Call us — 01223 655015
Mon–Fri · 9am–5:30pm · No fix, no fee
Start a free diagnostic →

Data Recovery Case File · Portable Drives · Half Connected Is Worse Than Unplugged

A Partial Disconnection Can Interrupt Data While Power Continues, or Cycle Both Repeatedly

His enquiry names the moment precisely, which narrows things a great deal. A drive that "stopped working after it partially disconnected from the computer", which now "seems to have power, doesn't make any odd noises and doesn't appear to be damaged" but is recognised by nothing — with cables and driver updates already tried. A partial disconnection is a specific and unkind event: it is not a clean removal, and the difference matters.

MediaExternal hard drive following partial disconnection during use — apparently powered; no unusual mechanical sounds; not enumerating on any host
Reported situationExternal drive in use · connector partially withdrawn from the host during operation · drive ceasing to function afterwards · power apparently present · no unusual sounds reported · no visible damage · not recognised by any computer · cables substituted and driver updates attempted without effect
Fault classInterruption during operation with possible supply cycling — interface board or firmware state affected; mechanism not implicated by the reported symptoms
Equipment usedPartial disconnection assessed as an interruption rather than a clean removal · drive removed from the enclosure and addressed directly with imposed timeouts · current draw measured across the start-up cycle on a controlled bench supply · service area read under strict timeouts in vendor technological modes · imaged write-blocked once initialisation was achieved

The decode: why partial is worse than fully out, and what his observations exclude

Why unplugging a drive cleanly is usually harmless: everything stops together. Power and data are removed at once, and a drive that loses power completely simply stops.

What a partial withdrawal does instead: separates them. Connector pins are of different lengths by design, so a connector half out can leave power connected while data lines are broken.

Why that state is specifically damaging: the drive keeps running. A powered drive with no host is left mid-operation, potentially with a write incomplete and nothing to complete it.

What the other possibility is, and it is worse: repeated cycling. A loose connector making and breaking contact rapidly can power the drive on and off many times in seconds.

Why that matters more than a single interruption: each cycle is a start-up. A drive subjected to rapid repeated power cycling performs its most demanding sequence over and over.

Now what his observations rule out, and they rule out a lot: the mechanism. No unusual sounds, apparent power and no visible damage together argue against a mechanical fault.

What that leaves: the interface board or the drive's own state. A board damaged by cycling, or a drive left with inconsistent internal state, both present as a healthy-sounding drive that nothing recognises.

Why his testing was worth doing: it eliminated the outside. Substituted cables and multiple hosts remove the lead, the port and the machine.

Why driver updates could not have helped: nothing is being detected. A driver acts on a device the system has found, and the system is finding nothing.

Why the enclosure comes off first: it is the likeliest casualty. The interface board sits directly behind the connector that moved, and a drive removed from a failed enclosure frequently works normally.

On the bench

Partial disconnection was assessed as an interruption rather than a clean removal — connector pins differing in length by design, so a half-withdrawn connector can maintain power while breaking data lines, leaving a running drive mid-operation, or make and break contact rapidly and subject the drive to repeated power cycling. Absence of unusual sounds with apparent power argues against mechanical involvement. The drive was addressed directly outside its enclosure.

The outcome

Partial disconnection assessed as an interruption, the drive addressed outside its enclosure, and the service area read under strict timeouts. Free assessment, one fixed written figure including VAT, 50% of parts and labour upfront with the balance only on successful recovery where the drive is opened. The decode: half out is worse than all the way out. Connector pins differ in length deliberately, so power can stay connected while data stops — or contact can make and break repeatedly, which is many start-ups in a few seconds.

After a cable comes half out during use

Stop reconnecting it, and understand why the event matters: connector pins are different lengths by design, so a half-withdrawn plug can keep power connected while breaking the data lines, leaving a running drive mid-operation — or make and break contact rapidly, which is dozens of start-ups in seconds. Your observations are useful, though: no unusual sounds with power apparently present argues against the mechanism, and points at the interface board sitting right behind the connector that moved.

Drive stopped after a cable came half out?
Stop reconnecting — call Cambridge Data Recovery on 01223 655015; the event assessed as an interruption rather than a removal, drive addressed outside the enclosure, service area read under strict timeouts.
Request a quote online →

Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.