Diagnostics, Testing and Technical Documentation
How does a professional technician prove a machine is fixed, and record it so someone else can follow the work?
- Interpret POST beep codes and motherboard error LEDs.
- Run burn-in and stress tests to verify stability under load.
- Apply a systematic troubleshooting method.
- Produce technical documentation that meets professional quality standards.
Overview
Repairing a computer is only half the job. The other half is proving the repair holds under load, and writing it down so the next technician — or the school auditor — can see what was done, why, and what it cost. This chapter closes the Computer Architecture strand and feeds directly into the Week 10 project portfolio.
A systematic troubleshooting method
Work through the steps in order. Guessing wastes parts and time.
Six steps used industry-wide:
- 1. Identify the problem: Question the user, note symptoms, ask what changed recently.
- 2. Establish a theory: List probable causes, simplest and cheapest first.
- 3. Test the theory: Swap, reseat or measure to confirm or eliminate each cause.
- 4. Plan and act: Decide the repair, consider the effect on the user and on data, then carry it out.
- 5. Verify full functionality: Test under real load, not just at the desktop, and apply preventive measures.
- 6. Document: Record findings, actions, outcomes and time.
POST codes and error indicators
At start-up the firmware tests essential hardware. One short beep usually means all is well. Repeating or patterned beeps map to a fault in the motherboard manufacturer's beep-code table — most commonly RAM, then graphics, then the board itself. Modern boards add debug LEDs (CPU / DRAM / VGA / BOOT) or a two-digit code display that names the failing stage directly.
Burn-in and stress testing
A machine that boots is not proven. Burn-in runs the hardware hard for a sustained period to expose faults that appear only when hot: memory tests (such as MemTest86) for RAM errors, CPU stress tests while monitoring temperature, and disk surface scans and SMART reports for drives. Record the peak temperature and whether the machine stayed stable for the whole run.
Technical documentation standards
A service report is a professional record. Include every heading below.
Required sections of a service report:
- Job identification: Job number, date, technician name, machine asset tag and serial.
- Reported fault: In the user's own words, plus your observed symptoms.
- Diagnostic steps: What you tested, the readings you took, what you ruled out.
- Action taken: Parts replaced with part numbers, settings changed, software installed.
- Verification: Tests run after the repair and their results, including burn-in duration.
- Recommendations and sign-off: Preventive advice, outstanding risks, time taken, cost, signatures.
Write in plain, factual language. Anyone reading it in a year should be able to repeat exactly what you did.
Practical — Diagnose, repair and report
- Your teacher will introduce a fault into a lab machine (for example an unseated RAM module or a disconnected power cable).
- Apply the six troubleshooting steps, recording your evidence at each one.
- Repair the fault and verify it with a boot test plus at least ten minutes of stress or memory testing.
- Write a complete service report using all six required sections.
- Submit the report as part of your Term 3 project portfolio.
- What are the first two steps of the troubleshooting method?
Reveal answer
Identify the problem (question the user, note symptoms), then establish a theory of probable cause.
- What does a repeating beep pattern at start-up usually indicate?
Reveal answer
A hardware failure detected by POST — most commonly a RAM fault; check the board's beep-code table.
- Why run a burn-in test after a repair?
Reveal answer
To prove the machine is stable under sustained load and expose faults that only appear when components are hot.
- Name four sections that must appear in a service report.
Reveal answer
Any four of: job identification, reported fault, diagnostic steps, action taken, verification, recommendations and sign-off.
- What is the difference between a symptom and a root cause?
Reveal answer
A symptom is what the user sees; the root cause is the underlying fault producing it.
Create a reusable service-report template in a word processor for your school's ICT department, with all six sections and space for signatures.