Random PC Restarts: How to Rule Out Bad Wall Power

Quick Answer

Rule out wall power by comparing the restart pattern across a known sound outlet or circuit, recording UPS or voltage-monitor events, and testing the PC under the same workload. Keep each change controlled.

A plug-in monitor can show useful voltage events, but it cannot certify the outlet or capture every transient. Never improvise mains test leads or open a distribution board. A qualified electrician should assess hot outlets, loose plugs, tingling equipment, burning smells and repeated trips.

The goal is to separate three possibilities: the supply path, the desktop and the workload. Replacing a PSU before that split can leave the real fault untouched.

Step 1: Write Down the Restart Pattern

Record the exact time, application, CPU or GPU load and whether the screen went black before the PC restarted. Note what happened to the router, room lights and other equipment.

Check Windows Event Viewer or the operating system's equivalent after the reboot. An unexpected-shutdown event confirms the system did not close cleanly. It does not tell you whether wall power, PSU protection, temperature or software caused it.

Photograph motherboard diagnostic codes or UPS alarms. Keep the observations together so you can compare events rather than relying on memory.

Step 2: Inspect Without Opening Mains Equipment

Shut the PC down and check the removable mains lead, plug and outlet for damage, looseness, discoloration or heat. Follow the PC and PSU maker's safety instructions.

Do not keep using a plug that crackles, smells burnt or becomes hot. Do not tighten an outlet or inspect fixed wiring yourself. Call a qualified electrician.

Check the PC-side power connectors only if you are competent to work inside the computer and the warranty permits it. Use the correct PSU cables for that exact unit. Modular cables are not universal.

Step 3: Reproduce the Workload

Run the same game, render or application on the original setup after important data is backed up. Watch temperatures and hardware monitoring alongside the wall-power record.

A restart tied to rising CPU or GPU temperature moves the diagnosis towards cooling. A restart at a repeatable combined load can involve the PSU, cabling, motherboard, GPU or supply path.

Do not stress-test an unstable system unattended. Stop for smell, smoke, abnormal noise or excessive heat.

Step 4: Compare a Known Sound Supply

Use another outlet only when a qualified person has established it is on a suitable, sound circuit. Running a long extension lead across rooms can create voltage drop and a trip hazard, so follow equipment and electrical guidance.

If the PC is stable on the known sound supply under the same workload, the original outlet or circuit becomes a stronger suspect. If it restarts in both locations, return attention to the desktop and any supply issue shared by both points.

One successful hour does not prove a rare fault gone. Match the test window to how often the restart occurs.

Step 5: Add Measurement

A suitable plug-in voltage monitor can show whether input moved during the event. Record minimum, maximum and time if the unit supports it. Check its accuracy and rating before use.

Normal nominal voltage is not a single fixed reading every second. Use current South African supply requirements and a qualified electrician to judge whether a measured pattern is acceptable. Do not invent a pass or fail threshold from an internet comment.

For an intermittent problem, professional power-quality logging can capture voltage, frequency and event duration over several days. That evidence is more useful than watching a live display and hoping to see the fault.

Step 6: Use a UPS as a Diagnostic Layer

A correctly sized line-interactive UPS can record input events, regulate some voltage variation and bridge short interruptions. If the PC becomes stable while the UPS repeatedly transfers, the upstream supply deserves investigation.

The UPS does not prove that wall power caused every restart. It may change the load, provide a different output waveform or hide the symptom. Read the event log and continue the diagnosis.

Measure the tower and required monitor under their heaviest normal task. Stay within both UPS watts and VA, and use the maker's runtime chart.

Step 7: Rule Out the PSU Properly

Software usually cannot identify the installed PSU model, rated capacity or internal health. Read the physical label only after safe shutdown and access, or use the build record.

Check whether the capacity and connectors suit the installed GPU and CPU. Look for the correct dedicated cables where the component maker calls for them. Never mix modular leads from another PSU.

A bench PSU tester does not reproduce a high-end gaming load or prove every protection circuit. A known sound replacement unit of suitable capacity can be a useful workshop comparison when fitted by a competent technician.

If the evidence points to replacement, compare Corsair PSU options by exact specification and best-selling power supplies. Match capacity, form factor, GPU connectors and warranty to the complete build.

Step 8: Check the Rest of the Desktop

Reset unstable overclocks or undervolts to supported defaults for testing. Update firmware only from official sources and on stable power.

Run memory and storage diagnostics through their supported tools. Check the operating-system crash dump when one exists. A software driver crash may restart the PC without any mains event.

Change one variable at a time. Swapping the outlet, PSU settings, memory profile and driver on the same day destroys the comparison.

Build Your Evidence Table

For each restart, capture location, workload, PC temperatures, wall reading, UPS state and whether other equipment flickered. Add the change tested and its result.

After three or four events, patterns often become clearer. If they do not, the log still helps the electrician or PC technician choose the next measurement.

Keep backups current throughout. An unexpected reset can interrupt file writes even when the hardware survives.

FAQ

How do you test wall power for a suspected PC issue?

Compare the same PC workload on a known sound supply, log input events with suitable equipment and use a qualified electrician for circuit assessment.

Can bad wall power look like a failing PSU?

Yes. Both can produce black screens and unexpected restarts. Controlled comparisons and event records separate them.

Does a UPS fix the wall fault?

No. It may regulate or bridge some events and add diagnostic evidence. The source still needs repair.

How long should you monitor?

Long enough to cover the normal recurrence. A daily fault needs less time than one appearing every few weeks.

Should you test before replacing a PSU?

Yes, when it is safe to do so. A measured diagnosis reduces unnecessary replacement and helps find hazards.