Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC TR 62878-2-2: Electrical Testing Guidelines for Device Embedded Substrates
💡 As electronic packages become more miniaturized and functionally integrated, traditional PCB testing methods are no longer sufficient. IEC TR 62878-2-2 introduces a structured five-level electrical testing framework for substrates with embedded active and passive devices, providing essential guidance for quality assurance in advanced electronic manufacturing.
1. The Challenge of Testing Embedded Devices
Device embedded substrates represent a paradigm shift in electronic packaging. Unlike traditional PCBs where all components are surface-mounted after board fabrication, embedded substrates integrate components within the substrate layers during the lamination process. Resistors, capacitors, inductors, and even semiconductor chips are buried inside the laminate structure. This creates a fundamental testing challenge: conventional flying-probe or bed-of-nails testing can verify interconnect continuity but cannot validate the functionality of devices hidden within the board. An open circuit in an embedded via or a parametric drift in an embedded resistor may go undetected until final system test, causing expensive rework or field failures. The standard defines five test levels to address this challenge comprehensively, from basic interconnect verification through full functional validation.
| Level | Test Scope | Target | Method |
|---|---|---|---|
| 1A | Interconnects not connected to embedded components | Substrate wiring | Open/short test (conventional) |
| 1B | Interconnects connected to embedded components | Component wiring | Open/short with probe access |
| 2A | Active device functionality | Embedded ICs, diodes, transistors | Device function verification under bias |
| 2B | Passive device parameters | Embedded R, L, C | Impedance/LCR measurement |
| 3 | Full system functionality | Complete embedded circuit | Circuit model vs. measured response |
⚠️ Test Level 1A is identical to conventional PCB testing and is sufficient for substrates without embedded components. However, once devices are embedded within the substrate, Level 1A cannot detect internal defects such as parametric drift or bonding failures. Levels 2A/2B or Level 3 become mandatory for meaningful quality assurance.
2. Detailed Test Methodologies
Each test level has specific procedural requirements. For Level 1A, the standard describes the conventional interconnection open/short test method where a voltage is applied across each net and the resulting current indicates continuity or isolation. Level 1B extends this to nets connected to embedded devices, requiring careful consideration of the device’s impedance to avoid false failure indications. For Level 2A (active device functional test), the substrate must be connected to a test fixture that can apply appropriate bias voltages and measure output responses — for example, verifying that an embedded operational amplifier exhibits correct gain and bandwidth. For Level 2B (passive devices), the standard recommends four-wire (Kelvin) measurement for resistors below 10 Ω to eliminate lead and contact resistance errors. Inductors and capacitors require measurement at specified frequencies relevant to the application.
The most comprehensive level, Level 3, involves constructing a complete circuit model of the embedded system, simulating the expected electrical behavior under defined stimulus conditions, and comparing measured responses against the simulation. The standard provides an explicit example (Figure 8 in the original document) demonstrating circuit model creation and simulation result validation for an embedded RC network. The test procedure flow diagram (Figure 10) guides the user through setup, measurement, and pass/fail determination for each level. This level is particularly valuable for analog embedded circuits where parametric accuracy directly impacts system performance.
✅ Engineering insight: The key difficulty in embedded substrate testing is limited test access — nodes buried under multiple laminate layers cannot be directly probed. The standard therefore strongly encourages a design-for-test (DfT) approach: embedding dedicated test pads, measurement vias, and even test access ports during the substrate design phase. Retrofitting test access after fabrication is often impossible, making DfT a prerequisite for Level 3 testing capability. The cost of adding test pads is typically less than 2 % of substrate area but can reduce diagnostic time by orders of magnitude.
3. Selecting the Appropriate Test Level
The test level required depends on the complexity of the embedded circuitry and the reliability requirements of the target application. For consumer electronics with simple passive embedding (fewer than five embedded passives per substrate), Level 2B testing with automated LCR meters integrated into the production line is usually adequate. For automotive or aerospace applications with embedded active devices (power management ICs, sensor interfaces), Level 3 functional testing may be required on every substrate. The standard notes that the electrical test report should clearly document which test levels were applied and the results for each embedded component or functional block, enabling traceability throughout the product lifecycle.
🚨 A common pitfall is assuming that embedded passive devices have identical characteristics to their surface-mounted equivalents. Embedded resistors fabricated by thin-film deposition exhibit different temperature coefficients (TCR) and voltage coefficients than discrete SMD resistors. Embedded capacitors formed by interlayer dielectric structures may show voltage derating and aging effects not seen in MLCCs. Level 2B testing must use measurement conditions specified in the relevant component specification, not generic LCR meter defaults.
4. Frequently Asked Questions
Q: Does this standard apply to redistribution layers (RDL) and fan-out wafer-level packages?
A: No. The standard explicitly excludes RDL and electronic modules defined as M-type business models in IEC 62421. It applies only to device embedded substrates fabricated using organic base materials (typical PCB laminate materials). Semiconductor wafer-level processes are covered by other standards.
Q: What sample size is required for testing?
A: The standard does not specify sample sizes — this should be determined by the product quality plan following accepted statistical methods (e.g., ANSI/ASQ Z1.4 or equivalent). In production, 100 % testing of Levels 1A/1B is typical, with Levels 2A/2B/3 applied on a sampling basis. For process qualification, the standard recommends 30 samples minimum for meaningful statistical characterization.
Q: Can boundary scan (IEEE 1149.1/JTAG) be used for Level 3 testing?
A: The standard does not specifically address boundary scan, but Level 3’s circuit-model-based approach is compatible with JTAG testing when the embedded chip supports boundary-scan features. The JTAG chain can be incorporated into the circuit model and used to verify interconnect integrity and device presence in the manufactured substrate.
