Hardware Troubleshooting

POST (Power-On Self-Test) and No-Boot Issues

At power-on, firmware (BIOS/UEFI) runs POST to verify the CPU, RAM, and basic devices before handing off to the bootloader. A failure stops the boot and is signaled by beep codes, motherboard debug LEDs (commonly a 4-light row labeled CPU/DRAM/VGA/BOOT), or a 2-digit hexadecimal POST code on enthusiast boards.

Representative Beep Codes (vary by AMI / Phoenix / Award)

Beep codes are not standardized — always confirm against the board manual. On modern UEFI systems, the DRAM or VGA debug LED staying lit tells you exactly which subsystem stalled.

Minimum-Hardware Isolation (No POST)

1. Confirm the outlet, the PSU rocker switch (rear), and a known-good power cable.
2. Look for the standby power LED on the motherboard — proves the PSU delivers 5VSB.
3. Verify the front-panel power button header is connected.
4. Strip to minimum hardware: CPU, one RAM stick, integrated graphics, no drives.
5. Reseat CPU, RAM, and GPU; clear CMOS (jumper or remove the CR2032).
6. Swap in a known-good PSU to rule out a dead supply.

Motherboard, CPU & Power Symptom Matrix

CPU Overheating

Throttling, instability, or a shutdown seconds after boot signals a thermal problem. Fix order: blow dust from the heatsink/fans with compressed air; confirm the CPU fan spins and its header reads RPM in firmware; remove the cooler, clean old paste with 90%+ isopropyl alcohol, and apply a fresh pea-sized dot; verify all mounting points are seated; improve case airflow; and confirm the cooler is rated for the CPU's TDP. Most desktop CPUs throttle around 95-100C and shut down at their thermal limit, so temperatures under sustained load should sit well below that.

RAM Troubleshooting

Tools: Windows Memory Diagnostic (mdsched.exe, reboots and scans) and MemTest86 (bootable USB, multiple thorough passes — let it run several cycles to catch intermittent faults).

Storage Troubleshooting

Never defragment an SSD. It has no heads, so fragmentation does not slow reads; defrag only burns finite write cycles. Rely on TRIM, enabled by default on modern Windows/macOS/Linux.

Power Supply & Safety

Random reboots under load, fans that twitch then die (over-current protection tripping), or a burning smell all implicate the PSU. Test with a PSU tester or a multimeter on the 24-pin (3.3V/5V/12V rails). Never open a PSU — its primary capacitors can hold a lethal charge even when unplugged. Replace it as a sealed unit and size it to the build's total wattage plus headroom.

A subtle PSU symptom is a system that powers on fine at idle but reboots only when the GPU spins up under a game or render — the supply cannot deliver peak 12V current to the graphics card. Recalculate the build's draw (CPU TDP plus GPU board power plus drives and fans) and add roughly 20-30% headroom; a supply running near 100% load runs hot and ages quickly. Loose or partially seated PCIe and EPS (CPU) power connectors produce the same intermittent reboots, so reseat every modular cable on both ends before condemning the unit.

Interpreting Symptoms vs. Components

Domain 5 questions reward pattern recognition — matching a described symptom to the single most likely failed part. Keep these high-value associations sharp:

• Clock/date keeps resetting and BIOS settings are lost → dead CMOS battery (CR2032), not the motherboard.
• Distended (swollen) capacitors near the CPU socket → degraded board; replace the motherboard.
• Burning smell or visible scorch → power off immediately, unplug, and isolate the shorted component; do not keep powering on.
• Grinding or clicking from a spinning drive → mechanical HDD failure; back up before it dies.
• Sluggish performance plus occasional BSODs → run a SMART check and a memory test; either a marginal drive or marginal RAM can produce both.
• System indicators: a steady standby LED with no fans means the PSU has power but the board isn't starting; no LED at all means no standby voltage reaches the board.

A Worked Isolation Example

A workstation powers on (fans spin) but shows no display and gives no beep. Apply minimum-hardware isolation: remove the discrete GPU and connect the monitor to the integrated graphics output. If video returns, the discrete card is the fault — reseat it, test it in another machine, or replace it. If there is still no display, remove all but one RAM stick and move it to the first recommended slot; watch the DRAM debug LED. If the LED clears and the system POSTs, the original stick or slot was bad.

Methodically changing one variable at a time — exactly Step 3 of the methodology applied to hardware — isolates the failed part without guesswork or unnecessary parts replacement.