What Was Happening
The first symptom had been the machine taking noticeably longer to start over a few weeks. Then it started occasionally getting stuck on the Windows logo for a long time before reaching the desktop. The morning it came in, it had cycled through “Diagnosing your PC” and “Attempting repairs” three times and then dropped to a black screen with no further activity.
It was an office machine running a handful of standard applications — nothing exotic — so the customer’s priority was getting it back to work, plus recovering a folder of documents on the desktop they did not have backed up elsewhere.
Our Diagnosis
Boot failure on a desktop is one of the more pleasant faults to diagnose, because most components are accessible and swappable individually:
- POST check. Powered on with the side panel open. Single beep, normal POST progression on screen, BIOS reachable. Confirmed the motherboard, CPU and RAM were healthy enough to begin booting.
- Drive enumeration. BIOS listed the drive but described it with the disk’s manufacturer name only, no model — sometimes a hint that the drive’s firmware isn’t responding fully. SMART read attempt from a bootable diagnostic USB succeeded with warnings: reallocated sectors well above threshold, pending sectors growing, an offline scan history showing read errors at multiple addresses.
- Boot attempt from the failing drive. Got further than the customer reported by selectively skipping past the automatic repair loop. Reached the loading bar, then a stop error mid-boot. Re-attempted, this time it would not progress past the BIOS handover. Classic intermittent-read failure for a drive that is past saving.
- Boot from a known-good Windows installer USB. Reached setup cleanly. Confirmed the machine itself was healthy and the only fault was the drive.
Decision point: do we try to repair Windows in place (risk: it works for a day then the next read failure breaks something else), or do we accept that the drive needs to come out and start fresh? With a drive showing this SMART profile the answer is always option two — repair-in-place is throwing away time.
How We Fixed It
Step 1 — image the failing drive. Connected the original drive to a SMART-aware imaging rig that skips and retries around bad sectors rather than failing the whole operation. Got most of the data off, including the folder of documents the customer flagged as important. A small percentage of the user profile was unreadable — fortunately none of it was critical to the customer.
Step 2 — fit the SSD. SATA SSDs are the simplest desktop upgrade in PC service. Disconnected the failing drive, fitted the new 240GB SSD in the same bay, used the existing SATA cable and power connector.
Step 3 — clean Windows install. Booted from a Windows installation USB, deleted any leftover partitions, let Windows create its own partition layout on the fresh drive, ran setup through to first boot.
Step 4 — Lenovo driver bundle. This is the step that gets skipped on home installs and matters more than people realise. Lenovo publishes a curated driver bundle per ThinkCentre model — chipset, network, sound, video, system management, BIOS, firmware. Installing those before adding any user applications gives the cleanest possible foundation. We use Lenovo’s official tooling rather than third-party “driver updater” software.
Step 5 — Windows updates run to current state on the bench so the customer is not staring at “preparing updates” on their first day back.
Step 6 — restore the recovered user data from the imaged old drive into the user profile on the new install. Documents, desktop, downloads, browser bookmarks and saved passwords (where the customer’s master credentials were available) all replaced.
Step 7 — verification. Three full power-cycle reboots, application launch tests, network connectivity, printer setup if applicable, then a final SMART read on the new drive to confirm clean baseline.
The Result
- Boot to login screen: under 20 seconds (down from “several minutes if at all”)
- Office applications launching instantly
- A drive baseline that should give the customer years of use
- Most user data recovered; the unrecoverable portion identified honestly rather than glossed over
One working day on the bench, including the time-consuming image of the failing drive.
Why This Happens
Mechanical hard drives are wear items. The two failure modes we see most often:
- Bad sectors accumulating slowly. The drive starts reallocating sectors as the magnetic platter develops imperfections. SMART tracks this in the background. Eventually too many sectors are bad in regions Windows needs to read frequently, and boot becomes unreliable.
- Mechanical wear on the read/write head. The drive starts taking longer per read, then begins missing reads outright. This is what you hear as “clicking” on a dying drive — though on a healthy bay it can be quiet enough that the only symptom is the slowdown.
Desktop drives often die earlier than people expect because they are running 8+ hours a day in a warm case for years on end. Five to seven years is a reasonable working life for a mechanical drive in an office machine; some last longer, but plan for a replacement in that window.
The good news: replacing the boot drive with a SATA SSD is one of the lowest-cost, highest-impact upgrades available for a 2014-and-later desktop. It does not require a new motherboard, new RAM, or any other change — and it turns a mid-life machine into one that feels new.
How to know your desktop’s drive is going
- Boot time creeping up over weeks. Worth a SMART check.
- Random “scanning and repairing drive” messages at startup. Always worth a SMART check.
- Applications opening slower than they used to, even though nothing has changed.
- Activity light on the case constantly lit, even when idle.
- Strange clicking or grinding from the case. Stop using the machine and seek help — that drive is at end of life.
A SMART check takes a few minutes. A failed drive that takes your documents with it takes much longer to recover from.
Local Help in Brixton SW9
Tower and all-in-one PCs benefit from component-level diagnosis rather than guessed-at part swaps.
We work on every major brand plus custom-built machines — and we’re happy to handle customer-supplied parts on a clear protocol.
Book a workshop visit on 020 7610 0500 or contact us first.