Effective documentation isn't about writing a novel—it's about capturing the right details efficiently. Follow these steps to ensure your failure notes drive action:
Step 1: Start with the Basics—Identify the "What" and "Where"
Begin with the foundational details that answer:
Which PCB failed, and during which test?
This includes:
-
Unique PCB serial number (critical for traceability)
-
Test station/operator ID (to spot if one station is consistently flagging failures)
-
Test type (ICT, functional, environmental, etc.)
-
Timestamp (when the failure occurred—useful for tracking shifts or machine maintenance cycles)
Pro tip: Use a barcode scanner to auto-populate serial numbers and test station info. This reduces human error and saves time.
Step 2: Describe the Failure—Be Specific, Not Sloppy
"Doesn't work" is useless. "Fails to transmit data over UART at 9600 baud; error code 0x03 appears in debug log" is actionable. Focus on
observable, measurable details
:
-
What exactly happened? (e.g., "Voltage at pin 7 drops to 0.5V when load is applied; spec is 3.3V ± 0.2V")
-
What didn't happen? (e.g., "No response to I2C command 0x7A; oscilloscope shows no signal on SDA line")
-
Were there error messages or alerts? (Screenshots or log snippets are gold here)
Avoid jargon unless everyone on your team uses it consistently. If a technician writes "SMT pad lifted," that's clear. If they write "BGA voiding," make sure your engineers know what that means (and if not, add a glossary to your documentation template).
Step 3: Capture Environmental and Operational Context
Failures often depend on conditions. A PCB might pass functional testing at 25°C but fail at 60°C. Note:
-
Ambient conditions (temperature, humidity, voltage fluctuations during testing)
-
Test parameters (e.g., "Tested at 12V input; failure occurred after 5 minutes of continuous operation")
-
Pre-test history (e.g., "PCB was reworked after wave soldering; this was its second test pass")
Step 4: Link to Components and Assembly Data
This is where your component management system and SMT assembly records shine. Note:
-
Component IDs (e.g., "IC U2: Part number ABC123, lot code L2345")
-
Assembly details (e.g., "SMT assembly performed on Line 3, Feeder 7; solder paste batch XYZ789")
-
Any known component issues (e.g., "Capacitor C10: Supplier X had a recall for this part in Q2 2024")
By linking failures to specific components and assembly steps, you can quickly spot if a problem is component-related (e.g., a bad resistor batch) or process-related (e.g., a misaligned stencil in SMT).
Step 5: Attach Evidence—A Picture (or Log) Is Worth 1,000 Words
Words can be ambiguous; images and data logs aren't. Always include:
-
Photos/videos of physical defects (e.g., a cracked solder joint, a bent pin)
-
Oscilloscope/waveform captures (for signal integrity issues)
-
Test equipment logs (e.g., ICT test reports, functional test software outputs)
Modern custom PCBA test systems can auto-attach these files to failure records, so technicians don't have to remember to upload them manually.
Step 6: Record the Fix—And Whether It Stuck
Documentation doesn't end when the PCB is fixed. Note:
-
What was done to resolve the failure (e.g., "Replaced resistor R15 with 1kΩ 1% tolerance")
-
Post-repair test results (e.g., "Passed functional test after 10 cycles of thermal stress")
-
Follow-up actions (e.g., "Flagged resistor supplier for batch inspection")