Data Recovery Case File · Mac & Apple Ecosystem · Reaching It Is the Work
The Drive Behind the Glued Screen
His enquiry was straightforward and the complication is one people do not anticipate. A 27-inch iMac from 2015 that "has recently been unable to power up, and thus I have lost access to the files on it as they are not cloud based." The good news is the usual and reliable kind: a machine that will not power up has almost certainly failed on the power side or the mainboard, which leaves the drive untouched. The complication is that on an all-in-one of that design, the drive is not behind a hatch — it is behind the display, which is bonded to the chassis with adhesive. That does not change whether the data is recoverable. It changes what reaching it involves, and therefore the price.
| Media | Internal storage from a 2015 27-inch iMac — machine not powering on; storage accessible only after removal of an adhesively bonded display panel |
| Reported situation | Machine failing to power up · no symptoms attributed to storage · files not held in cloud storage · owner seeking recovery and costing |
| Fault class | Host failure with storage unimplicated — access requiring controlled display removal; configuration to be confirmed as single drive or paired volume |
| Equipment used | Display separated under controlled conditions with adhesive replaced on reassembly · storage read write-blocked on native equipment · paired-volume configuration confirmed before imaging · contents verified by opening |
The decode: why the machine is not the problem, and why access is
Why his data is probably fine: the reliable part first. A machine that will not power on at all has usually failed at the power supply, the power path or the mainboard — and in all of those the drive is a bystander that simply stopped being asked for anything. Storage failure more typically produces a machine that powers up and then cannot boot, which is a different symptom. So the drive is very likely intact, and the recovery is a matter of reaching it.
Why reaching it is a job: on a desktop tower the drive comes out in two minutes with a screwdriver. On an all-in-one of this design the display is the front of the machine and it is held on with adhesive strips rather than clips or screws. Separating it means cutting that adhesive cleanly, supporting a large glass panel while its cables are disconnected, and replacing the adhesive properly on reassembly. Done carelessly, the panel cracks, the cables tear, or the machine cannot be closed up again. It is entirely routine work for anyone equipped for it and a genuinely poor idea as a first attempt with a kitchen knife.
Why that affects the quotation rather than the outcome: worth being explicit, because customers are often surprised. Two machines with identical drives and identical faults can carry different figures purely because one has a hatch and the other has a bonded display. The work of reading the drive is the same; the work of getting to it is not.
The configuration question worth settling first: a 27-inch machine of that year may have a single hard disk, a single solid-state drive, or a paired volume combining a small solid-state module with a large disk into one logical drive. That last configuration matters enormously, because the data is distributed across both devices and neither is readable alone — so both must be found and imaged. Establishing which he has is part of the assessment rather than a detail.
What he should not do meanwhile: repeatedly trying to power the machine achieves nothing if the fault is power-side, and if the drive is in any way involved, each attempt asks a possibly-struggling drive to work again. And the machine should not be taken somewhere for a repair before the data question is settled — repair work means powering, handling and sometimes replacing components, none of which should happen while the drive holds the only copy.
On the bench
The display was separated under controlled conditions rather than levered — the adhesive cut cleanly, the panel supported throughout and its cables released before any weight was taken, with fresh adhesive fitted on reassembly so the machine could be closed properly afterwards. The configuration was confirmed before imaging rather than assumed, since a paired volume requires both devices and finding only one would have produced an unreadable result. Storage was read write-blocked on native equipment, taking the dead machine out of the question entirely, and the contents were verified by opening before delivery on fresh media.
The outcome
The display removed and refitted under controlled conditions, the configuration confirmed and the contents imaged, verified and delivered. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone with an all-in-one that will not start: the machine failing to power on usually leaves the drive completely untouched, so your data is probably fine; but on a bonded-display design the drive sits behind a glass panel held on with adhesive, so reaching it is controlled work rather than a two-minute step — which affects the figure without affecting the outcome; and establish whether your model uses a paired volume across two devices, because both would be needed. Don't have the machine repaired before the data is off it.
All-in-one that will not power up
Take encouragement from the symptom: a machine that won't power on at all has usually failed on the power side or the mainboard, and in both cases the drive is untouched. A failing drive normally gives you a machine that starts and then can't boot. What's different about an all-in-one is access — the display is bonded to the chassis with adhesive rather than screwed on, so getting to the drive means cutting that adhesive cleanly, supporting a large glass panel while cables are released, and refitting adhesive afterwards. That's routine for anyone equipped for it and a bad first project. It also means your quote reflects the access as much as the recovery. Ask whether your model uses a paired volume across a small SSD and a large disk, because both devices would be needed. And don't send it for repair before the data is off.
The drive is probably fine — call Glasgow Data Recovery on 0141 404 0294; display separated under controlled conditions, configuration confirmed, storage read write-blocked and verified.
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.