3.5 Test Strategy: ICT vs Functional vs Burn-In

Testing is the only mechanism that converts a manufacturing gamble into a guaranteed product. Without a robust test strategy, you are not shipping a product; you are shipping a probability. The goal of testing is not to debug the design—that is the engineer’s job. The goal of manufacturing test is to verify that the physical unit was built correctly and will not fail immediately upon arrival. Different test methods catch different types of defects, and relying on just one leaves gaping holes in your quality defense.

ICT (In-Circuit Test): The "Bed of Nails"

Scope: Structural Electrical Integrity.

Speed: Extremely Fast (Seconds).

ICT does not check if the product "works"; it checks if the product was built to the schematic. The board is pressed onto a fixture of spring-loaded pins ("Bed of Nails") that contact specific Test Points. The machine isolates each component and measures it individually.

The Engineering Reality

ICT is a Sherlock Holmes investigation. It asks: "Is Resistor R1 actually 10kΩ? Is Capacitor C4 actually 10µF? Is there a short between Power and Ground?"

• If a wrong component was placed (e.g., a 1kΩ resistor instead of 10kΩ) → Then ICT catches it instantly.
• If there is a microscopic solder bridge under a chip → Then ICT detects the short circuit before power is applied, preventing the board from blowing up.
• Tradeoff: High Setup Cost. You must pay thousands of dollars for the custom fixture. It is worth it only for volumes > 1,000 units.

FCT (Functional Test): The Simulation

Scope: Performance verification.

Speed: Slow (Minutes).

FCT treats the device as a "Black Box." It powers up the unit, loads the firmware, and simulates the real-world environment. It pushes buttons, reads sensors, and checks displays. It answers the ultimate question: "Does it do what it is supposed to do?"

The Engineering Reality

FCT is the only way to catch logical or system-level failures.

• If the solder joints are perfect but the component is internally defective → Then ICT passes, but FCT fails.
• If the firmware has a bug or the calibration is off → Then only FCT will detect it.
• Risk: Bottleneck. If the test takes 3 minutes and you need to build 1,000 units a day, you need multiple test stations, doubling your equipment cost.

Pro-Tip: Avoid binary "Pass/Fail" results. Your FCT should log quantitative data (e.g., "Output Voltage: 5.01V"). This allows you to spot trends—if the voltage drifts from 5.0V to 5.2V over a week, you can stop the line before it hits the 5.5V failure limit.

Burn-In: The Stress Test

Scope: Infant Mortality (Early Life Failure).

Speed: Extremely Slow (Hours to Days).

Burn-In is the process of running the device at elevated temperatures and voltages for an extended period. It is designed to kill weak components before they leave the factory.

The Bathtub Curve

Electronics reliability follows a "Bathtub Curve." Failures are high at the beginning (Infant Mortality), low in the middle (Useful Life), and high at the end (Wear Out).

• If you ship without Burn-In → Then the customer becomes the filter for Infant Mortality failures.
• If you run Burn-In for 4–24 hours → Then you force these weak units to fail inside the factory, ensuring that shipped units are in the stable "Useful Life" phase.
• Target: Mandatory for medical, automotive, and high-reliability industrial gear. Optional for cheap consumer goods.

Flying Probe: The Fixtureless Alternative

Scope: Prototype / Low Volume ICT.

Speed: Very Slow.

For prototypes, you cannot afford a $5,000 ICT fixture. A "Flying Probe" machine uses robotic arms to move probes around the board, measuring components one by one.

• Pro: Zero tooling cost. You can test a single board immediately.
• Con: Time. Testing one board can take 20 minutes.
• Use Case: NPI builds and troubleshooting complex defects.

Final Checklist

Test Type	Catches...	Blind Spot	Cost Structure	Best For...
ICT	Shorts, Opens, Wrong Values	Functional Logic, Firmware	High Fixed Cost (Fixture) / Low Unit Cost	Mass Production (>1k units).
FCT	System Failures, Bad Parts	Specific Component Values	Low Fixed Cost / High Time Cost	Final validation for every product.
Burn-In	Latent Defects, Weak Parts	Assembly Errors	High Energy & Time Cost	High-Reliability / Safety Critical.
Flying Probe	Shorts, Values	Speed	Zero Fixed Cost / Very High Time Cost	Prototypes / NPI.