L David Tomei
Well-known
- Local time
- 10:14 PM
- Joined
- Dec 3, 2007
- Messages
- 363
The voltage, the type of plug and polarity of the contacts are the only important things. If that's known , then almost any DC power supply can be used.
What to do about failed external hard drive
- Thread starter Rob-F
- Start date
- Latest activity Latest activity:
- Replies 48
- Views 5K
Rob-F
Likes Leicas
Right. But in this case the power supply is the good part. Something has gone dead in the drive itself.
Sparrow
Veteran
I imagine you've checked for a blown fuse?
Sparrow
Veteran
... back in the days when a pair of 40mb Winchesters were state of the art my technician once used a Dremel to run a duff drive up to speed and get the data back
Vince Lupo
Whatever
I used their services a couple of years ago to recover an external hard drive of mine that went belly up. It cost me $1600, but I got everything back.
craygc
Well-known
Well you're only part way there ...a SATA drive has 3 power taps, one at 3.3V, a second at 5V and the third at 12V. The 3.3V is usually never used, the 5V is for the circuitry and the 12V is only for the platter and head motors.
EDIT: Ooops ...just realised you're actually referring to the power brick of the enclosure and not the plug to the physical drive.
wpb
Well-known
Often recovery services don't remove the platters; they find an identical drive and swap out the hard drive controller. If your drive spins up but makes "ticking" noises this is often due to the controller malfunctioning. The controller (the circuit board attached to almost all HDDs) can also sometimes be brought back to life if chilled in the freezer. Crunching or grinding is really the worst noises from a HDD as the read/write head or platters are physically damaged- that requires real data recovery expertise.
BobYIL
Well-known
Data safety + performance..
Data safety + performance..
Most of us are using external hard drives for extended capacity; and for the safety considerations multiple units are used with overlapping contents. There is also a trend for employing SSDs as C drive and paying some substantial money per $/GB basis. I would suggest RAID10 configuration to be built with conventional components (HD only) and plausible cost as long as your motherboard complies to.
Provided that you have employed proper components:
- it will cost you not more than an SSD + an HD
- extreme fault tolerance, highest data protection.
- excellent speed, highest performance (no need for an SSD).
RAID10 requires min. 4x HD. I've built one with 8, costed me around $600 for the RAID but at the end not more than the same PC with SSD+ 1xHD of the same total capacity.. Now I do not even need to concern about which one of the HDs would fail..
Data safety + performance..
Most of us are using external hard drives for extended capacity; and for the safety considerations multiple units are used with overlapping contents. There is also a trend for employing SSDs as C drive and paying some substantial money per $/GB basis. I would suggest RAID10 configuration to be built with conventional components (HD only) and plausible cost as long as your motherboard complies to.
Provided that you have employed proper components:
- it will cost you not more than an SSD + an HD
- extreme fault tolerance, highest data protection.
- excellent speed, highest performance (no need for an SSD).
RAID10 requires min. 4x HD. I've built one with 8, costed me around $600 for the RAID but at the end not more than the same PC with SSD+ 1xHD of the same total capacity.. Now I do not even need to concern about which one of the HDs would fail..
BobYIL
Well-known
First take the HD out of the housing. Power the housing and proceed as follows:
Where do you measure this 12VDC? If it's IDE type, then there is a 4-pin supply socket inside the housing, yellow 12VDC, red 5VDC you must measure. If SATA type then measure the same not on the narrow data socket but on the wide power socket. If the voltage could not reach to the sockets as stated then a. short circuit the switch b. test all 0 ohm resistors (fuse resistors) from the input of 12VDC.. (follow the +) Test them as diode.
Depending on your findings I will try to assist you..
kuzano
Veteran
Nature of the failure
Nature of the failure
The majority of the time, drive failure is not complete. If it still spins up, the motor is working. If the shock did not cause the write/read heads to score one or more of the data platters (the original source of the term "hard drive crash"), then a reasonably astute tecnical person can simply hook the drive up to another computer and retrieve the data or other files from the drive. That's what your wife's friend did.
No, that is not worth $1000 and only takes a matter of minutes to see if it can be done.
As Chris noted, the $1000 price tag comes in when the motor in a hard drive fails completely and the drive does not spin up, as we say in techy words.
Then the drive must be competely disassembled and the platters moved into a bench drive that has the same motor and circuit board as the original. Being equipped to do that for all the drives out in the world is very costly and the process is time consuming. It also requires a very expensive environment (clean room and hair nets and such). That is costly.
However in many companies, corporations and banks, or where the livelihood of a person depends on the drive contents, the price is small at $1000.
Most failures, as I said before are not total drive failures. You often get warnings of impending failure. However a total drive crash gives no warning and is instantaneous. It is simpy a mechanical failure of the drive. Dropping a drive from a few feet up to a concrete surface can cause slow impending failure, or it can shock/break the drive motor, read/write assembly, or circuit boards.
You were lucky. You'll stay luckier if you have three or more fully synchronized hard drives. (full sets of data, and don't waste time on full HD backups. No one really makes one that will swap into a computer and be up/running in minutes. Only redundant RAID can do that reliably)
You can duplicate a drive in your system with cloning software, but that's really just a software solution that acts a bit like RAID. I do it sometimes, and store the second drive or unhook it to get a system running again in short order. However, it's not ever truly up to date.
This has been a real bugaboo in data security for many years. There are a thousand pretenders to data security in the marketplace. Never seen one that is strong and secure. Usually some weakness. Why.... those systems are designed and written by Human Beings.
The closest you can come to secure data is:
1) Plan for, in fact, expect failure.
2) Reduntant... many levels of redundancy
3) Off site duplication storage
4) Forget the "cloud". Who do you call when the web site disappears over the weekend. One that purported to be for Pro Photographers did just that. There have been others. Have any of you examined the financial statements of web sites that say they can protect your images. Will they succeed in the internet market?
5) Who is ultimately responsible for your data. YOU... not even your wife or kids.
My personal take on LaCie is that they are just a marketing company using branded drives from various manufacturers. Their history is no better or worse than the drive that is inside. How they handle their drives may be another story.
I do know, from gutting a broken LaCie that they have used Seagate, and am told other hard drives. There are really a very small number of surviving actual manufactures of hard drive. Offhand, I can think of Seagate, Western Digital, and Hitachi. Possibly Fujitsu.
In fact if you google or yahoo LaCie hard drive reliabilty, you may not be so pleased. Your only solution is build true redundancy in your backups.
Would you have lost images in the recent debacle. If yes, perfect the system. Drive reliability won't truly protect you. Redundancy will. And forget the Cloud... they don't guarantee anything.
True reduncancy is two complete sources at home, and not in partitions on the same drive, or second drives in the computer. Then the final step is one more completely redundant system off site (outside the home--safe deposit, whatever)
Don't lose too much sleep simply because I would never buy a LaCie. Get the system in place and it won't matter. There are NO truly reliable drives on a long term basis. Current time before you are pushing a drive is 4-5 years. And that's only estimated life span on the bearings inside. Motors, circuit boards, and platters all have their own MTBF, which comes from the results of testing on mfr... it means "mean time between failures".
25 years consulting, teching and WMAO on PC's. (No MACs/Apples allowed. If someone comes to my door, driving up in a SAAB, wearing striped T or polo shirts, Plaid knee length shorts or 10 pocket cargo shorts [calf length], birkenstocks, and a pocket protector, I presume He/She has an Apple product and I don't answer the door. No money in it. When I did try to do MACs/Apples, I got no calls. I could seriously relate to the Maytag repair man. Now you know why they are so Freaking expensive up front.)
Nature of the failure
The majority of the time, drive failure is not complete. If it still spins up, the motor is working. If the shock did not cause the write/read heads to score one or more of the data platters (the original source of the term "hard drive crash"), then a reasonably astute tecnical person can simply hook the drive up to another computer and retrieve the data or other files from the drive. That's what your wife's friend did.
No, that is not worth $1000 and only takes a matter of minutes to see if it can be done.
As Chris noted, the $1000 price tag comes in when the motor in a hard drive fails completely and the drive does not spin up, as we say in techy words.
Then the drive must be competely disassembled and the platters moved into a bench drive that has the same motor and circuit board as the original. Being equipped to do that for all the drives out in the world is very costly and the process is time consuming. It also requires a very expensive environment (clean room and hair nets and such). That is costly.
However in many companies, corporations and banks, or where the livelihood of a person depends on the drive contents, the price is small at $1000.
Most failures, as I said before are not total drive failures. You often get warnings of impending failure. However a total drive crash gives no warning and is instantaneous. It is simpy a mechanical failure of the drive. Dropping a drive from a few feet up to a concrete surface can cause slow impending failure, or it can shock/break the drive motor, read/write assembly, or circuit boards.
You were lucky. You'll stay luckier if you have three or more fully synchronized hard drives. (full sets of data, and don't waste time on full HD backups. No one really makes one that will swap into a computer and be up/running in minutes. Only redundant RAID can do that reliably)
You can duplicate a drive in your system with cloning software, but that's really just a software solution that acts a bit like RAID. I do it sometimes, and store the second drive or unhook it to get a system running again in short order. However, it's not ever truly up to date.
This has been a real bugaboo in data security for many years. There are a thousand pretenders to data security in the marketplace. Never seen one that is strong and secure. Usually some weakness. Why.... those systems are designed and written by Human Beings.
The closest you can come to secure data is:
1) Plan for, in fact, expect failure.
2) Reduntant... many levels of redundancy
3) Off site duplication storage
4) Forget the "cloud". Who do you call when the web site disappears over the weekend. One that purported to be for Pro Photographers did just that. There have been others. Have any of you examined the financial statements of web sites that say they can protect your images. Will they succeed in the internet market?
5) Who is ultimately responsible for your data. YOU... not even your wife or kids.
My personal take on LaCie is that they are just a marketing company using branded drives from various manufacturers. Their history is no better or worse than the drive that is inside. How they handle their drives may be another story.
I do know, from gutting a broken LaCie that they have used Seagate, and am told other hard drives. There are really a very small number of surviving actual manufactures of hard drive. Offhand, I can think of Seagate, Western Digital, and Hitachi. Possibly Fujitsu.
In fact if you google or yahoo LaCie hard drive reliabilty, you may not be so pleased. Your only solution is build true redundancy in your backups.
Would you have lost images in the recent debacle. If yes, perfect the system. Drive reliability won't truly protect you. Redundancy will. And forget the Cloud... they don't guarantee anything.
True reduncancy is two complete sources at home, and not in partitions on the same drive, or second drives in the computer. Then the final step is one more completely redundant system off site (outside the home--safe deposit, whatever)
Don't lose too much sleep simply because I would never buy a LaCie. Get the system in place and it won't matter. There are NO truly reliable drives on a long term basis. Current time before you are pushing a drive is 4-5 years. And that's only estimated life span on the bearings inside. Motors, circuit boards, and platters all have their own MTBF, which comes from the results of testing on mfr... it means "mean time between failures".
25 years consulting, teching and WMAO on PC's. (No MACs/Apples allowed. If someone comes to my door, driving up in a SAAB, wearing striped T or polo shirts, Plaid knee length shorts or 10 pocket cargo shorts [calf length], birkenstocks, and a pocket protector, I presume He/She has an Apple product and I don't answer the door. No money in it. When I did try to do MACs/Apples, I got no calls. I could seriously relate to the Maytag repair man. Now you know why they are so Freaking expensive up front.)
L David Tomei
Well-known
- Local time
- 10:14 PM
- Joined
- Dec 3, 2007
- Messages
- 363
Thanks for the advice. It's up and running with the image files recoverd. However, everything is backed up on Carbonite.
Rob-F
Likes Leicas
I got it out of the case and found the socket with two black leads, a yellow lead, and a red lead. I got no voltage reading from red to black and none between yellow and black (or between any two leads.) Same thing after pressing the switch. I see some resistors--the smallest I ever saw--on a small circuit board, the same one the switch is on. Should I go ahead and measure them with a DVM? Or with my Simpson 260XL on the low-power ohms range? They are tiny! I hope I can get my test leads on them! And I hope they are the right ones to check.
BobYIL
Well-known
Hmmm, then your HD is IDE.. Did you short the switch? Check either the LED will light up and the voltage will come to the leads (or not). (Better check the switch on diode mode first when the device not connected to power..)
Should the switch intact then check the tiny resistors one by one. Any DVM is OK..
(Is the pcb powered by a external power supply or in-case power supply? Any picture of the inside of the case?)
I do not think that your HD is powered properly. Trying to find out the architecture (design) of the power circuitry. A picture would help a lot as there are different configurations for them.
Rob-F
Likes Leicas
The switch checks good in diode mode. Shorting it with power on does not bring the unit alive. I'm trying to check these little resistors. They are smaller than my meter probes! I haven't found a bad one so far. There are evidently some capacitors that look pretty much like the resistors. Still looking around . . .
BobYIL
Well-known
Since you did not have the voltages I have stated, the problem seems to be with the power supply section of the HD.. we need to "see" that 12 and 5 volts at the leads, then the HD will be checked if its intact or not..
BobYIL
Well-known
Do you use desktop or laptop? If desktop then try with an IDE socket out of the main power supply to test the HD portion.. First look at the back of the HD portion you took out of the external drive, there is a jumper socket (4x2) by the IDE socket of the HD but only ONE jumper.. if it's not on "slave" position then place it on "slave" (not on "master" or elsewhere. BTW, if there's no jumper use it as it is). Do these while the power was switched off. Then connect the HD to IDE socket and also to a standard 4-lead (yellow-red) power socket. Attention: If on the mainboard the IDE connection was only one, take off the IDE cable to the DVD-ROM and use it for a temporary test. Now you can turn on the PC.. Check whether it would give any Boot-disc error; if it does take off the jumper.
(Then advise me what happened..)
(Then advise me what happened..)
MartinP
Veteran
Buy another hard-drive caddy and try the drive in there. They cost very little and if you find the hard-drive itself is dead, then you will probably want to put another disk in the caddy and use that one anyway - so the small cost is not wasted. Using hours of time to diagnose an non-repairable component is not sensible.
This is not the place to get reliable advice explained at length, so that you can understand what you are trying to do and the ways developed in industry over decades to solve these problems.
All hard-drives, and caddies, and PC's (and Mac's) etcetera, etcetera, come with a 100% guarantee of failure. The only question is when. Remember, 100% - ALL of them.
Please put these three paragraphs together and realise that you would do well to ask around your locality for a specialist shop or contractor to help you out! 🙂
This is not the place to get reliable advice explained at length, so that you can understand what you are trying to do and the ways developed in industry over decades to solve these problems.
All hard-drives, and caddies, and PC's (and Mac's) etcetera, etcetera, come with a 100% guarantee of failure. The only question is when. Remember, 100% - ALL of them.
Please put these three paragraphs together and realise that you would do well to ask around your locality for a specialist shop or contractor to help you out! 🙂
BobYIL
Well-known
It takes 5 (five) minutes once you learn how to check it. Also for an FCC licensed radar tech to learn such things, IMO, should be fun, not trouble.
Rob-F
Likes Leicas
I want to know more about the "hard drive caddy." Does this mean another identical chassis, but without a hard drive in it? And whatever it is, where do I get one? Online? Are they standardized so that any hard drive fits any caddy? Or what?
Thanks!
Rob
neelin
Established
I've recovered a drive that would not read by putting a small (1lb) block of dry ice directly on the internal HDD drive case. Whatever it did, it allowed the drive electronics, spindle, whatever to work long enough to recover all my data. The machine was off, and I turned it on a couple of minutes after it was cooling the HDD case, leaving the dry ice in place all through the operation.
robert
robert
Similar threads
- Article
- Article
