Destroyed hard disk

[Bevelhead]

I have a work situation that is worrying.
We have an old system running windows NT based program controlling a manufacturing machine.The tower and hard drive are about 10 yrs old and I was not aware the HDD was not copied as a back up..
Recently I turned the power off this computer and when the power was restored the operating system wouldnt initiate and the computer was asking for a boot disk to run a recovery.
When the boot disk was inserted into the cd rom, the computer wouldnt load it.
After changing BIOS preferences with no positive outcome,It was then decided to try the non working HDD into another working tower running NT to see if we could achieve a result and eliminate the original computer as being faulty.
The result was, it toasted the running computer.The computer gave off a burning odour and indicated a red flashing LED at the HDD.
After this we installed a new PSU into the second tower hoping this would restore the second computer.
All we got was a motherboard trying to boot then the motherboard cooling fan stopping and the system shutting down. Another tower was assembled from old components in the workshop to get this machine operating.
The old HDD was then installed back to its original computer and the system is still asking for a boot disk for recovery. Not surprised there but this unit hasnt burned like the other. So back to square one !!
Has anyone else had a similar experience and have any advice to getting a positive result?

 

[wdmfiber]

Too much to cut and paste here, but years ago I read a good article in Popular Mechanics.
I managed to find it online:
http://www.popularmechanics.com/technology/how-to/tips/4294038

In your case it sounds like money would not be much of a problem. You may want to outsource the work to a specialty data recovery service.
http://www.krollontrack.com/data-recovery/
Chat with them, a box will "pop up" at their home page.
See what they think and they're also give you a quote. It won't be cheap, but I suppose it would be a business write off.

 

[chrisso]

The experiences you describe are the unusual,. However. Getting a known good ide hard drive and installing again would likely solve your problem.
The mtbf on a pc hard drive running 8 hours a day is around the 5 years mark, so get a couple of spares for the rest of your pc's (that still work)

 

[Bevelhead]

Thanks for the fast replies folks. Especially the mean time between failures info. I wasnt aware of that statistic.
I cant wait to tell the boss and make him aware we need to back up all the HDD's in the plant.There must be about 20 PC's that probably dont have back ups. The environment the PC are housed doesnt help either. Very hot and dusty.
We can handle a hardware failure but we need that software and will have to do a forensic recovery on the hard disk. Interesting times ahead !

 

[Tom Tancredi]

To address the technical aspect: Your most likely correct the Drive (as you seen the other hardware) has failed and your lucky it last this long (BUSINESS STANDARD: Replace entire PCs ever 3 -Low Risk- years and no longer then 5 -HIGH RISK- years max). Depending on if you have the 'code' of the program controlling this manufacturing machine either be the original programmer, backup or on PAPER; then calculating the cost of the manufacturing machine being idle, does the cost of a Data Recovery company seem justifiable to extract that code, which MAY be corrupt (i.e. the program may never work even if recovered) anyway and a total loss and require the solutions below anyway.

To address the business aspect: Someone needs to be "retired" effective immediately because they just impacted your production and bottom line HEAVILY. Any RISK Analysis or better yet paying for a Business System Analysis Consultation would have pointed out this very thing years ago that was HIGH RISK for failure, with NO conceivable recovery methodology (are you sure NT, not 2000, XP or even Vista? If your still on NT, assuming 4.0 which was the last edition, that is WAY WAY past any support measures even Microsoft will not support that several years ago... and worse the type of hardware, IDE drives, hasn't been manufactured in years, you can't replace the hardware!).

IMHO I believe your company has failed itself in effect Business Systems Administration, and created a HIGH Risk to the bottom line that now your going to have to pay more then is necessary to resolve the issue, than if the initial cost to address this 'known issue' (support ended for the system time for upgrade) at the beginning.

Your best solutions would be
A) Is there a replacement system by the Manufacturer of that Manufacturing Machine? If so what is the cost including long term support and again, costs long term (since your apparent model for ALL devices, electronic or mechanical is more then 10 years) for the replacement PC side of the machine? What is the overall costs?
B) Is there a COMPETING system (system being the entire manufacturing machine, components, parts, PC, software, all of it) that performs the same function as this machine, and with the same support and long term costs, what is the overall costs?
C) Now invest in the cost for a BSA Consultant to come in (yes no matter what your gonna have to suck up this cost no matter what decision you make because your company FAILED to adhere to basic RISK Avoidance protocols) and with NO CODE at all, seek them to develop, install, backup and migration paths, copies in both data, long term OFFSITE STORAGE backups, and PRINTED copies of the code kept in OFFSITE STORAGE (think like your other important document in the Safe/etc.) with the long term (10 years + ) projection for upkeep and migration support, what would the estimated costs be for this Development project? Is there a third party solution software that can be modified to meet these same requirements thus driving down developmental costs? Again LONG TERM SUPPORT from that company possible? What are the risks and costs doing all this (say for example you fire the programmer, how can you get someone else to do this work)?
D) Estimated costs to remove the Manufacturing Machine and those product sales to the bottom line? Is the COST to MAKE the products HIGHER then the LOSS of NOT doing that business? Taking that amount of LOSS and comparing it to A, B or C, does it justify the expenditure.


ALL this I laid out SHOULD have been what the person in charge did YEARS ago and has a Business PLAN to address this very issue (just as much as if what happens if the building caught fire today and is a total loss? what happens if the watermain broke and flooded the building ruining all the electrical equipment - fried them - and now rusted all the manufacturing equipment? etc. etc. all part of a Risk Analysis).

 

[fzabkar]

@Bevelhead, could you at least tell us the model number?

Could you upload a photo of the component side of your HDD's PCB?

Does the drive still spin up?

 

[Bevelhead]

Thanks again for the taking the time to help me with this problem as Im located in Sydney, Australia.
Its great feeling to know that theres helpful assistance and sound advice from complete strangers from different parts of the globe as my experience with computer technology is remedial.
I should give you some background.
My true role at the plant is as an electrical technician on afternoon shift. Im the only one on that shift.
I repair the machinery if it breaks down and perform preventative maintenance. But computers are where I fall behind.The plant works 24/5 with preventative maintenance on the 6th or 7th day. At previous plants that I have worked at we always had a computer tech to keep up with technology but here we dont seem to have that and this falls back on to us, even though its a larger company.
This issue has arisen when the power was turned off for some PM. The circumstances that I believe led to this issue are the following.
1. HDD not copied or backed up.
2. PC hard drive in excess of 10 yrs old.
3.. Operating software unsupportable.
4. PC had a UPS in the cabinet not connected.
5. PC installed in a hostile environment ( very hot and dusty ie paper dust)
These problems were highlighted to the management but often not acted upon and now we pay the price. Ive posted here to find a solution and to help others to avoid this. Old adage is prevention is better than the cure.
The problem has been passed on to my supervisor and he's running with it now so i will post about its progress. Unfortunately I cant post a pic of the items until Monday. Maybe that will help too.
Thanks to everyone Pete

 

[fzabkar]

With any luck the problem will be a simple PCB fault, in which case the maximum repair cost shouldn't be more than US$50.

BTW, I'm not too far from you.

 

[Bevelhead]

'Ive just received a message from my colleague stating the original disk drive wont spin and an alternative hd backup wouldn't load either. I guess its time to try some new hardware first then try again.

 

[fzabkar]

A no-spin symptom is a much better proposition than a drive that spins but can't be detected. There is a very strong chance that a PCB swap will get you going again.

 

[BillyDataGuy]

'Ive just received a message from my colleague stating the original disk drive wont spin and an alternative hd backup wouldn't load either. I guess its time to try some new hardware first then try again.

Can you please explain this: and an alternative hd backup wouldn't load either.

Also it sounds like your drive has a short in it somewhere or an electrical failure, possibly a motor malfunction and needs to be checked by a professional or you will continue to destroy more electronics and never get your data back.

 

[fzabkar]

"Wouldn't load" is a vague statement. It doesn't necessarily imply that the "alternative backup" drive is faulty or has been damaged by the external hardware.

Before jumping to conclusions and engaging a "professional", most of whom couldn't measure a short circuit in an iron bar, I would be prepared to at least visually examine the original PCB and assist the OP in making relevant diagnostic measurements. (BTW, I'm an electrical engineer with several decades of experience in the component level repair of minicomputers and peripherals, so I guess that makes me a professional.)

 

[BillyDataGuy]

Yes it's vague that's why clarification was asked for on the "wouldn't load" statement.

With regards to "jumping to conclusions" it's really just jumping on the facts already provided, combining that with experience with drives behaving this way. Not claiming to have a PHd in electrical engineering.(When I was speaking of professional I was referring to someone who specializes in data recovery. ) Just going on what OP has already stated the drive has caused damage previously:

"The result was, it toasted the running computer.The computer gave off a burning odour and indicated a red flashing LED at the HDD.
After this we installed a new PSU into the second tower hoping this would restore the second computer.
All we got was a motherboard trying to boot then the motherboard cooling fan stopping and the system shutting down. Another tower was assembled from old components in the workshop to get this machine operating.
The old HDD was then installed back to its original computer and the system is still asking for a boot disk for recovery. Not surprised there but this unit hasn't burned like the other. So back to square one !!"

It's pretty clear there is some failure that is causing an electrical problem. Whenever that odour is present that means to usually stop putting power to the device, its fried.

Asking the OP to visually examine a PCB for after making the declaration that most "professionals" couldn't even measure a short circuit seems like a less than optimal approach to resolving OP's problem.

Sort of similar to implying that swapping out a PCB board is some easy magic cure for a drive that won't spin. I mean if we are just after getting it to spin, maybe it is, however just because it spins doesn't mean it is can be read. Most PCB transplants require the firmware chip to be transplanted as well in order to be usable, and even with that happening successfully .. getting the drive to spin up and then read is about 50% or less.

Seems like a lot of trouble to save a few hundred dollars and risk losing everything on there in the process. But what do I know. The data must not be that important.

Never the less the floor is yours fzabkar. Good luck OP

 

[fzabkar]

@BillyDataGuy, the OP still has not told us the model number of the drive, so there's essentially no information for us to work with. Is the OP's supervisor even capable of providing such information? If not, then perhaps you're right and "professional" recovery is their only recourse.

Assuming that relevant information will be forthcoming, my approach would be to examine the damage to the PCB and then determine whether this damage was the result of a fault on the PCB, or whether it was the result of an overvoltage from the PSU.

Knowing the model and brand would help us assess the likelihood that a PCB swap would fix the problem. Drives manufactured within the last decade will have TVS protection diodes on each of the +5V and +12V inputs. If either of these diodes are shorted, then this will confirm that the PSU killed the drive. Furthermore, if the +5V input sustained the overvoltage, then there will be a serious possibility that the preamp on the headstack will have been damaged, in which case professional data recovery will be the only option. Otherwise, if the +12V supply was the culprit, then chances are good that a PCB swap will fix the problem.

Since the OP claims to have electrical skills, then some simple electrical testing should be possible. A cheap digital multimeter would suffice. To safeguard the data, the PCB need not be mounted on the drive.

As for what constitutes a data recovery "professional", unlike the OP who no doubt has verifiable technical qualifications from a recognised institution, DR pros have no such official accreditation. In fact ANYONE can hang out a shingle and claim to offer DR services, and that's exactly what happens. I've been watching the professional storage forums for over 4 years and have seen amazing ignorance and ineptitude during that time. In fact there is a well known Data Recovery Expert course that claims to make anyone an expert in only 5 days, yet there are only two pages devoted to electronics. The author of the course doesn't understand the difference between a resistor and a diode, and that goes for the majority of the professionals who work on your precious data. There are DR pros who don't know how to use a multimeter (one person didn't realise that a board had to be powered in order for voltages to be present), others who can't recognise the difference between a MOSFET and a flash memory chip (they are confused by the fact that both have 8-pins and are in identical packages), still others who can't recognise resistors, capacitors or fuses even though they are marked with R, C, and F on the board. I've seen people who will take a good and bad board and then use a multimeter to compare the resistances of every component. They really have no idea how to perform even the most rudimentary troubleshooting. Most of them are just ex-IT keyboard punchers or PC board jockeys with exaggerated opinions of themselves. The DR profession is the definitive example of the Dunning-Kruger Effect.

Imagine you wanted to be a surgeon. You would go to university, study for 4 years or more, and then do an internship at a hospital. Only then would you be ready to go it on your own. OTOH, a DR professional would start by amassing a collection of cadavers, practice until he thought he was competent, and then hang out his shingle as an MD. Moreover, a DR pro who doesn't understand the difference between a resistor and a diode, but who can do headstack replacements, would be like a brain surgeon who didn't understand the difference between an arm and a leg. That said, there are some very competent DR professionals, but how do you find them, especially since everything they do is a secret?

 

[Tom Tancredi]

Guys, based on what has been said to date, I think your both missing a point here:

OP: "My true role at the plant is as an electrical technician on afternoon shift. Im the only one on that shift. "

So first off this is a one deep position, meaning he is busy all the time as he works solo. Second while he is a electrical technician, he may not be nor even have the equipment necessary to work on a device like the HDD, but more for heavier machinery (think scale here) and thus the tools 'on hand' may not even suffice to do 1/10 of the exploring methodology your saying.

Lastly, it is HIGHLY apparent that this is a environment that is focused on 'mechanical' then 'electrical' solutions (i.e. where can I tighten my wrench, not how do I measure the Ohm calibration), so they ignore things (this PC, basic Risk Analysis, etc.) till they break then they go 'buy another' and carry on because things all should last 20-50 years like they always have.

This is a HIGH LEVEL decision, one the 'Master's that be' will decide what to do, not the OP. As I pointed out, the code on it, even if recoverable, may not be usable and worse can't be just 'loaded' onto another PC and back to work. There is a major problem here of Management Failure to assess its assets, liabilities and risks to ensure minimal to zero downtime, and now they will have to play catch up. So at this point, whether the drive needs to be 'DR' or the PCs themselves are just DOA, doesn't matter as there is no possible 'PLAN B' to resolve the issue (no computers are sold with IDE drive chipsets, or connectors, and NT is NOT supported nor even purchasable, much less 'what' did the 'code' actually 'say' to the Mechanical Machine no one has a clue!).

As Bill Gates told IBM when they showed him the first Personal Computer, "is that all it does? Blink lights?", it isn't just the hardware (mechanical) that makes things work anymore, it is the software.

 

[Guest]

...

 

[BillyDataGuy]

@fzabkar LOL .. you are actually dead on, and there is no way I could have said it better.
@Tom very true observation + analysis ... however I did want to add that where there is a will there is a way, and yes there are bones still left in the dinosaur graveyard (no computers are sold with IDE drive chipsets, or connectors, and NT is NOT supported nor even purchasable), and since antiques are exempt so are outdated operating systems . Just saying if the price was right anything is possible.

Never the less ...

 

[USAFRet]

IMHO, this is no longer a recovery situation, but a replace situation.

We have a 1 deep non-PC person, trying to repair a 1 deep decade+, badly neglected PC that has failed in multiple stinky ways.
How much is the company losing every day, while this guy flails about trying to source and replace a PCB on an ancient drive. With zero guarantee of it even booting, and far, far less guarantee of it actually working to prefail condition.

Assuming this is a CNC controller....would you trust it to output correct parts? I wouldn't.

Time to replace with something from this decade. With actual IT support.
And all that which comes along with "IT support".

 

[USAFRet]

IMHO, this is no longer a recovery situation, but a replace situation.

We have a 1 deep non-PC person, trying to repair a 1 deep, decade+, badly neglected PC that has failed in multiple stinky ways.
How much is the company losing every day, while this guy flails about trying to source and replace a PCB on an ancient drive. With zero guarantee of it even booting, and far, far less guarantee of it actually working to prefail condition.

Assuming this is a CNC controller....would you trust it to output correct parts? I wouldn't.

Time to replace with something from this decade. With actual IT support.
And all that which comes along with "IT support".

 

[fzabkar]

Just to explain why I have to vent occasionally ...

Imagine if a data recovery "professional" contacted you privately and asked you how to partition and format a hard drive. You would be appalled, wouldn't you? In fact that's the level of question that I regularly get in my inbox, although of an electronic nature.

For example, a few weeks ago I was asked how to differentiate between a resistor and a diode. Ironically this person (who seemed like a nice guy) had only recently completed the 5-day expert course that I alluded to previously. He counted a major US car company as one of his data recovery clients, and he had a business that had sold more than ten thousand computers, yet he had never used a multimeter. Instead all his "repairs" had been effected by replacing FRUs. Now he was attempting a data recovery from a USB flash drive. Two minutes with a multimeter would have told him that his mystery component was a resistor (in any case diodes have a polarity stripe). The flash drive had two chips, the flash controller and a NAND flash memory. I located a datasheet for the controller and a circuit diagram (which showed the resistor) for the drive. His next question was, which of the two chips in the circuit diagram was the NAND flash? God give me strength ...

Getting back to the original problem, ISTM that updating the technology would be advisable. However, there still remains the problem of recovering the data. In most cases you can't get a much easier fix than a board swap, if that is indeed what is required. There are times when you will need to transfer the firmware from patient to donor, but some PCB suppliers will include this service at no additional cost.

Depending on the capacity of the drive, and assuming it is an IDE type, it may be possible to substitute it with a CompactFlash card. It may even be possible to replace it with a recent SATA drive via an appropriate adapter. If the drive capacity is too large, it could always be reduced by way of a HPA. For example, you could use a tool such as HDAT2 to reduce a 2TB drive to 128GiB, if it were necessary for compatibility reasons.