners.
A useful DVT test plan defines the scope, requirements covered, sample allocation, test sequence, equipment, fixtures, firmware version, owners, schedule, acceptance criteria, and reporting format. It should also state whether each test is destructive and whether the same unit can be reused. A board that has completed vibration testing should not automatically become the sample for a precision calibration study unless the plan explicitly allows it.
Create a sample-allocation matrix so certification units, reliability units, beta units, and engineering spares do not compete for the same hardware. Record calibration requirements for measurement equipment and identify external laboratories early. The plan also needs a deviation process: who can approve a changed method, how failed tests are investigated, and when a design change requires regression testing. Without that control, DVT turns into a collection of disconnected experiments rather than evidence that the design is ready.
3. Build the Right Sample Size
Plan approximately 50–200 production-intent units so testing can run in parallel and expose variation.
One or two hand-tuned prototypes can prove a concept, but they cannot represent normal component tolerances, assembly variation, and different environmental histories. Titoma commonly uses 50–200 units during DVT. The appropriate number depends on product risk, destructive test count, certification needs, beta-test plans, and the confidence required for critical performance claims.
Divide the build before testing starts. For example, allocate separate groups to thermal cycling, vibration and drop, ingress testing, electrical characterization, regulatory pre-compliance, user trials, teardown analysis, and untouched control samples. Keep spares for confirmation tests and failure analysis. All units should use production-intent PCB materials, components, enclosure materials, connectors, and firmware-loading procedures. If a deviation is unavoidable, document it and explain which results the deviation may affect.
4. Run Environmental and Stress Testing
Expose complete units to the temperature, humidity, vibration, impact, and handling conditions expected in their real service life.
Environmental testing should be derived from the actual use case, not copied from another product. Typical DVT work includes high- and low-temperature operation, thermal cycling, damp heat or humidity exposure, vibration, mechanical shock, drop testing, UV exposure, ingress protection, chemical exposure, and accelerated life testing. Industrial electronics mounted near motors need a different profile from a handheld indoor controller.
The PCB-specific goal is to reveal interactions that bench testing misses. Thermal expansion can crack marginal solder joints or stress large BGAs. Humidity can expose contamination, leakage paths, poor sealing, or inadequate conformal coating coverage. Repeated cable movement can loosen a connector or fracture a soldered termination. Record functional performance during and after exposure, not only visible damage. When a unit fails, preserve its history and perform failure analysis before changing the design.
5. Run PCB-Level Functional and In-Circuit Testing
Use complementary PCB inspection and test methods to distinguish assembly defects from real functional failures.
DVT is the right time to prove the future test strategy, not merely inspect boards manually. As explained in Titoma’s guide to PCB testing methods, each method sees a different class of defect:
- AOI (Automated Optical Inspection) quickly finds missing, shifted, reversed, or poorly soldered SMT components, but cannot prove internal electrical performance.
- Flying probe is flexible and requires no dedicated bed-of-nails fixture, making it practical for EVT and early DVT quantities. It is slower and does not replace a system-level functional test.
- ICT (In-Circuit Test) uses test pads and usually a bed-of-nails fixture to check nets and components rapidly. The fixture cost is easier to justify at volume, but DVT must confirm adequate test-point access before the layout is frozen.
- Functional test (FCT) powers the assembly, loads or verifies firmware, exercises interfaces and loads, and confirms that the board performs its intended job.
- Boundary scan or JTAG can provide access to dense digital assemblies where physical probing is limited.
Review the netlist against accessible test points, confirm probe clearance, and identify critical nets with no coverage. Exercise the functional fixture across the supply and temperature limits defined in the PRD. Save measurements by serial number so failures can be correlated with component lots, PCB panels, firmware, and assembly data. Finding missing ICT access during DVT may require a manageable PCB revision; finding it after the production fixture is built can mean expensive fixture rework and line downtime.
6. Validate Certification and Regulatory Compliance
Use production-intent DVT samples for pre-compliance and formal testing required in every target market.
Build a compliance matrix covering the applicable product and market requirements, which may include CE, FCC, UKCA, UL or IEC safety standards, RoHS, REACH, radio approvals, EMC, electrical safety, battery transport, and ingress ratings. The exact list depends on the device, power source, radios, installation environment, and countries of sale.
Certification planning should begin before DVT, but the DVT build is often when parallel laboratory work becomes possible. Reserve sufficient units, cables, power supplies, support equipment, firmware, and technical documentation. Run EMC pre-scans before the formal submission when practical. A late filter, shielding, antenna, plastics, or power-supply change can affect PCB layout, thermal performance, tooling, and previously completed tests. Treat every compliance-driven modification as a controlled design change with defined regression coverage.
7. Test Usability With Real Users
Give representative users production-intent units and observe whether they can install, operate, recover, and maintain the product correctly.
Engineering teams know how the product is supposed to work, which makes them poor substitutes for first-time users. DVT beta units should go to people who match the actual operator, installer, service technician, or customer profile. Give them realistic instructions and tasks rather than a guided demonstration.
For electronics, usability failures often cross mechanical, hardware, and firmware boundaries: connectors are easy to insert incorrectly, status LEDs cannot be interpreted, buttons are difficult to reach, setup fails after a power interruption, cables interfere with mounting, or the device cannot recover from a failed firmware update. Record task success, time, errors, user comments, and physical damage. A product can meet every electrical requirement and still create returns if users cannot operate it reliably.
8. Run Failure Mode and Effects Analysis (FMEA)
Prioritize credible failure modes and make sure the DVT plan tests the highest risks.
Use a Design FMEA to review the product by function, subsystem, and interface. Identify each failure mode, its effect, likely cause, current prevention or detection control, severity, occurrence, and detectability. Focus on risks supported by the architecture and use environment: cracked solder joints under cycling, connector wear, ESD damage, moisture ingress, sensor drift, overheating, reversed connections, corrupted memory, or hardware–firmware timing faults.
FMEA is useful only when it changes action. High-priority items should lead to a design control, a specific test, improved diagnostics, a supplier control, or a service procedure. Update the analysis when DVT discovers an unexpected failure. The completed DVT report should show how critical failure modes were addressed rather than presenting FMEA as a document created only for a review meeting.
9. Validate Assembly and Serviceability
Confirm that production-intent parts can be assembled, inspected, opened, repaired, and reassembled without damage or ambiguity.
Run a structured assembly review with the contract manufacturer and service team. Check component clearances, polarity markings, cable routing, connector access, screw access, torque requirements, adhesive application, thermal interfaces, ESD controls, programming access, calibration steps, and the order of assembly. Time representative operations and document any special tools or skills.
Serviceability matters even when board-level repair is not planned. A technician may still need to replace a battery, display, cable, sealed module, or complete PCBA. Confirm that fasteners survive the specified number of service cycles, seals can be replaced correctly, and the unit returns to its safety or ingress state after reassembly. Apply the principles in Titoma’s electronics DFM guide before the design reaches PVT, when tooling and line processes are harder to change.
10. Get Supplier Sign-Off on Design-Locked Components
Confirm that every critical component and custom part can be supplied at the required specification, volume, quality, and lead time.
The DVT bill of materials should identify approved manufacturer part numbers, alternates, lifecycle status, minimum order quantities, lead times, tooling ownership, and inspection requirements. Give special attention to long-lead ICs, custom magnetics, displays, batteries, connectors, antennas, molded parts, and components that cannot be substituted without PCB or firmware changes.
Supplier sign-off is more than receiving a quotation. Confirm specifications, tolerances, drawings, cosmetic limits, test data, packaging, traceability, and change-notification expectations. Where a second source is planned, test it rather than assuming equivalence. Capture component and material lots for the DVT build. A product that passes DVT with parts that cannot be purchased consistently is not ready for PVT.
11. Document Everything in a DVT Report
Create a controlled DVT report that links every requirement, sample, test result, failure, corrective action, and approval.
The report should include the product revision, BOM, PCB and mechanical revisions, firmware version, unit serial numbers, component deviations, test equipment, calibration status, test methods, raw data locations, photographs, pass/fail results, failure analysis, corrective actions, and regression results. Summarize open issues with an owner, risk, and due date.
A failure is not closed because a later unit passed. Document the root cause, the implemented change, and evidence that the correction solved the problem without creating a new one. The final approval should identify which requirements passed, which deviations were formally accepted, and whether the design is released to PVT. This report becomes the technical baseline for production fixtures, work instructions, inspection criteria, and the pilot build.
COMMON DVT MISTAKES—AND WHAT THEY ACTUALLY COST
Most DVT overruns are not caused by one difficult laboratory test. They come from unclear acceptance criteria, poor sample control, missing PCB test access, and late supply-chain decisions. The table below shows where common mistakes normally surface.
| Mistake | When It Surfaces | Cost Impact |
|---|---|---|
| No formal DVT test plan | At DVT start | Scope creep, sample conflicts, missed requirements, and repeated tests |
| Skipping real-user testing | Beta release or early production | Mechanical or firmware redesign after tooling and documentation are nearly complete |
| Testing only nominal conditions | Certification or field use | Compliance failure, returns, warranty cost, and reputation damage |
| Missing ICT coverage on new nets | PVT fixture debug | Bed-of-nails fixture rework, PCB revision, reduced fault coverage, or line stoppage |
| No supplier sign-off | PVT purchasing or ramp | Part substitutions, tooling changes, recertification, and schedule delays |
WHEN IS THE DESIGN READY TO MOVE FROM DVT TO PVT?
A design is ready for PVT when every critical requirement has objective evidence, failures have documented root causes and corrective actions, regression tests are complete, compliance work is passed or on an approved path, production test access is confirmed, and suppliers have signed off on the locked parts. Minor open items need named owners and an explicit risk acceptance; they cannot be hidden in meeting notes.
DVT typically needs a minimum allocation of about eight weeks, but the correct schedule depends on laboratory lead times, environmental test duration, failure investigation, and design iterations. Moving early does not remove unfinished work—it transfers that work to PVT, where fixtures, tooling, materials, and production schedules make every change more expensive.
If your team needs a DVT partner who can connect product requirements to PCB test coverage, production-intent builds, certification, and the future factory line, talk to Titoma. The goal is not simply to complete a checklist; it is to enter PVT with evidence that the design deserves to scale.
