The 'main' HD on my old Acorn has failed. Most of the material is backed up - but not all including some CAD files I did recently, which took quite some time. Silly me, but it doesn't eat HDs in the same way as a PC does, so I'd got a bit . According to my records - which were backed up - it's over 15 years old.
It's an IDE Hitachi Deskstar 160 Gb. The software recognises there's a drive present, but says it's either not formatted or not initialised. The replacement drive works just fine.
As it happens I have a second identical one which still works. Was on a PC which is no more - the current ones are SATA.
Tried swapping the PCB - but no luck. Also tried putting it in the freezer for a day.
Anything else I can try? Not worth spending serious money on a data recovery service - but I don't mind having a fiddle.
I don't recollect whether more recent Acorns have a PC type architecture but maybe you have access to a PC, an external drive caddy and a live Linux CD? I've had success reading and downloading files from a dodgy HDD that way. The fact that the computer is not running the same OS as the faulty disc is a bonus in this situation.
It was pretty daft not trying software recovery utilities before taking the drive apart or sticking it in the freezer, both pretty much measures of last resort. You've probably massively reduced your chances of a successful recovery. Anyway, put the original PCB back on, and if it still spins up and doesn't make endless clicking noises, try running this:
formatting link
or this:
formatting link
(yes, I know it says fsck, which implies *nix, but this version is for RISC OS format discs.)
What was the error message? Elsewhere you said it had 'failed', which is generally taken to mean some mechanical/electrical failure, like not being detected, not spinning up, making grinding noises or the PCB going pop.
If it is merely 'corrupted' then the mechanism is working fine, you just need software to make sense of the data. Possibly there are parts that can't be read any more, but if some parts can you want to recover them before further damage.
The imperative when faced with an unhappy drive is to clone it to a disc image asap. Then write that to another drive and see if you can repair the damage to the data. If you mess up, re-write it from the image and try again. The image is your safe 'master' copy, and having done this you no longer depend on the failing drive.
It is better to image it in a different machine so you don't get things trying to access it. I'd use ddrescue or dd_rescue (those are different things) on Linux, but there are other approaches. It is possible to do this on RISC OS but if it's confused I wouldn't necessarily trust its block reading code.
Linux will mount ADFS disc images, so a Linux live CD could be used to create an image of the original hard disc, then to mount the image and see if it's recognisable, e.g.
take a sector by sector disk image of the failed drive # dd if=/dev/sdX of=diskimage.img
make sure the adfs kernel module is installed # depmod adfs.ko
mount the image as a Linux loopback mount # mkdir /mnt/RiscOSdisk # mount -v -t adfs -o ro,loop diskimage.img /mnt/RiscOSdisk
Didn't mention it specifically - but did say software was giving an error message 'not formatted or not initialised.' That is the same with the IDEFS software and DiscKight.
These seem to be PC based, but surely you have an ancient PC somewhere? Or a mate has one? If you are in North London, you can come and borrow one of mine.
Reload it from your backup. If you don't have a backup then it isn't worth keeping.
Recuva works on some file systems even after formatting but I don't know if it works on riscos file systems.
You should make a backup using a sector copy before you tray anything. If you can't read anything using the sector copy then you are going to find it difficult without swapping bits like heads and platters around and you need specialist equipment to do that.
It means the drive can't servo; it can't find the data tracks on the disc surface, so it cannot calibrate to track zero. The noise you hear is the heads pinging back and forth as the drive firmware attempts to find a servo track.
It can be caused by several things - head crash, head failure, head amplifier failure, controller PCB failure, etc. It's pretty terminal.
If you haven't done anything daft like open it and there hasn't been a head crash (which scores gouges in the disc surface, removing the magnetic media) the data is probably still recoverable by a recovery firm, but it won't be cheap. Whether it's worth the cost is up to you - it's your data, only you know its value.
En el artículo , snipped-for-privacy@gmail.com escribió:
Some are just inherently noisy when accessing (i.e. the heads are moving about). In earlier days I used to like being able to hear the disc working away.
Most modern spinning rust drives now configurable AAM (acoustic management) - you can set a balance between performance and noise.
Yes, though that's quite a separate issue with a different symptom. The one I'm talking about frequently goes into headspin, everything has to wait un til it behaves again. It's been doing backup service for a few years withou t incident, other than pausing here and there. It seems to have been caused by an overheat event.
Here's a heads up for you. When attempting to recover data from a failed HDD by 'borrowing' the PCB from another identical, perfectly working drive, you need to take extreme ESD precautions when you seperate the boards from the drive casings[1].
Whilst the user accessable connections (power and data ports) are well protected against modest ESD hazards, the interfaces between the board and the spindle and head actuator motors and the read/write heads aren't so blessed. It is all too easy to damage these interfaces with modest levels of ESD that otherwise wouldn't any cause damage to the SATA data and power interfaces.
I reckon you've already tried your best option with the board swap. It can work but note my sorry tale below. Out of interest, did you swap the 'borrowed' board back? And, importantly, did the donor drive still work ok afterwards?
[1] I successfully recovered a customer's data off a 1TB Samsung using this technique but landed up with TWO dead drives in the end. The trick had worked just fine on the first board swap but when swapping the boards back, the good one also stopped working.
I'm guessing my usual cavalier attitude with ESD precautions that had proved more than sufficient in safely handling drives and adapter cards over the previous two or three decades was a bit too cavalier to prevent the boards from succumbing to the residual levels of ESD still being generated by my every day anti-ESD handling technique. :-(
That was a few years ago when 1TB HDDs were still a useful size and a lot more expensive. I've long since given up trying to track down examples which have failed due to bad sectors from which to cannibalise the controller boards.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.