PC Troubleshooting & Diagnostics
PC Troubleshooting & Diagnostics
Diagnose no-boot, POST errors, overheating, and common hardware failures.
The Diagnostic Approach
Before replacing hardware, follow a systematic process:
- Identify the symptom ߀” no power, no display, crashes, beeps, slow performance
- Isolate the cause ߀” remove/swap components to narrow it down
- Test with minimal hardware ߀” one RAM stick, no GPU (use iGPU), no extra storage
- Check the simple things first ߀” reseated connectors fix more problems than anything else
No Power At All
The system does nothing when you press the power button.
- Check the PSU rocker switch on the rear (must be in the ON / I position)
- Check the front panel power button header on the motherboard (refer to manual for correct pins)
- Try a known-good power cable
- Test the PSU with a paper clip: short the green wire (PS_ON) to any black wire on the 24-pin connector ߀” the PSU fan should spin. Warning: do this only briefly for testing.
- Check for shorts: remove the motherboard from the case and test outside on a cardboard box
POST Failures ߀” Beep Codes
During Power-On Self-Test (POST), the motherboard speaker emits beep codes to signal errors. Codes vary by BIOS manufacturer:
| BIOS | Beep Pattern | Likely Cause |
|---|---|---|
| AMI | 1 short | POST passed OK |
| AMI | 3 long | RAM not detected |
| AMI | 1 long + 2 short | GPU error |
| AMI | 5 short | CPU error |
| Award | Continuous long beeps | RAM not seated |
| Award | 1 long + 2 short | GPU error |
Modern boards often have a Q-LED or POST code display instead of beep codes.
No Display
System powers on but nothing appears on the monitor:
- Check monitor input selection and cable connection
- If using a discrete GPU: reseat it, try another PCIe slot
- Remove the GPU and use the iGPU (CPU must have integrated graphics)
- Try one RAM stick at a time in different slots
- Clear CMOS: remove the motherboard battery for 30 seconds, or use the CLRTC jumper
RAM Diagnostics
- Use MemTest86 ߀” boots from USB, runs extensive memory tests overnight
- Swap RAM sticks one at a time to find a faulty module
- Try each stick in different slots (some slots are faulty, not the RAM)
Storage Diagnostics
| Tool | What it checks |
|---|---|
| CrystalDiskInfo | S.M.A.R.T. data ߀” reallocated sectors, pending sectors, spin-up failures |
| CrystalDiskMark | Read/write speed ߀” identify degraded drives |
| HD Tune | Error scan ߀” maps bad sectors on the drive surface |
chkdsk /r |
Filesystem-level check and repair |
A drive with reallocated sectors > 0 is failing. Back up immediately.
Overheating
Symptoms: sudden shutdowns, thermal throttling (CPU/GPU drops to minimum speed), loud fan noise.
- Check CPU/GPU temperatures with HWiNFO64 or Core Temp
- CPU should idle below 50 ?°C; under load below 90 ?°C for most CPUs
- If temps are high: clean the cooler and heatsink of dust
- Re-apply thermal paste (dried paste is a common cause of high temps after 3߀“5 years)
- Check that all case fans are spinning
GPU Diagnostics
- Artefacts (coloured pixels, corrupted textures): dying GPU VRAM or driver issue
- Black screen during gaming: GPU crashing ߀” check temperatures and power delivery
- FurMark: stress test that identifies GPU instability under load
- Try the GPU in another PCIe slot or a different system
Blue Screens (BSOD) ߀” Common Stop Codes
| Stop Code | Common Cause |
|---|---|
| MEMORY_MANAGEMENT | Faulty or overclocked RAM |
| WHEA_UNCORRECTABLE_ERROR | CPU/RAM/PSU instability; hardware defect |
| SYSTEM_SERVICE_EXCEPTION | Driver issue |
| CRITICAL_PROCESS_DIED | OS file corruption; bad storage |
| IRQL_NOT_LESS_OR_EQUAL | Driver or RAM issue |
Check the Event Viewer (Windows + X ߆’ Event Viewer ߆’ Windows Logs ߆’ System) for error details after a crash.