Problem Statement
PC exhibiting random restarts, occasional BSOD (MEMORY_MANAGEMENT), and slower-than-expected boot times. Goal: identify root cause without replacing parts unnecessarily.
Tools Used
- Windows Memory Diagnostic
- CrystalDiskInfo (drive health)
- Event Viewer (System / Application logs)
- Task Manager & Resource Monitor
- CPU-Z (hardware identification)
- Compressed air + thermal paste
Diagnostic process
Before touching anything, wrote down: when crashes occur, error codes seen, how long the machine has been running, and any recent changes (new software, hardware added). Random restarts under load vs. at idle tells a very different story.
Opened Event Viewer → Windows Logs → System. Filtered by Critical and Error levels. Found repeated BugCheck events with code 0x0000001A (MEMORY_MANAGEMENT). This pointed to RAM or driver corruption before any physical testing.
Scheduled a memory test via mdsched.exe. Let it run the Extended test overnight (two passes). Results came back with errors on the second DIMM slot. Re-seated both sticks and ran again — errors cleared on one stick, persisted on the other.
Removed one stick, booted, tested. Then swapped sticks. The PC was stable with stick A in slot 1, unstable with stick B in any slot. Confirmed: one bad RAM module.
Ran CrystalDiskInfo — overall health showed "Caution" on the HDD with 47 reallocated sectors. Not failing immediately but noted for monitoring. SSD checked out healthy.
Found heavy dust buildup on the CPU cooler and case fans. Cleaned with compressed air, reapplied thermal paste to CPU. Idle temps dropped from 68°C to 44°C. This also likely contributed to the crashes under load.
Key diagnostic commands
# Pull recent critical/error events from System log Get-EventLog -LogName System -EntryType Error,Warning -Newest 50 # Check for BugCheck (BSOD) events specifically Get-WinEvent -FilterHashtable @{LogName='System'; Id=1001} | Select -First 10 # Check disk health via SMART data Get-PhysicalDisk | Select FriendlyName, HealthStatus, OperationalStatus # Check drive for errors chkdsk C: /scan # Launch memory diagnostic mdsched.exe # (schedules test on next reboot)
Screenshots
Outcome
One RAM module was faulty. Removed it, ran the PC on the single good stick — no crashes in 72 hours of testing. Recommended replacing the faulty module. Also flagged the HDD for monitoring due to reallocated sectors.
CPU was thermal throttling under load due to dust and dried thermal paste. After cleaning, temps dropped 24°C and perceived performance improved. This was masked by the RAM issue but would have caused problems independently.
What I learned
Event Viewer pointed directly at MEMORY_MANAGEMENT before I touched a single component. Always check logs first — it narrows the field significantly.
Testing one RAM stick at a time is the only way to be sure which module is faulty. Guessing which one to replace is expensive and often wrong.
High CPU temp can cause instability that looks like software or RAM issues. Cleaning and re-pasting should be on the checklist for any machine that hasn't been serviced in 2+ years.
Writing down pass/fail for each step meant I could hand the notes to anyone and they'd understand exactly what was tried, what the result was, and what was left to do.