Lacie Hard drive failure

[Gosub]

Ok,

Friends hard drive failed but now makes no noise at all so I said pop it over

The Drive is a Samsung Model HM320ii in a Lacie powered external enclosure.

Taking advantage of my new Bench power supply, I set the current to the minimum load of 2A which is 200mA under the dedicated power supply for this device.

This is where I learned how important it is to have some of theses current limiting bench power supplies is.

1. I worked out where positive and negative were oriented on the chassis by doing a quick examination and then confirming with a continuity check
2. After setting the correct voltage and current, I connected my power supply to the correct rails on the chassis power board.
3. Immediately after enabling the powered circuit I observed that the voltage dropped to 1.2V and the current had ceilinged to my pre-set of 2A.
Something was shorted or blown
4. I detached the drive from the chassis and powered the chassis drive free, same result.
I deducted that the short was on the logic board as I got the same result with or without the drive attached.
5. Detached the Logic board from the drive and found this



Damn, Damn

What would cause this and what part has blown, has the memory been damaged?

What would you guys suggest?

I'm thinking is buy an identical drive and retrieve the data as quick as possible , BUT

Has the drive been damaged or if not, is the drive going to blow a new logic board

Any ideas are welcome

[German_EE]

It is almost certain that the blown component is a decoupling capacitor on the power supply rail so for starters you need to extract the blown component, clean the PCB and remove all char and carbon then solder in a replacement capacitor. The fun will be finding out the component value, I suggest 100 nF as this value is quite common in this position.

Then,cross your fingers and test with the hard drive connected.

Edit: A friend has just suggested that the capacitor may be much larger in value because of the physical size, maybe 1uF.

[Gosub]

I took a picture of the PCB revision and here is the picture



From this I found what the blown component should look like



If you ride up the letter t from the word 'store' until you hit the Memory chip you'll see the possible cap that's blown.

Can you identify this chip?

Answers on a postcard

[Gosub]

[wraper]

Does not look like capacitor, rather TVS diode. Modern hard drives don't contain tantalum capacitors, which have similar look but different leads. As there is big MLCC in parallel, likely this trace is directly connected to +12V input. Removing this TVS most likely will make HDD operational again.
If drive works again, check that enclosure if supply voltages are correct and there are no dead electrolytic capacitors.

[Gosub]

I removed the blown component and get the same high current so maybe something else is wrong here, a visual inspection shows nothing.

I see someone is selling the same drive but working on Ebay. I think I should buy that part and use this as a learning bag for me seeing that I am new to electronics

[wraper]

Then I guess enclosure put overvoltage to HDD. Likely motor driver IC is shorted. However other parts might be dead as well. Main IC contains firmware, therefore replacing it might not give you desired result. I repaired HDD only a few times, and not sure about particular Samsung. But in many cases it is impossible to repair  HDD when replacing the whole board as data stored on the board and platters are dependent. Need to note that you won't be able to replace motor driver and main ICs without hot air or IR rework station as they have thermal pad under them.

[Gosub]

Thanks for the inputs,

I will buy the new drive and see if I can get data from the platters under low current conditions and then play with this more

Thanks to all the replies guys, really appreciated

[wraper]

If you try to replace the board, buy hdd with the same firmware version and revision if it is visible on the label. Also do some research with Google if replacing PCB works on this model. Also worth to look on the date code and get closest. Manufacturing place must be the same.

[poot36]

If you want to try and find any other shorted components try using a hi rez ohm meter on the board.  If you do not have that use a low voltage 5 to 7 V 1A or so power supply that is not current limited and will not shutdown under a short (Try a small transformer based wall wart) and feel which components are getting warm or hot and remove them and test them out of circuit for shorts.  I managed to partially fix a laptop motherboard this way.

[Timmay]

For me I learned these lessons in the early 90s. Back-up, back-up and back-up. I always back-up to what I call a shelf drive. As in, I back to it and never use it but to check the back-up or write to it. As a photographer I have three back-ups of my data. I never use my computers main drive for anything but most used programs, games and OS.

As for repair it could be a logical failure in which case you need to use a HD recovery service. Really nothing else to say here,

Mechanical failures could be the read-write head, board or other components. You could replace these parts (the board especially and many times its easier and faster to do this). Other mechanical failures are generally beyond the repair of mere mortals.

Physical damage could be heat, water, shock, etc.This could be un-recoverable and your data could be lost forever. Once upon a time I had a head hit and damage the disk, but using it externally I was able to copy the entire drive and babysit it and give input to every error that came up and got most of my data back, but not all.

MOST of the time when friends send me "bad" hard drives I replace the board and it works, not all the time but I have had a lot of success with this. I also recommend betting the entire similar drive, swapping boards and if it doesn't work you still get a hard drive out of it and if it does you can buy the board and you get two drives out of it (drunk logic )

[BradC]

If you try to replace the board, buy hdd with the same firmware version and revision if it is visible on the label. Also do some research with Google if replacing PCB works on this model. Also worth to look on the date code and get closest. Manufacturing place must be the same.

That might get you close, or in lightning strike sort of probabilities access to the data. In most cases on top of the bootloader firmware (the majority of the drive firmware is stored on the platters) there are calibration constants stored on the board also and these are required as they account for all the mechanical variables of the drive. On drives I've managed to get to work I've been lucky in that all the calibration has been stored in serial flash chips that I could swap over (mostly Maxtor and early Seagate).

You can try a cold board swap, but it's more likely that not that you'll get no result. Failure of the TVS diodes on hard drives are quite common. Next up seems to be MLC caps. I get best results by following the 12V rail around and progressively removing and replacing parts until I find the dead one. It's a brutal approach but with the right tools (SMD tweezers and a forked iron bit) it's very easy and relatively quick.

A pro data recovery service has all the gear to be able to swap boards and re-write calibration if required, but you pay for the privilege.

I second what Timmay wrote. I have backups, backups of those backups and then another backup sitting in a safe that belongs to someone I trust. Never rely on external hard disks. They tend to use the cheapest of the cheap drives because the general population supports the race to the bottom.

[Gosub]

Hi guys,

Thanks for all the replies, I sourced an identical drive and got it shipped pretty quick so onto the next step
Well,, before I take that next step I have taken a picture, see below

The chip that's blown is the one that says ZA 8

Before I power the new drive, do you still think this is a tvs diode

[macboy]

It is a very likely a TVS or Zener diode. If you google "TVS diode marking ZA", then the first hit is a datasheet describing a TVS diode for 5 V supplies, marked "ZA". You will also find similar results replacing TVS with Zener. The reverse breakdown voltage is in the 6.2 to 7.3 V range. It is designed to provide some best-effort protection against reverse voltage and over-voltage on the power supply rail, usually in combination with a series fuse or a fusible resistor.

You can try to operate the drive without this, it is not a critical part for normal operation. Just don't connect the power supply backwards please! Or you will find that it won't just be a little ten cent diode that is charred

[TheNewLab]

Please let us know your results!

I have always backed up and replaced the drives. honestly?
I have never known of anyone actually repairing the circuit. Let us know! And, if the data is OK, or did it get corrupted?