Is it feasible to replace an HD controller board?

Topic

New Reply

A friend’s elderly neighbour has just asked me for help. He showed me how his external 500 Gb hard disk – a mains-powered 3.5″ Seagate unit – is no longer recognised by either his Windows PC or laptop. He said he isn’t worried about the hard disk itself, just the data on it, i.e. years of photos. I took the HD home and connected it to several other PC’s but none of them would recognise it. In each case, Windows Explorer would just show the Windows 7 equivalent of a never-ending egg-timer. With his permission I opened the HD case, disconnected the actual HD from its USB interface board and connected the drive directly to a SATA connector on my PC’s mainboard.

This time the drive WAS recognised by the PC’s BIOS but the following warning appeared:
42577-hd1
Click to enlarge

I tried to boot into Windows 7 but the PC refused to go past the Loading Windows screen.

I tried to boot into the latest SeaTools for DOS (from a USB stick) but the PC hung on the FreeDOS boot screen.

The only thing I was able to do was to carry out a low-level DPS test (it’s an HP base unit), which failed immediately:
42578-hd2
Click to enlarge

During all this I found out that what I thought was a (S.M.A.R.T.) standard method of reporting HD diagnostics doesn’t appear to be a standard at all…
42580-hd3
Click to enlarge

… and that HD manufacturer’s like Seagate (on its particularly unhelpful website) don’t actually let you know what the S.M.A.R.T. error codes mean:
42579-hd4
Click to enlarge

(Thanks, Seagate… I’ll remember this.)

Anyway, when I told the chap the news, he told me that he had been moving his data to the external drive rather than copying his files to the backup HD. That’s right… he has no backups of years of photos other than his damaged external HD nor can he afford a professional company to remove the platters from his ‘dead’ hard disk (in a clean room) and mount them on a substitute drive for reading.

Instead, he asked this question (which I can’t answer as I have no experience)… is it possible to buy the exact same model of HD and transfer its controller board to the damaged HD to read the physical data (in order to copy to another drive)?

The controller board of the HD appears to be secured by 8 Torx screws:
42581-hd5
Click to enlarge

I have no idea what lays beneath and thought I would ask before exploring further.

Reply | Quote

Replies

jwoods – I looked at SpinRite. IMHO it’s quite clever how Steve Gibson describes SpinRite’s capabilities but doesn’t make its limitations particularly clear. Instead you have to interpret statements like ‘SpinRite works with HD’s “that can be made visible to DOS through the addition of controller BIOS or add-on DOS drivers” ‘.

I could be wrong but I take that to mean that SpinRite would not work in my case… ‘cos the drive is not visible to either the BIOS or (Free)DOS (or Windows) as an accessible/addressable device (… hence me asking for help from anyone/everyone who has previous practical experience of perhaps swapping controller boards).

Reply | Quote

Have you used SpinRite yourself? Did you recommend it from personal experience?

Reply | Quote

Rick there are SATA drivers available on GRC, but as it has always seemed like a pain to get anything going I have not expressed my S/W iunky gene. Assuming that there is no translation problems w/ removing the usb/sate I/F, I would want the same, exact same controller board. Not just the same model #. Stuff frequently gets changed w/in model numbers, cheaper components, or maybe less flaky . Any dif COULD cause an incompatibility. I would explore the costly pro option if he feels it is really important to him. And be warned these pros operate in a clean room, if you are not a recovery specialist you do NOT have one no matter how much scrubbing and vacuuming you do. (Have you ever tried to put on a screen protector?) :cheers:

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

Hi David… yep, gone through all the possible/probable costs of professional (clean room) recovery with him. He and his wife are pensioners with limited savings and I hate to see an 80-year-old trying to hold back his tears… but I don’t know what else I can do.

I’m more than happy to spend whatever time is necessary to try to repair his HD if it’s possible to swap the controller board with one from an identical HD. I just need to know if anyone has done it before.

He’s given me the OK to remove the Torx screws to see whether data connections are soldered or sockets. I guess that’s the next step in the absence of any feedback.

Reply | Quote

That data could mean an awful lot to the owner, from the description/screenshots, I can’t tell which part is most likely to be at fault there – attempt the wrong ‘fix’ and it could make things worse, even irretrievably so.

Rick, how does the drive smell & what kind of noise does it make during boot? How long before it sounds like it starts to spin up & does it sound like it’s hitting full speed without fluctuating?

Yes, it’s possible to do a ‘board swap, providing the donor ‘board is close enough in all details – but it will almost certainly require the firmware info to be transferred from the old drive to the new ‘board – most reputable companies will do this FOC, often by return of post and usually at no extra cost. Prices do vary…

Try HD Sentinel to check the SMART details, the author is knowledgable and approachable, if you can get some details give him a nudge, he might have some specifics about that Seagate range’s SMART report.

Turn off Explorer’s autorun ‘feature’ before you connect it again, you don’t want it hanging around being hogged by that.

Reply | Quote

From experience autopsying HDD (ok I just like taking things apart) it is all ribbon cable in a header. older might have some pin connectors IIRC, but likely not. You may have to get a few to get the exact same board. I do not see any identifying marks on the board.
Best of luck
:cheers:
PS some maybe a different type but I don not recall any soldered, that would be an expensive op for mftrs.

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

From experience autopsying HDD (ok I just like taking things apart) it is all ribbon cable in a header. older might have some pin connectors IIRC, but likely not. You may have to get a few to get the exact same board. I do not see any identifying marks on the board.
Best of luck
:cheers:
PS some maybe a different type but I don not recall any soldered, that would be an expensive op for mftrs.

David – Sounds like you’ve been drinking. ROFL.

However, it does sound like you’ve been playing around with these controller boards… and I don’t think I’ll break anything if I take the 8 Torx screws out to have a quick look to see whether any connections are socketted or soldered.

Here goes!

Reply | Quote

David – Sounds like you’ve been drinking. ROFL.

Here goes!

I worked for a quasi gov agency for a quarter century how could I not?? :p

:cheers:

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

I worked for a quasi gov agency for a quarter century how could I not?? :p

ROFL

It turns out that my Torx set doesn’t go down in size to what I need so I’m going to try Satrow’s suggestion of HD Sentinel before proceeding further.

Reply | Quote

…is it possible to buy the exact same model of HD and transfer its controller board to the damaged HD to read the physical data (in order to copy to another drive)?…

Yes.

I have done that quite a few times. Most of the time it has worked at least well enough to get the data off the drive.

But you will need to make sure the drives’ model numbers match.

…The controller board of the HD appears to be secured by 8 Torx screws…I have no idea what lays beneath and thought I would ask before exploring further.

Either the controller board or the underside of the drive will have pins that fit neatly into a slot on the opposite part and/or spring-loaded contact arms that bear against gold-plated contacts on the opposite part.

…It turns out that my Torx set doesn’t go down in size to what I need…

I have searched many times for torx bits that will fit those screws but without success. In each case I have dealt with I used a small pair of short-nose pliers to (carefully) grip-and-loosen the head of the torx screws then wind them out using a suitable-sized flat-blade jewellers screwdriver.

If you manage to get access to the data on the drive by swapping the controller board then I would advise against trying to run SpinRite or HD Sentinel until after you have copied all (or at least as much as possible) of the data off the drive.

Reply | Quote

Yes.

I have done that quite a few times. Most of the time it has worked at least well enough to get the data off the drive.

But you will need to make sure the drives’ model numbers match.

Either the controller board or the underside of the drive will have pins that fit neatly into a slot on the opposite part and/or spring-loaded contact arms that bear against gold-plated contacts on the opposite part.

I have searched many times for torx bits that will fit those screws but without success. In each case I have dealt with I used a small pair of short-nose pliers to (carefully) grip-and-loosen the head of the torx screws then wind them out using a suitable-sized flat-blade jewellers screwdriver.

If you manage to get access to the data on the drive by swapping the controller board then I would advise against trying to run SpinRite or HD Sentinel until after you have copied all (or at least as much as possible) of the data off the drive.

42597-20151110_151945
The script is small but I believe it reads ‘Wiha’
Now that you have refreshed my memory it was mostly spring-loaded contact arms on recent models.

:cheers:

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

That Hddzone site has some really good info (it is in chinglish, but mostly understandable). Thanks again Tonyl
:cheers:

Of course I am a bit leery of the microwave advicehttp://www.hddzone.com/how_to_repair_the_damaged_hdd_ii.html:confused:

And now for the movie!
https://youtu.be/ZCGLNYyd03A?t=221

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

http://www.myharddrivedied.com/blog/why-spinrite-not-my-data-recovery-software-list

http://www.onepcbsolution.com/

PCB Solutions: “If the BIOS can see the drive at all, then the card does not need to be replaced.”

Reply | Quote

Hi Satrow – Many thanks for your reply.

There’s neither any smell nor noise whatsover, no matter how long I leave the drive connected and the PC powered up. It sounds (and feels, when I put my hand on it) like it spins up immediately without any problem and I haven’t heard any speed fluctuations once it spins up. It’s a guess at this point but I think the issue may be totally electronic, not mechanical.

There appears to be a contact/friction mark on the front of the drive’s original plastic casing but the owner swears blind he hasn’t dropped it or knocked it. I’ve listened very carefully and there’s absolutely no sound of any ‘clicking’ that would indicate undue activity of the HD’s actuator arm.

Reply | Quote

Sounds like the best option would be to send the drive to a professional data recovery service.

I agree… but they’re pensioners who don’t have the money to spend hundreds and hundreds of pounds on the off-chance they’ll be able to recover their photos.

If you can’t help with either practical advice or personal experience then that’s fine.

Reply | Quote

Hmm. See what SMART attributes you can grab with HD Sentinel Rick, you might be able to setup the BIOS to take more time over drive detection? Turn off SMART in the BIOS and swap your secondary drive with the dodgy one (set the dodgy one as #3)?

Attribute #5 is sector reallocation count (or it usually is!) – bad blocks with no spare ones to reallocate – along with the difficulty of getting into Windows with it, it might mean that some recent damage has struck the first few sectors, where most of the baseline data is stored (MBR etc.).

Reply | Quote

My best practical advice is that the drive could be damaged further by tinkering, and could eliminate any chance of data recovery entirely.

I agree completely and have already discussed this with the drive’s owner so he is constantly aware of any risks due to any rescue attempt that I make on his behalf.

You could pay for it yourself…sort of “paying it forward”.

ROFL

Reply | Quote

@ Rick

Further info.

I just checked through my stack of old HDDs and one is a 500GB Seagate ST3500418AS (seems to be same model as in the first screenshot in your #1 post). It has a sticker on the PCB “PCB failed” which I placed there, obviously after I had retrieved the user’s data by swapping the PCB w/ another HDD.

Also, as a hint, in many cases when customers had this problem I searched my customer database for the drive’s model number and often found where I had fitted a same-model HDD to another customer’s PC. In such cases I would offer to fit a larger drive to the “donor’s” PC – in all cases they accepted (why not? a new larger capacity HDD at no cost?).

Reply | Quote

in many cases when customers had this problem I searched my customer database for the drive’s model number and often found where I had fitted a same-model HDD to another customer’s PC. In such cases I would offer to fit a larger drive to the “donor’s” PC – in all cases they accepted (why not? a new larger capacity HDD at no cost?).

That’s a GREAT idea! Not only did you save the customer with the bad hard drive, but you created a lot of good will with the customer who “donated” the old hard drive!

Group "L" (Linux Mint)
with Windows 10 running in a remote session on my file server

Reply | Quote

@ Rick

Further info.

I just checked through my stack of old HDDs and one is a 500GB Seagate ST3500418AS (seems to be same model as in the first screenshot in your #1 post). It has a sticker on the PCB “PCB failed” which I placed there, obviously after I had retrieved the user’s data by swapping the PCB w/ another HDD.
.

Ok some times it works, but a surer way is to make sure the controller chip/board is the same as I said. Follow the link inhttp://windowssecrets.com/forums/showthread//172730-Is-it-feasible-to-replace-an-HD-controller-board?p=1030400&viewfull=1#post1030400 from Tonyl. Lots of good info. Thanks Tonyl

:cheers:

Seehttp://www.hddzone.com/seagate-pcb-c-1.html

PS you may need that magnifying head set!;)

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

I have a dismantled Seagate drive with me and the board has not solder or pins, just 20 pads and 3 pads that are pressure connected – a screw goes through the middle of the 20 pads to hold the board onto the connector. I needed a T6 driver to remove the screws.

A satrow said, attribute 05 indicates the platters have more errors than the drive can handle and I don’t think a controller will fix that – my dead Seagate has a similar problem but I replaced it when I saw the number of errors increasing and I still couldn’t recover much, used a backup.

Maybe we could all chip in a few shekels to pay for a copy of Spinrite etc?

cheers, Paul

Reply | Quote

…attribute 05 indicates the platters have more errors than the drive can handle and I don’t think a controller will fix that…

Any data read from or written to a modern HDD must necessarily pass through the HDD’s “controller board” (PCB – Printed Circuit Board). If the PCB is faulty then it is likely that is the real cause of the 05 errors.

At least one of Rick’s earlier posts says the HDD seems to spin up immediately, there are no “clicking” noises, and Rick doesn’t mention any overheating of the HDD.

Reply | Quote

It sounds like a couple of you have had success transferring controller boards and that the boards are only pressure fit, not soldered. I’ve spoken to the owner and he’s asked me to track down an identical drive. Many thanks for the feedback, everyone.

Reply | Quote

Worst case scenario, I think, is that it doesn’t work and you have to butcher the new board for parts. Provided you can figure out which part you need, of course. It depends how handy you are with a soldering iron. Have a quick read here.

Reply | Quote

I’ve been shopping and now have a Torx drive bit that fits the 8 screws on the HD’s controller board. I checked a couple of them and was surprised to find they didn’t need any effort to loosen them. I was expecting to find them really tight. (Hmmm… assumptions, right?)

Before I do anything else I’m going to see why I couldn’t get SeaTools for DOS to work from a USB flash disk then I’m going to check out Hard Disk Sentinel for DOS (suggested by Satrow, thank you) as I can’t get Windows 7 to boot whilst the damaged drive is connected.

Reply | Quote

… I can’t get Windows 7 to boot whilst the damaged drive is connected.

In case the BIOS checks are freezing the damaged drive (and the boot sequence) out, have you tried plugging it in after Windows starts? I had a feeling I had done that with my SATA internal caddy, but on trying just now it didn’t recognise the drive, until I went into Device Manager and selected ‘scan for h/w changes’ – and up it popped. I’m not sure that standard SATA connections are supposed to be hot-swappable, but if it works…

Reply | Quote

In case the BIOS checks are freezing the damaged drive (and the boot sequence) out, have you tried plugging it in after Windows starts? I had a feeling I had done that with my SATA internal caddy, but on trying just now it didn’t recognise the drive, until I went into Device Manager and selected ‘scan for h/w changes’ – and up it popped. I’m not sure that standard SATA connections are supposed to be hot-swappable, but if it works…

If you do that you need to check bios to set that sata port a hot startable

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

If you do that you need to check bios to set that sata port a hot startable

That may be necessary for proper operation, but I am pretty sure my 9-year old BIOS doesn’t have that function, but it still worked.

Reply | Quote

FWIW, I once had a HDD controller die (still under warranty). I solved the data-recovery problem by buying another HDD (possibly a 2nd-hand one) of the same make & model & swapping the controllers over. Having recovered the data, I swapped them back & returned the faulty one for replacement. I still have both drives in a now-dormant PC.

Cheers,
Paul Edstein
[Fmr MS MVP - Word]

Reply | Quote

When the damaged hard disk was connected I found it impossible to boot into Windows… the PC would just hang at “Starting Windows”. I also tried using Ubuntu and Linux Mint live CD’s but neither was able to mount the hard disk. This made me think that the problem may be due to the drive’s PCB board.

I removed the PCB and found that the 3 contacts for the motor (1) looked clean but the 2 rows of contacts for the actuator (2) looked dirty.
42640-hd13
Click to enlarge

I used a soft eraser to clean the contacts but this didn’t help when I reattached the PCB and tried booting into Windows with the damaged drive attached.

I still wasn’t able to boot into SeaTools for DOS so had a second go at creating a DOS-bootable USB stick using RUFUS. This worked and I was able to copy the SeaTools for DOS executable over to it and boot the PC from it with the damaged hard disk attached.

A Short Test failed immediately with the message DST — FAILED – Unknown. As this wasn’t very informative I tried a Long Test. This failed after a few seconds with the message The drive is not responding to commands. Check cables and drive power connection. I changed the SATA data cable as a precaution but a second run produced the same error.

I downloaded the DOS edition of Hard Disk Sentinel and added it to the bootable USB stick. This gave me far more information (Thank you Satrow for the suggestion):
42641-hd12
Click to enlarge

I was a little bit confused why Seagate’s own utility suggested a control problem (which could have been electronic) whilst Hard Disk Sentinel appeared to show quite clearly that the problem was a physical one, i.e. damage to the surface of one or more platters.

As the drive was now actually being recognised I updated the drive’s firmware from CC34 to CC49 in case the problem was corruption (instead of/as well as physical damage). The firmware updated without a problem but further tests with SeaTools and Hard Disk Sentinel showed no change in the results.

I’m now researching data recovery from DOS… so if anyone has any suggestions – preferably from practical experience – then I’m all ears.

Reply | Quote

…I’m now researching data recovery from DOS… so if anyone has any suggestions – preferably from practical experience – then I’m all ears.

I very much doubt you’ll find a free data-recovery program that will do what you want.

During the past 20+ years I have tried several data-recovery programs to recover userdata from failed/failing customers’ HDDs. I had some success with “Ontrack EZRecoveryPro” but much better success with “Active File Recovery”: http://www.lsoft.net/

I purchased the “Enterprise” version. Have had best success by creating the bootable CD, then boot from the CD w/ the sick HDD connected to the computer. AFR should list the sick HDD and allow you to scan for recoverable files/folders. AFR also allows you to specify a location (on another drive) to save the recovered data.

Reply | Quote

As the drive was now actually being recognised I updated the drive’s firmware from CC34 to CC49 in case the problem was corruption (instead of/as well as physical damage). The firmware updated without a problem but further tests with SeaTools and Hard Disk Sentinel showed no change in the results.

I’m now researching data recovery from DOS… so if anyone has any suggestions – preferably from practical experience – then I’m all ears.

May be just maybe you need to do the chip swap???

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

As the drive is now recognised, I think the usual way would be to buy a caddy to plug it into and then USB it into a working computer and also plug in an external HDD then copy across, but as I’ve never had to do this, other experienced members will probably be able to give you a walk through.

EDIT – Found this YT tutorial – https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CDEQtwIwAWoVChMIl5-85J6SyQIVzO8UCh1-7Qjc&url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DSLLQusrmzvI&usg=AFQjCNG-XQKwm4GNYqM4CvWaLXeuzdhu7g&bvm=bv.107467506,d.ZWU&cad=rjt

Reply | Quote

I’ve just tried that on 2 (Win 7) PC’s. On the first I get the following popup:
42644-hd14
Click to enlarge

On the second, the drive is recognised in Device Manager under Disk drives but doesn’t appear in Windows Explorer. When I check in Disk Management, no drives are detected and, instead, the status bar just shows Connecting to Virtual Disk Service… but goes no further.

Reply | Quote

Don’t suppose you still had the bootable USB connected when you tried this ?

Don’t know why one machine can see it but not the other and don’t suppose right clicking on it to see if an updated driver could be found would also work.

While 4GB of bad sectors is serious, if the electronics were working then it should still be recognised.

Can you hear it spinning or trying to ?

Does DISKPART/List disks recognise it ?

Reply | Quote

Don’t suppose you still had the bootable USB connected when you tried this ?

No, I didn’t have the bootable USB stick still connected.

While 4GB of bad sectors is serious, if the electronics were working then it should still be recognised.

My thoughts too… unless the damaged sectors include partition tables and/or boot sectors?

Can you hear it spinning or trying to ?

This is the quietest drive I’ve ever encountered. You can’t hear it at all. I have to hold it in my hand whilst connected to know that it has spun up.

Does DISKPART/List disks recognise it ?

DiskPart just hangs after showing version no, copyright and the name of the PC it’s running on.

Funnily enough, the second time I tried with Computer Management, I waited until it was open and showing the drives before connecting the damaged one in its USB caddy. This time the drive was recognised as an online disk but the moment I clicked on it Disk Management changed to Not Responding.

Reply | Quote

Perhaps a repair shop may have compatible electronic bits for you to get it up and running again sufficiently for you to recover the data – otherwise I can’t see you getting past that, but where you have brown contacts, that suggests a burn out somewhere.

Reply | Quote

Perhaps a repair shop may have compatible electronic bits for you to get it up and running again sufficiently for you to recover the data – otherwise I can’t see you getting past that, but where you have brown contacts, that suggests a burn out somewhere.

I can’t help but feel that I should be able to smell something off if there was a burn out somewhere… plus the output from Hard Disk Sentinel suggests it’s damaged sectors, not a problem with the PCB board.

I’m seeing the owner later today so I’ll soon know whether he wants me to keep on trying or accepts that it needs a data recovery company (and all the expense that will entail).

Reply | Quote

If the BIOS correctly recognizes the drive then it is not an issue with the PCB board.

But decent job of inspecting, documenting and cleaning up the contacts anyway. Interesting the three rightmost on both seem oxidized to a greater extent, yes? Have no idea what that means, but interesting.

Reply | Quote

Brown contacts usually come from the overheating of more current flowing through than should be.

I think I would pursue the repair shop before going the expensive recovery process as they will have to swap the circuit board anyway before they could recover the data.

Reply | Quote

Oxidation is usually caused by exposure to the atmosphere but there’s also something called Fretting Oxidation which is caused by movement between the contacts – I think.

Not sure if that would be possible with a HDD unless it may have been vibration, but I’ve seen a product called Deoxit recommended for cleaning oxidation, although a Google to Amazon, it looks pricey and WD-40 applied with a cotton bud may work equally well.

Oxidation will reduce the power going through those contacts which is probably the reason why it doesn’t respond.

Reply | Quote

WD-40 applied with a cotton bud may work equally well.

NO just PLAIN NO

Just because you don't know where you are going doesn't mean any road will get you there.

Reply | Quote

I agree, do not go anywhere near electronics with WD40.
To clean contacts rub them with a clean plain tissue, it is slightly abrasive and will clean the oxidisation.
Do not use tissues with extras, like lanolin, you will be sorry.

cheers, Paul

Reply | Quote

I’ve seen reports where oxidisation just returns when a protective coating hasn’t been applied to the contacts which deoxidising agents do and so would WD-40 – but I’ll bow to your experience if you are speaking from experience.

Reply | Quote

The contacts should either be tinned (a light solder coat) or be gold plated. Both resist oxidisation and shouldn’t need additional protection.
WD40 or similar leaves a lubricating residue which is not necessarily inert and may interfere with the connector making contact when re-assembled.

cheers, Paul

Reply | Quote

It’s recommended to clean off excess after cleaning, but I’m still wondering how a sealed unit can oxidise.

Reply | Quote

You might want to try TestDisk and treat it as an unformatted/reformatted/raw drive, not much point in trying to fix partition tables and other base disk id, I’m pretty sure they have a fixed location in the first few blocks – which appear to have been damaged and are probably unrepairable.

Reply | Quote