Your external hard drive is not showing up because the fault lies somewhere between the USB connector and the filesystem driver: a power or cable failure, a USB bridge chip that never enumerates, a partition table the operating system cannot parse, or a drive that has dropped into a firmware-safe mode after too many bad sectors. In most cases the drive is detected at the hardware level but never mounts, which is why it spins and lights up while staying invisible in File Explorer or Finder.

The order matters. A drive that does not appear in firmware-level tools is a different problem from one that appears but shows no letter, and the repair path diverges sharply between them. Working through the layers in sequence — physical, USB bridge, partition, filesystem — is faster than reinstalling drivers at random.

The enumeration chain from port to mount point

Every external drive must pass four gates before a file manager lists it. First the host controller detects a device on the bus and assigns an address. Second, the drive's USB-to-SATA bridge chip answers a SCSI inquiry and reports a vendor and product string. Third, the operating system reads sector zero and finds a partition table. Fourth, the filesystem driver mounts each partition and assigns a letter or mount point.

Each gate can fail silently. A bridge chip that draws more current than the port supplies may reset repeatedly, a condition called enumeration looping. The drive spins up, the light blinks, and the host sees a device appear and vanish several times per second. Windows may play the connect chime on a loop; macOS may log repeated "device not configured" errors.

Sector zero is a single 512-byte or 4096-byte block. If it is unreadable — a bad sector, a corrupted master boot record, a partition table overwritten by a failed format — the drive presents as uninitialized or raw even though every other sector is intact. The operating system has no map of where the data lives, so it offers to initialize the disk, which writes a fresh partition table and destroys the old layout.

A practical test separates these cases in under a minute. Open Disk Management on Windows or Disk Utility on macOS and check whether the drive appears at all. If it shows with a capacity but no volume, the bridge and controller are fine and the fault is logical. If nothing appears, the fault is physical or in the bridge.

Quantified context: how often each layer fails

The failure distribution is not even. Cable and connector faults account for a large share of "dead" drives that work on a second attempt, because USB-C and micro-B connectors are rated for roughly 10,000 insertion cycles, and the micro-B design concentrates mechanical stress on a thin internal tongue. A cable that has been coiled tightly in a bag for a year often develops an intermittent data pair while the power pair still conducts, producing a drive that spins but never enumerates.

Power is the second common cause. A 2.5-inch portable drive typically draws 4.5 to 5 watts at spin-up, which sits close to the 4.5-watt budget of a single USB 3.0 port. Bus-powered drives on unpowered hubs, laptop ports, or long thin cables frequently brown out during spin-up. A 3.5-inch desktop drive needs 12 volts and ships with its own power adapter for exactly this reason; using the wrong adapter, or one rated below the drive's label, causes the same symptom.

Firmware-level head parking and safe-mode behavior explains another cluster. When a drive logs repeated read errors, many internal firmware routines reduce head speed, disable write caching, or park the heads and present a minimal identity. The drive still answers basic commands but refuses to expose the full capacity. This is protective behavior, not a virus, and it is one reason a drive can be visible in a BIOS or a disk utility but invisible to the operating system's mount layer.

Counterintuitively, the most destructive failure mode is often triggered by the user. A drive that fails to mount prompts a dialog offering to format or initialize it. Accepting writes a new partition table over the old one. The old table is typically a few kilobytes in a known location and can usually be recovered with partition-recovery software, but only if the user stops writing to the drive immediately.

Edge cases: enclosures, hubs, and filesystem mismatches

The enclosure is a separate computer. Every external drive contains a small bridge board running its own firmware, and that firmware can fail independently of the disk inside. A bridge that negotiates USB 3.0 incorrectly causes the drive to work at USB 2.0 speeds or not at all on certain hosts. Replacing the enclosure or moving the bare drive to a known-good dock is the definitive test, and it requires knowing whether the drive's USB port is soldered to the bridge or the drive is a bare SATA unit inside a shell.

Filesystem mismatches produce a drive that is detected and healthy but unreadable. Windows cannot natively read Apple's APFS or the older HFS+ without third-party software; macOS can read but not natively write NTFS. A drive formatted on one platform and moved to the other may appear in disk tools with a recognizable capacity yet mount nowhere. The same applies to Linux ext4 volumes on a Windows machine.

Hubs and port topology matter more than most guides admit. A USB 3.0 hub shares one upstream link among all downstream ports, so a bus-powered drive on a hub alongside a webcam and a keyboard may fail to spin up even though it works alone. Some laptops reduce port power in battery-saving modes, and some desktop front-panel ports deliver less current than rear ports wired directly to the motherboard.

A less obvious edge case is the drive that mounts only after a delay. Spinning drives take 3 to 10 seconds to become ready, and some bridge chips report ready before the platters reach operating speed. On fast-booting systems the mount attempt can time out before the drive is ready, leaving it invisible until the user unplugs and replugs.

Misconceptions that send people down the wrong path

The first misconception is that a drive that lights up is working. The LED is wired to the bridge's power rail, not to the disk's read channel or the host link, so a lit drive can have a dead controller or an unreadable sector zero.

The second is that reinstalling USB drivers fixes most cases. Driver faults are real but uncommon on current Windows and macOS builds; the enumeration chain fails far more often at the cable, power, or partition layer, and driver reinstalls do not touch those.

The third is that data recovery requires the drive to mount. Recovery tools read raw sectors through the operating system's block device, which works whenever the drive enumerates, even if the filesystem is unreadable. That is why "uninitialized" is often recoverable while "not detected at all" is not.

The fourth is that a clicking drive should be imaged at home. Clicking indicates a mechanical fault — a failed head, a stuck spindle, or media damage — and every power-on cycle can grind the head across the platters. Professional clean-room recovery exists precisely because continued operation destroys the remaining data.

Red-flag warnings: Clicking, beeping, or repeated spin-up and spin-down cycles: power off and stop, do not keep trying. A drive that becomes hot to the touch within minutes: disconnect and do not reconnect until the enclosure is inspected. The "initialize disk" or "format" prompt: decline it; this writes a new partition table. Burning smell or a swollen enclosure: unplug from the wall, not just the computer. A drive that works intermittently after a drop: treat every successful mount as possibly the last.

Practical actions in diagnostic order

Start at the physical layer. Try a different cable of known quality, then a different port, then a different computer, then a powered hub or direct motherboard port. Each substitution isolates one link. If the drive is 3.5-inch, confirm the power adapter matches the label's voltage and current rating; a mismatched adapter is a common cause of a drive that never spins.

Check firmware-level visibility. On Windows, open Disk Management and Device Manager and look for the drive under disk drives and under universal serial bus controllers. On macOS, use Disk Utility and System Information's USB section. If the drive appears with a capacity but no volume, proceed to logical recovery. If it does not appear at all, the fault is physical or in the bridge.

For a detected but unmounted drive, do not format. Image the drive first with a sector-level tool if the data matters, then attempt partition-table recovery on the image. The original partition table is often recoverable because it sits in a fixed location and is rarely overwritten except by a format the user approved.

For a bridge-level failure, remove the bare drive and connect it through a known-good SATA dock or a replacement enclosure of the same interface. This distinguishes a dead bridge from a dead disk in minutes.

For a drive that mounts slowly or intermittently, disable USB selective suspend and check whether the host port delivers adequate current. A powered hub resolves most bus-power problems.

Timeline table for a typical failure and recovery attempt:

Stage What happens Typical duration Action
First symptom Drive spins, no letter appears Minutes Check cable and port
Enumeration loop Device appears and vanishes Seconds per cycle Move to powered port
Drive detected, no volume Disk Management shows capacity Minutes Stop; do not format
Partition table lost Drive shows as uninitialized Immediate Image before repair
Firmware safe mode Drive answers basics, not capacity Persistent Test bare drive in dock
Mechanical fault Clicking or beeping Worsens with use Power off; professional recovery
Clean-room recovery Platters imaged in isolation Days to weeks Expect high cost

Comparison of recovery paths when the drive is detected but unreadable:

Path What it recovers Risk Cost range (USD) When to use
Partition-table repair Intact data on lost partition Low if imaged first 0 to 30 for software Drive detected, uninitialized
Filesystem repair (chkdsk, fsck) Minor corruption Moderate; can worsen 0 Filesystem errors, not hardware
Sector-level imaging Full raw copy Low; read-only image 30 to 100 for tools Before any repair
Enclosure or dock swap Access to bare drive Low 20 to 60 Bridge suspected dead
Professional clean-room Data from mechanical failure Low if prompt 300 to 2,000+ Clicking, beeping, dropped drive

Reconcile the thresholds before acting. A drive that appears in firmware tools but not in the file manager is a logical problem; a drive absent from firmware tools is physical or bridge-level. A drive that mounts after a delay is a power or enumeration timing problem, not data loss. A drive that clicks is a mechanical problem regardless of what any software reports. Each threshold points to a different tool, and using the wrong tool on the wrong layer wastes time and can destroy data.

The professional caveat is that "not showing up" covers at least six distinct failure classes, and the most common advice online — reformat, reinstall drivers, run a generic repair command — is correct only for a narrow slice. The safe default is to stop writing to the drive the moment data matters, image it before any repair, and treat every prompt to initialize or format as destructive until proven otherwise.

For drives under warranty, contact the manufacturer before opening the enclosure; opening voids coverage on many models. For drives holding irreplaceable data, the cost of professional recovery is often lower than the cost of a botched home repair.