The fault · a click, a beep or a grind from the drive
Take the power off it before you read another line. A drive that has begun to click or beep is generally still holding what it held an hour ago; what disappears afterwards disappears in the retries. The head assembly cannot work out where it is, so it withdraws to the ramp, sweeps out again, gets no answer and resets — round and round while the power is on it, and each sweep is one more chance of contact with the surface. Pull the plug and the damage stops at whatever it has reached. One drive costs £300 + VAT; the look that comes first is free, one written figure reaches you before any of the work begins, and on most jobs no invoice follows where nothing comes back.
Put it to an engineer — the first look costs nothing
0203 0868633
| How it shows itself | The usual reason for it | Where that leaves you |
|---|---|---|
| An even click, roughly one a second, that carries on | Position comes from servo data laid down at manufacture. With none of it legible, the actuator resets to the ramp and sweeps out once more | Cut it off at the wall |
| A faint beeping, usually out of a 2.5" portable | Nothing is turning and nothing else can be heard: a spindle bearing gone solid, or heads bonded to the platter | Nothing to be gained by tapping it |
| A dry rasp, or a grinding sound | One head is resting on a platter, shaving coating away on every rotation | Data goes with each turn |
| One short chirp on each power-up | The heads will not calibrate, which usually follows a knock | Switch off, and leave it there |
| Two or three clicks, then silence | It gives up, resets itself, and runs the same sequence again | Unplug it to end that loop |
| The clicking begins on particular files | A head has failed, or one band of platter is damaged | Stop copying — it needs an image |
A case number goes on it the day the parcel lands, and an engineer settles what has truly failed before anything else happens — free of charge, and first in the order of work. Back to you come two things together: a straight note of what is liftable and what is not, plus one figure, fixed and written down. Accept it, or decline and owe us nothing.
Heads and platter surfaces are examined first, with nothing connected to the drive at all. Should a noisy unit really need to be run, it is run by hand, a second or two at a time, with somebody watching it. No drive here is left turning on its own.
A donor head stack matched to your exact model goes in where the failed one was. A bearing gone solid is worked free. Heads resting on a platter are raised off it without being dragged anywhere. Then, and only then, the drive is put to work on a single task — a full image, and no other demand on it.
A hardware imager reads gently, in passes. Difficult areas are put off, the straightforward sectors are banked first, and the imager returns for the rest until the copy is as full as the disk allows. Every step from there on is performed against that copy; your drive is read from and never written to.
What was recovered is listed for you first, and only then does a bill exist. Approve the list and it is invoiced; turn it down and it is not — and where nothing has come back, most jobs carry no charge whatever. Recovered data travels home on fresh media bought in for your job, with the postage at our end. Your case is not closed until you have read those files on a machine of your own.
This page comes down to one word: stop. A head rides over a platter turning thousands of times a minute, at a clearance measured in nanometres, and engineering held to that tolerance does not put itself right. On noisy drives, whether a recovery ends up complete or partial usually comes down to how quickly somebody got to the plug. Post it and an engineer opens it on the day of arrival; a written figure follows, and nothing is chargeable until you accept it.
Five years of takings were on the drive an Ealing upholstery firm sent in, knocking away at about one beat a second. It was the second donor stack that held. Imaging was given all the time it asked for, and the books opened again with that quarter's VAT return still to file.
Most likely, yes. A click is the head assembly failing to position itself, and it says nothing at all about the platters below, which is where your data physically lives. With a matched donor stack and an unhurried image, the greater part of a disk normally comes back. What costs you is the running time after the noise starts.
No — it adds a second fault to the one you already have. The tip is older than any drive now in service. Put a sealed head assembly in the cold and two unrelated things begin together: metals of different kinds contract by different amounts, and water forms where there is meant to be none. Let it warm up again and there is oxide across the contacts, so the bench is looking at two problems rather than one.
Every platter leaves the factory with servo marks written across it, and the arm hunts for those. When none of them answers, it withdraws to the ramp and repeats the whole approach; that repetition is what you can hear. The one thing the sound will never give you is why. A head out of alignment, a preamplifier that has given up and contact marks on one surface cannot be told apart through an alloy lid. Separating them needs the cover off, under laminar air.
Software cannot. Whatever it asks for is still served by a head assembly that has already failed, so the only result is a drive kept doing the very thing wearing it out.
Looking at it is free. Back comes a list of what opened and what did not, together with a single price to finish, set down in writing while you are still free to say no. On most jobs an invoice only follows the data. Until that list reaches you, leave the drive switched off.