Hard Disk Kharab Hai Ya Nahi? HDD Test Full Method Explained

Listen First, Before Any Software Test

A full test starts with your ears, not with software. A healthy hard disk hums quietly and spins up smoothly. Repetitive clicking, grinding or beeping — or a drive that spins up and down again and again — means mechanical damage inside the drive. No software test can fix that, and every extra minute of power makes data recovery harder.

If the drive sounds normal, note the symptoms that brought you here — slow loading, freezing, files that fail to copy, Windows repairing itself at boot. The tests below will tell you whether the disk explains them or not.

Read the SMART Report with CrystalDiskInfo

Download CrystalDiskInfo from its official site (crystalmark.info) and open it. It reads the drive's own self-monitoring system, called SMART — the health diary the drive has been keeping since the day it was made. Look at two things. First, the big Health Status box: Good (blue), Caution (yellow) or Bad (red) — that is the drive maker's own summary.

Second, the raw values of three attributes in the table: 05 Reallocated Sectors Count, C5 Current Pending Sector Count and C6 Uncorrectable Sector Count. On a healthy drive all three raw values are 0. On a failing drive they climb — as an example, a dying disk might show raw values like 312 reallocated, 47 pending and 47 uncorrectable, because it is already eating through its spare sectors.

Run a Full Surface Scan

SMART only knows about sectors the drive has already tried to use. A surface scan reads every single sector on the disk — including the forgotten corners — and draws a map of the whole surface. Use the Error Scan tab in HD Tune, or your drive maker's own long test (SeaTools for Seagate, WD Dashboard for Western Digital).

Expect it to take time: roughly 1 to 3 hours per terabyte on a mechanical drive is normal. Every block should come back green. Red blocks are damaged sectors — note how many there are and where they sit. One isolated red block on an old drive is a warning; red blocks scattered across the map, or a scan that stalls and crawls at one spot, is a failing surface.

Check the Speed Pattern

Run a benchmark such as CrystalDiskMark, or the benchmark tab in HD Tune, and look at the shape of the result, not just the headline number. A healthy mechanical drive gives a fairly smooth, predictable speed curve from start to finish. A drive with weak or damaged areas shows sudden deep drops — the speed falls to near zero and struggles back — because the heads keep re-reading the same weak spots.

For a rough reference only: a healthy 7200 RPM desktop drive reads around 120–200 MB/s sequentially and a 5400 RPM laptop drive around 80–120 MB/s. Age, model and how full the drive is all move these numbers, so treat this as a pattern check, not a pass mark.

Cross-Check with Windows' Own Disk Status

Windows keeps a basic health record you can ask for without any extra software. Open Command Prompt and run the command this video uses:

wmic diskdrive get model,status

Each disk is listed with its model and a status: OK means the drive's own failure prediction has not tripped; Pred Fail means the drive itself is predicting failure — treat that as a replace-now signal. This step only confirms or contradicts the SMART picture; it never replaces the surface scan.

Get-PhysicalDisk | Select FriendlyName, MediaType, HealthStatus, OperationalStatus

The Verdict — Combine All Three Results

No single number decides. Combine the evidence like a technician does:

Healthy: SMART Good with 05, C5 and C6 raw values at 0, surface scan fully green, smooth speed curve. Trust the drive — and still keep normal backups, because healthy drives can also die suddenly.

Backup and watch: SMART Caution with small non-zero values, a surface scan with one or two red blocks, mild speed dips. Copy everything important today, then re-test in a week. If the numbers grow, the drive is on its way out.

Replace now: SMART Bad, growing reallocated or pending counts, red blocks spreading, clicking sounds, or a scan that cannot finish. Stop testing, rescue the data and replace the drive — testing a disk in this state only spends the life it has left.