Sign in with
Sign up | Sign in
Your question
Solved

None of my hard drives work anywhere I try them

Last response: in Storage
Share
August 18, 2012 6:15:40 AM

Hello,
I have a Gigabyte EX58-UD4P motherboard with one 500G Seagate SATA drive running on the Gigabyte controller with my OS Win7 on it. On the ICH10 controller, I am running a RAID 5 using 3 Seagate 500g drives. The other day I hit the power button, and nothing happened. I suspected the power supply because no LED's etc. I replaced the Power supply. Still nothing. I proceeded to unplug one item at a time (fans, expansion cards etc) and hit the power button and still no activity until I unplugged the last and final thing, my RAID drives. After they were removed, I got LED and Fan activity. Installing one PCI card and fan at a time I powered up, with success after success. Finally I installed my System and RAID drives, but they were all unrecognized in the BIOS, unresponsive and none would spin. I suspected a bad motherboard or disk controller. I put my OS drive into a toaster drive on another computer to retrieve it's data, but it would not spin or be recognized by the eSATA input. Tried other unrelated drives in the other system, and they were recognized in the toaster drive so I suspected my 500G system drive was bad. I put another unrelated spare dive into my computer attached to the Gigabyte controller and it was recognized, onto which I installed a temporary OS. My RAID drives still were not recognized in the BIOS even with the driver installed, nor were any of them spinning. I took one out and brought it to the other system and put that in the toaster drive (hard drive bay), and it did not spin nor was it recognized. So why are all 4 drives in my system (and in another system) unresponsive and unrecognized? What would have caused them to all do that? Power surge? Bad motherboard? Bad power supply? How can I get them to work? thanks guys.

More about : hard drives work

August 18, 2012 7:08:11 AM

i tried once tow 500g on RAID 0 with my x58-ud4p and this problem happened when i removed one of them and i couldn't enter the second HDD or download windows until i returned the second HDD then delete RAID from RAID bios check user manual to know how to delete it, and i also had to return the HDD sata port to the same motherboard sata port
so i suggest to return each HDD to the port where it was then enter RAID bios to delete RAID
m
0
l

Best solution

a c 288 G Storage
August 18, 2012 9:18:51 AM

If your drives were damaged by a power surge, then look for shorted TVS diodes.

See http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

... and http://www.users.on.net/~fzabkar/HDD/ (photo clips)

I suspect you might find damage like this:
http://www.users.on.net/~fzabkar/HDD/ST1000528AS_fix.jp...

If you do see the above, don't despair. The fix is an easy one and should cost you nothing, assuming that the diodes were able to restrict the damage.
Share
Related resources
August 18, 2012 4:22:35 PM

The RAID set up in the BIOS is correct and I already connected each SATA drive to it's original port, so I don't suspect it has to do with that.

It sounds like the TVS diodes are possibly the problem. I read the links that fzabkar gave me and the symptoms sound strangely similar to my problem. One thing the system was doing, (after installing the new power supply) was a brief sound, the fans would attempt to power up and then, nothing, all in just a few milliseconds. If I installed a separate unrelated hard drive, the system would recognize the drive and then start up. And because the drives aren't seen in another system, I suspect it's the drives and not my configuration or cable routing.

After reading all the links fzabkar gave me, I inspected my drives. From the, outside, I cannot see any visible damage and they "appear" fine, from the underside of the PCB board. Is it safe to remove the PCB board? Will I need to remove it to see the problem and fix it? It requires a special star screwdriver I don't currently have but will get if needed. What should I do next? Too bad it's not a simple switch or pop out fuse.
m
0
l
August 18, 2012 8:17:46 PM

this problem happened with every one tried to switch from RAID to single HDD and here are some ways to fix it http://www.tomshardware.com/forum/20882-63-cannot-remov...
http://www.sevenforums.com/hardware-devices/169907-cant...
http://forums.storagereview.com/index.php/topic/21364-c...
http://en.community.dell.com/support-forums/disk-drives...
http://windows.ifmdb.com/UQs3O6B3-how-to-completely-del...
i guess its not hardware problem because i dont think your 3 HDDs damaged at the same time
quote (It sounds like the TVS diodes are possibly the problem. I read the links that fzabkar gave me and the symptoms sound strangely similar to my problem. One thing the system was doing, (after installing the new power supply) was a brief sound, the fans would attempt to power up and then, nothing, all in just a few milliseconds. If I installed a separate unrelated hard drive, the system would recognize the drive and then start up. And because the drives aren't seen in another system, I suspect it's the drives and not my configuration or cable routing. ) i think your motherboard was clearing CMOS and maybe your motherboard now on IDE mode so make sure its on RAID mode before returning your HDDs after that enter RAID bios and see if your HDDs are recognized
m
0
l
a c 288 G Storage
August 18, 2012 9:24:53 PM

@dabito, all the information you need is in my FAQ. The fact that the fans kick once and immediately shut down confirms that the PSU's short circuit protection is being activated. This in turn confirms that one or more of your HDDs is, or was, short circuited. Continuing to power up your HDDs will only risk more serious damage, such as the burns in my last link.

You need to procure a Torx 6 screwdriver to remove the PCB. You may also need a multimeter, if the damage turns out not to be visible. A multimeter can be purchased for US$5 from places such as HarborFreight.

BTW, it would have helped immensely if you had told us the model numbers of your HDDs.
m
0
l
August 18, 2012 10:21:24 PM

fzabkar, I picked up a Torx 6 screwdriver today. The Seagate model #'s are

RAID 5 Drives: ST3500418AS

System Drive: ST3100528AS

Thank You
m
0
l
August 18, 2012 10:25:15 PM

Oh, forgot to mention, I do own a multimeter.
m
0
l
a c 288 G Storage
August 18, 2012 10:37:44 PM

The ST31000528AS has inductors in series with the TVS diodes. You can expect that these will be burnt. The fix is to remove the shorted diode and flow a blob of solder over the inductors.

The ST3500418AS drives also have inductors:
http://www.users.on.net/~fzabkar/HDD/ST3500418AS_5V_ind...

However, their TVS diodes are elsewhere on the PCB:
http://www.users.on.net/~fzabkar/HDD/ST3500418AS_12V_TV...
http://www.users.on.net/~fzabkar/HDD/ST3500418AS_EEPROM...

If a simple DIY repair doesn't recover your drives, then the worst case scenario will cost about US$200 for all 4 drives. This includes a firmware transfer. Each PCB stores unique, drive specific information in a serial EEPROM. This chip, or its contents, needs to be transferred from patient to donor.
m
0
l
August 19, 2012 4:38:31 AM

Wow, you hit the nail on the head! I removed my PCB board from the system drive (ST3100528AS) and saw the inductors all fried, just as you said. I think I counted 5 of them, all damaged. I took some pictures for you to see, but I don't know how to attach them to this message. While looking at the close up pictures, it appears that one of the inductors was entirely fried and incinerated to ashes. Another inductor is just barely hanging on, on one side. So it appears that there were more inductors. I'm sure the picture would be helpful. It looks bad; I hope it can be salvaged. I suspect my RAID drives will all have the same condition. Anyhow, how should I clean away all the carbon residue? Then, should I solder one big blob across all the inductors at once, or do I need to precisely apply one blob of solder across each individual inductor? Also, I believe that the TVS Diodes are directly above the inductors, am I correct? I haven’t tested them yet. Will I be testing and removing them both (in other words do they each play a part in my problem or do I just need to focus on one in particular)?
Thank you for your help so far. There is absolutely no way that I would have ever figured this one out on my own.

m
0
l
a c 288 G Storage
August 19, 2012 6:32:47 AM

You need to test both TVS diodes. Usually the overvoltage hits either the +5V supply or the +12V supply, but not both. If both diodes are shorted, then this may point to a more serious problem.

Instead of solder blobs, there is an easier way to restore connectivity to the affected supply rails. This involves soldering a wire between the SATA power pins and the cathode of the corresponding diode. I'll show you on your photo when you succeed in uploading it.

As for the charred remains, you can clean them up with a toothbrush and isopropyl alcohol (audio/video head cleaner).
m
0
l
August 19, 2012 9:06:52 PM

Here are 2 pictures of the ST3100528AS drive's PCB up close to the action.



Uploaded with ImageShack.us



Uploaded with ImageShack.us
m
0
l
August 19, 2012 9:09:43 PM

Will those photos work or do you need another picture showing more of the board?
I have yet to open the ST3500418AS drives, but I expect to see similar damage, although a different layout as you stated.
m
0
l
a c 288 G Storage
August 19, 2012 9:44:26 PM

"QA" is the 5V TVS diode, "LEK" is the 12V diode.

It appears that the overvoltage struck the +5V input. That's the more sensitive one, so you may want to verify that there is no visible damage to the motor controller chip. You need to provide the resistance measurements for each diode before we can proceed.

BTW, the following article should help you identify the components:
http://hddscan.com/doc/HDD_from_inside.html
m
0
l
August 20, 2012 4:56:06 AM

Ok, I tested the diodes. The LEK diode appears to be ok. It shows a voltage drop in one direction of .62 on my meter and no drop in the other direction. The resistance measures 120 kohm in one direction and 97.7 kohm in the other direction.

The QA appears to be bad. It shows no voltage drop and a resistance of .7 ohms in both directions.

I also opened one of the ST3500418AS Raid drives and measured its diodes. I found burnt up inductors also, though not as badly damaged as the system drives. The measurements on the one Raid drive were similar to the above drive. The larger diode showed a voltage drop in one direction of .61 v with no voltage drop in the other and a resistance of 112.6 kohms one way and 100.7 kohms the other way.

The smaller diode in that drive measured no voltage drop in either direction, and a resistance of only 1.8 ohms. I imagine that one is bad as well.

I did not notice any other visible signs of damage on either of the boards.
m
0
l
a c 288 G Storage
August 21, 2012 10:49:47 PM

You need to measure the resistances of all the TVS diodes, even if they look OK. BTW, when you say "no voltage drop", I presume you mean zero ohms. That's the way it should read on the 200 ohms resistance range.
m
0
l
August 22, 2012 9:08:32 PM

You are obviously more tech savvy than me, so it's highly probable that I misunderstood something you said or that I am doing my measurements incorrectly.
If I understood you correctly, there are 2 TVS diodes on each hard drives PCB: a 5v and a 12v correct? I measured each one and the following shows my results. Bear in mind That my muiltimeter is different than yours (you probably have a really nice Fluke or something comparable). Anyhow, mine has a diode tester. It measures a voltage drop across the diode in both directions. Using this meter, all of the 12v TVS diodes are shown as still good, but all of the 5v TVS diodes are shown as bad.
As far as measuring the resistances, my meter doesn't have 200ohm but it's a 400 ohm scale. When I measure the resistances of the 5v TVS diodes, I get a different reading when measuring forward bias vs reverse bias. When I measure the 12v TVS diodes using the 400 ohm scale, I get an over the limit reading s0 I have to change the meter to Kohms to get a reading.
Here are all the results of my resistance measurements each with a forward and reverse bias measurement:

System Drive ST3100528AS:
5v TVS .7ohms
12v TVS 97.7 Kohms and 120.0 Kohms

RAID Drive 0 ST3500418AS:
5v TVS 1.8 ohms
12v TVS 100.7 Kohms and 112.6 Kohms

RAID Drive 1 ST3500418AS:
5v TVS 1.5 ohms
12v TVS 219.2 Kohms and 279.7 Kohms


RAID Drive 2 ST3500418AS:
5v TVS .2 ohms
12v TVS 90.4 Kohms and 106.3 Kohms

I hope this information helps you help me.
m
0
l
a c 288 G Storage
August 22, 2012 11:28:26 PM

You have confirmed that all four drives have a shorted 5V TVS diode. Assuming that these diodes have succeeded in limiting the damage to themselves (and to the inductors), then the solution is to snip each of the diodes with flush cutters, and reconstruct the path from the +5V pins in the SATA power connector to the cathode of the 5V TVS diode. The cathode is identified by a stripe at one end of the diode, and a "V" on the PCB.

You can either reconstruct the path by replacing the burnt inductors with wire links or blobs of solder, or you can run a wire between the diode's cathode and pins 7,8, and 9 of the SATA power connector:

http://pinouts.ru/Power/sata-power_pinout.shtml

Cleaning up the charred remains would help you to distinguish between those components which have been damaged and those which have been sprayed by their neighbours.

You should be able to find quite a few threads at Seagate's forums, eg ...

http://forums.seagate.com/t5/Barracuda-XT-Barracuda-Bar...
m
0
l
August 23, 2012 9:00:15 PM

If I connect a wire from the SATA power connector as you said, can I just use 1 piece of wire and then solder pins 7 8 and 9 together? Or do I have to run a separate wire from each pin and carefully not get solder on the adjacent pin? The wire I have is 25 gauge, I think.
m
0
l
a c 288 G Storage
August 24, 2012 1:02:02 AM

You can connect all three SATA pins together. They are all fed from the same source.

Before you power up your drive, just confirm that there is no longer any short circuit between the +5V and ground pins at the SATA power connector. Also confirm that there is continuity (ie zero ohms) between the diode's anode (the non-striped end) and SATA ground.

It may be an idea to power up each board, and verify that it doesn't smoke, before you reinstall it on the drive's HDA.

Best of luck.
m
0
l
August 24, 2012 1:25:51 AM

Thank you. I'll be crossing my fingers.
m
0
l
August 24, 2012 7:17:23 PM

Fzabkar,

I got my system drive to work!!!!!!!!!! I removed the bad +5v TVS diode and ran a piece of wire from the 7,8, and 9 pins to the cathode, then tested on the bench just as you said. No smoke, so I reinstalled the system drive, Windows started and there is my desktop and all of my data!
Awesome!!!
Now I'll attempt to fix my RAID drives. Hopefully they go smoothly too and I can retrieve my data from them as well.

Thank you. You are a genius!
m
0
l
August 24, 2012 7:18:13 PM

Best answer selected by dabito.
m
0
l
a c 288 G Storage
August 24, 2012 8:21:10 PM

Congratulations. What you've essentially done is to remove a short, and bridge a fuse. Sometimes data recovery really is that easy. :-)

The root cause of your problem is most likely due to failed capacitors on the DC side of your PSU. You will probably find that several appear bloated or leaky. There are numerous examples at http://www.badcaps.net/forum. I suspect that the damage is most likely caused by overshoot at switch-on, in which case the outputs may test OK when the supply settles down.

Power Supply Design and Troubleshooting:
http://www.badcaps.net/forum/forumdisplay.php?s=d407e96...

BTW, for continued protection, you may like to consider replacing the 5V TVS diodes with an SMAJ5.0A. These are available for about US$1 from the suppliers in my FAQ.
m
0
l
!