PC keeps freezing? Try this 20-minute Safe Mode test

The mouse stops. The screen stays put. Ctrl+Alt+Delete does nothing. It feels like the machine has given up on you. Most freezes are heat, a driver fight, tired RAM, a drive on the way out, or a power supply that cannot keep up. Do not start buying parts yet. Get it off safely, copy anything you cannot lose, then run one test that splits software from hardware.

• Wait a few seconds first. Sometimes Windows is just catching up. If the cursor still will not move, force it off.
• Hold the power button for 8 to 10 seconds. On a laptop, unplug the charger as well. Give it 30 seconds before you turn it back on.
• Copy the important files the moment it comes back. USB stick or cloud. Do this before you tinker.
• Unplug extras. Printers, hubs, extra drives, a second screen. A bad peripheral can lock the whole PC.

If it keeps happening, the 20-minute Safe Mode test below is the one step worth doing before anything else. Computer Support Scientist covers homes and offices across NSW. Sydney can start with computer help across Sydney. Central Coast homes can use computer help on the Central Coast.

What to do in the first ten minutes

A freeze feels urgent. The first ten minutes still matter more than a clever fix later. Get this part wrong and you can lose files, or make a sick drive worse.

1. Wait briefly. If the cursor still moves, give it a minute. If Ctrl+Alt+Delete does nothing, move on.
2. Force it off properly. Hold the power button until it is fully dark. Do not tap it. Do not yank the cord as your first move unless you smell burning.
3. Wait 30 seconds before powering on. Repeated hard restarts in a row can pile file-system errors onto a drive that is already struggling.
4. Back up as soon as you reach the desktop. Photos, tax files, the school assignment. Anything you cannot replace.
5. Unplug printers, USB hubs, extra drives, and a second monitor. Dell's own lock-up guidance starts here for a reason. A flaky dongle can freeze a healthy PC.
6. Feel the vents. Listen. Smell. Roaring fans and a hot chassis point to heat. Clicking or grinding from the drive bay is a stop-now sign. A burning smell is a safety issue, not a software one.
7. Write down the pattern. On boot, only in a game, only after an update, or completely at random. Three notes on your phone beat a vague "it just does it."

Pro Tip: Log the time, what you were doing, and anything you noticed. Fan noise. Heat. A specific app. Four entries usually show a pattern your memory would miss.

The 20-minute Safe Mode test

This is the step most people skip. It is also the one that saves the most time. Guessing at fixes is how people reinstall Windows for a problem that was a failing stick of RAM.

Hold Shift while you click Restart. Then go to Troubleshoot, Advanced options, Startup Settings, and pick Safe Mode. Use the PC normally for 20 to 30 minutes. Browse. Open the app that usually locks it. Safe Mode loads only the essential drivers, so the split is clean.

• If the freezing stops, you are almost certainly looking at a software or driver conflict. Graphics drivers are a common one.
• If it still freezes in Safe Mode, treat it as hardware until something proves otherwise. RAM, the drive, heat, or power.

Twenty minutes of ordinary use is enough. You do not need a full day's work in Safe Mode. You only need a fair test of the thing that usually locks it.

What Event Viewer is actually telling you

Windows logs freezes quietly. Event Viewer is where that log lives. Search "Event Viewer" in the Start menu, then open Windows Logs > System. Look for red Error entries around the time it locked up.

A few entries matter more than the rest, and Microsoft's freeze troubleshooting notes treat them as hardware clues, not trivia:

• WHEA-Logger usually points to CPU, RAM, or PCIe instability.
• Kernel-Power, Event ID 41 on its own, with nothing else nearby, often means the PC lost power. That is more PSU than software.
• Display driver names such as nvlddmkm.sys or amdkmdap.sys point to a graphics crash or a driver fight.

Take a screenshot. If you later want a person to look, those logs cut the guessing. You are handing over evidence, not "it just froze." Newcastle and the Hunter can start from computer help in Newcastle if you would rather talk it through than chase Event IDs alone.

Heat, RAM, and the drive

Once Safe Mode and Event Viewer have pointed you, these three checks cover most home PCs.

Temperatures under load

Download HWiNFO or HWMonitor. Run something demanding: a 4K video, a game, an export. As a rough home guide, most consumer CPUs should sit well under 85C under load, and GPUs under 80C. A fan at full scream during light browsing is a warning too. XDA's overheating signs match what we see before a freeze or a sudden shutdown.

A laptop on a doona, or a desktop jammed against a wall, will run hot. Clear the vents. If it is still cooking after a dust-out, the paste between the CPU and the heatsink may have dried. Repasting a thin laptop is easy to get wrong. That is a fair point to stop.

RAM that looks "random"

Faulty memory is sneaky. The freeze looks unrelated to what you were doing. Search Windows Memory Diagnostic, let it restart and scan, then follow up with MemTest86 from a bootable USB if you can. A full pass can take hours. It catches intermittent errors the quick Windows tool misses. If you have more than one stick, test them one at a time. That isolates a single bad module instead of blaming the whole set.

Drive health

Open CrystalDiskInfo and check SMART on the main drive. "Good" means healthy. "Caution" or "Bad" means back up now, not later. Then open Command Prompt as administrator and run:

chkdsk C: /f /r

That scans for file system errors and bad sectors. It needs a restart on the system drive. Repeated bad-sector warnings mean replace the drive, not keep repairing it.

If you hear clicking, grinding, or clunking, stop using the PC. That is a data job first. See our NAS and drive recovery notes for the "stop and photograph" habit. Do not keep powering it on to test theories.

Fixes once you know the side

Match the cause. Do not run every fix in a row hoping one sticks.

If it is heat

• Clear dust from vents, fans, and heatsinks with compressed air. Do not spin a fan with the straw.
• Confirm fans actually spin under load, not only at idle.
• Give a desktop breathing room. Use a cooling pad on a laptop that cooks during calls or games.

If Safe Mode was steady

• Graphics suspects: Display Driver Uninstaller (DDU) in Safe Mode, then install the current driver from NVIDIA, AMD, or Intel. Not whatever Windows Update picked.
• If freezes started after a Wi-Fi update: Device Manager, Network adapters, your adapter, Properties, Driver, Roll Back Driver.
• If it started right after a Windows patch: Settings, Windows Update, Pause updates for a week so you can test without a moving target.
• Open Task Manager (Ctrl+Shift+Esc). Watch Processes while it feels wrong. Trim Startup apps. Run a Windows Security full scan. Two antivirus tools at once is a classic freeze cause. Keep one.

Corrupted Windows files are worth a pass if drivers look clean. In an administrator Command Prompt, in this order:

sfc /scannow
DISM /Online /Cleanup-Image /RestoreHealth

Both are safe. They do not delete your documents. Run DISM if sfc says it found errors it could not fix.

If RAM tests fail

Reseat each stick: out, then firmly back in. Loose contacts cause more freezes than people expect. Overnight MemTest86 is worth it if you can spare the hours. A stick or slot that always errors is your answer. Replace that module. Do not keep mixing it with a good one "to see."

If Event Viewer shows Kernel-Power 41 alone

A tired power supply mimics other faults: freezes under load, sudden reboots, nothing useful nearby in the log. Proper testing means measuring the 12V rail under load. That is not a visual check. Hand this one to a technician.

How to tell if it is actually fixed

One clean evening is not proof. Intermittent faults hide from a short test.

1. Do the thing that used to freeze it, for 30 to 60 minutes. Browse, stream, export, game. Match the real load.
2. Leave HWiNFO running. Glance at temperatures.
3. Check Event Viewer afterwards for new WHEA-Logger or Kernel-Power errors.
4. Turn on a SMART alert in CrystalDiskInfo so a dropping drive does not surprise you next month.

A two-minute monthly look at temperatures and SMART status catches problems while they are still cheap. Never run two antivirus programs at once. They fight each other, and that fight freezes PCs more often than people think.

If the machine is merely sluggish rather than frozen, start with our slow PC checks instead. Task Manager first, then startup apps. A freeze that ends in a black screen or a boot loop is a different article: Windows won't boot.

When to stop and call

Some symptoms are not a Saturday project. The risk of making it worse is the point, not your skill.

• Repeated Blue Screens with a different code each time. That pattern is hardware instability, not one naughty driver.
• WHEA-Logger showing up regularly. CPU, RAM, or the board needs proper stress testing.
• Clicking, grinding, or a burning smell. Power down. Unplug. A swollen laptop battery is the same rule.
• It still freezes after Safe Mode, Event Viewer, RAM, and the drive checks. You are likely looking at a PSU or a motherboard, and that needs swap testing.

Bring what you already have: Event Viewer screenshots, the CrystalDiskInfo report, and rough times. That handful of detail lets a remote session narrow the cause before anyone opens the case.

What we see on visits

A lot of "random" freezes are not random once the order is followed. A graphics driver after a Windows update. A laptop full of lint. One bad RAM stick in a pair. A drive that CrystalDiskInfo already flagged Caution three weeks ago. The jobs that drag on are the ones where Windows was reinstalled first, then RAM was bought, then the PSU was blamed, and the vents were still packed.

We start with Safe Mode, the logs, then heat and memory. Parts get replaced when the numbers support it, not because a forum thread listed them. Work is approved before it starts.

PC Scientist can help when it keeps locking up

Some of these checks take twenty minutes. Others hide under a sealed laptop, a PSU that needs a meter, or a drive that is starting to fail. That is a fair reason to stop. Computer Support Scientist can look with you, remotely first if that is enough, or onsite if the hardware itself is the problem.

Tell us whether it freezes on boot, under load, or at random. Mention what Safe Mode did, and what Event Viewer showed if you looked. We will say whether a remote session will do, or whether a visit makes more sense. Call 0493 563 381, or pick your area below.

Sources

These pages help when you want the maker's own steps:

• Windows-based computer freeze troubleshooting - Microsoft
• System locks up, freezes, or an error occurs - Dell
• Signs your CPU is overheating - XDA Developers