The PC That Kept Crashing — Diagnosing an Unstable Machine in Merthyr Tydfil

Mark from Merthyr Tydfil had been putting up with a crashing PC for weeks. Not a consistent crash — not the kind that happens every time you do a specific thing, which would at least give you something to go on. Just random. Sometimes mid-game, sometimes while browsing, sometimes seemingly for no reason at all. Blue screens, sudden restarts, the occasional complete freeze.

He'd tried the obvious things — restarting, running Windows Update, even a factory reset. Nothing made a difference. The crashes kept coming.

Consistent crashes are actually easier to deal with than random ones. If a PC crashes every time you open a specific program, or every time it's under heavy load, that's a clue. Random instability is harder because it could be caused by a wide range of things — and some of them look identical from the outside.

The main candidates for random crashing are:

RAM faults — faulty or failing RAM is one of the most common causes of random instability. The errors it produces are inconsistent because they depend on which part of the RAM is accessed at any given moment. A stick of RAM can pass basic tests and still cause crashes under certain conditions.

Storage issues — a failing hard drive or SSD can cause crashes, freezes, and data corruption. As sectors on the drive start to fail, reads and writes to those areas produce errors that the system can't handle gracefully.

Overheating — a CPU or GPU that's running too hot will throttle itself to reduce heat, and if it can't cool down fast enough, the system will crash as a protective measure. This often produces crashes under load rather than at idle, but not always.

Power supply problems — a PSU that's failing or undersized for the system can cause instability when the components draw more power than it can reliably deliver. This is often overlooked because the PSU is easy to ignore.

Driver or software conflicts — less common than hardware faults, but worth ruling out.

Diagnosing an unstable PC properly means working through the possibilities systematically rather than guessing. Mark's machine went through a full diagnostic: RAM tested with memtest, storage checked for bad sectors and SMART errors, temperatures monitored under load, event logs examined for error codes.

The culprit was a failing stick of RAM. One of the two sticks in the machine was producing errors under certain memory access patterns — not consistently enough to fail a quick test, but enough to cause the random crashes Mark had been experiencing. Running the machine with just the good stick confirmed it: stable, no crashes.

A replacement RAM stick of the same specification, properly seated and tested. Mark's PC has been running without issue since.

The whole job — diagnosis, part sourcing, repair, and testing — was done within the week. Mark picked it up knowing exactly what had been wrong and why, explained in plain English without jargon.

The temptation with a crashing PC is to try things at random — reinstall Windows, replace parts speculatively, hope something fixes it. This wastes time and money. A factory reset won't fix a hardware fault. Replacing the wrong component doesn't help.

Proper diagnostics take a bit longer upfront but mean you know what you're actually dealing with before any money is spent on parts. Steve will always diagnose first and explain what he's found before recommending any repair.

Random crashes, blue screens, freezes, or a machine that just doesn't feel right — bring it to the bench. Steve covers South Wales including Merthyr Tydfil, Dowlais, Aberfan, Treharris, and the surrounding areas.

Book a diagnostic or get in touch — free diagnostics, no fix no fee, always.