Help Needed: Fried Hitachi HDS721010CLA 1TB
August 8th, 2011, 12:42
WHAT HAPPENED (full drive details at bottom)
Based on hours of reading at the forum this weekend (thank you to this community for even existing), I hope my problem is a TVS diode burnout:
I replaced a faulty power supply on Friday night (Aug 5), which is something I haven't done in about 20 years. I (stupidly) believed that modern Molex connectors were mated and could not be plugged in backwards, but apparently they can be, as I plugged in my HDD power backwards without much force (thank goodness this didn't happen with the motherboard or other components).
At power-on, there was a momentary hesitation, then a puff of smoke from the HDD, followed by me diving for the power switch too late.
The result is that a hard drive that has tons of critical data will no longer spin up, power on, or be recognized by the motherboard.
WHERE I'M AT NOW
First I noticed that SATA power pinouts 6 and 12 (2 grounds) were blackened. In particular, a resistance check on the multimeter told me that pinout 12 was no longer connected to its PCB hole. I have applied solder carefully so it now reads as a connection, without crossing to any other holes, but the drive still won't spin up or be recognized (photos below taken before soldering).
I bought another (same model/size, one year later) HDD, and have tried swapping in the good PCB, but the PC still won't recognize it, although I understand this is common due to PCB changes and vagaries.
What I haven't found so far in reading several posts and articles on multimeters and diodes is which side of these black blobs on my HDD PCB that could be the TVS diodes is the cathode and which is the anode and what multimeter readings mean it's failed and can be cut.
QUESTIONS
Q1. It's not clear to me whether I take multimeter readings with power applied to the HDD or not?
Q2. In the attached PCB front-side image below, have I correctly circled the TVS diodes?
Q3. If so, which color circle is the 5V and 12V respectively?
Q4. With my Commercial Electric MAS830B multimeter set to 'Diode', and no power applied to the HDD, and assuming the top of the image is the anode (red lead), the multimeter reads 1 (not a diode) for the red circle component. Assuming Q2 is correct, does this mean this diode has failed and can be cut?
Q5. However, the reverse direction (red multimeter lead on the bottom) of the red circle component reads 161, and if it were a diode it should also read 1. Is this correct?
With no power applied to the HDD, and assuming the top of the image is the anode (red lead), the multimeter reads 1703 for the green circle. The reverse reads 452.
Q6. I see the sideways component between the two circled components has a gray stripe. Could it also be a diode? It reads 1 assuming A-C, and 169 assuming C-A.
I get comparable (within 10%) readings from all the analogous components on the new good version of this hard drive.
HDD DETAILS
MAKE: HITACHI
MODEL: HDS721010CLA332
CAPACITY: 1TB
P/N: 0F10383
MLC: JPT3EA
C P/N: H3D10003272S
MANUFACTURE DATE: MAY-2010
FRONT OF PCB READS ON 4 LINES:
OA72947
BA3321B
89BX
BACK OF PCB READS ON 3 LINES:
220 OA90233 01
88V0B E318699
94V-0
Based on hours of reading at the forum this weekend (thank you to this community for even existing), I hope my problem is a TVS diode burnout:
I replaced a faulty power supply on Friday night (Aug 5), which is something I haven't done in about 20 years. I (stupidly) believed that modern Molex connectors were mated and could not be plugged in backwards, but apparently they can be, as I plugged in my HDD power backwards without much force (thank goodness this didn't happen with the motherboard or other components).
At power-on, there was a momentary hesitation, then a puff of smoke from the HDD, followed by me diving for the power switch too late.
The result is that a hard drive that has tons of critical data will no longer spin up, power on, or be recognized by the motherboard.
WHERE I'M AT NOW
First I noticed that SATA power pinouts 6 and 12 (2 grounds) were blackened. In particular, a resistance check on the multimeter told me that pinout 12 was no longer connected to its PCB hole. I have applied solder carefully so it now reads as a connection, without crossing to any other holes, but the drive still won't spin up or be recognized (photos below taken before soldering).
I bought another (same model/size, one year later) HDD, and have tried swapping in the good PCB, but the PC still won't recognize it, although I understand this is common due to PCB changes and vagaries.
What I haven't found so far in reading several posts and articles on multimeters and diodes is which side of these black blobs on my HDD PCB that could be the TVS diodes is the cathode and which is the anode and what multimeter readings mean it's failed and can be cut.
QUESTIONS
Q1. It's not clear to me whether I take multimeter readings with power applied to the HDD or not?
Q2. In the attached PCB front-side image below, have I correctly circled the TVS diodes?
Q3. If so, which color circle is the 5V and 12V respectively?
Q4. With my Commercial Electric MAS830B multimeter set to 'Diode', and no power applied to the HDD, and assuming the top of the image is the anode (red lead), the multimeter reads 1 (not a diode) for the red circle component. Assuming Q2 is correct, does this mean this diode has failed and can be cut?
Q5. However, the reverse direction (red multimeter lead on the bottom) of the red circle component reads 161, and if it were a diode it should also read 1. Is this correct?
With no power applied to the HDD, and assuming the top of the image is the anode (red lead), the multimeter reads 1703 for the green circle. The reverse reads 452.
Q6. I see the sideways component between the two circled components has a gray stripe. Could it also be a diode? It reads 1 assuming A-C, and 169 assuming C-A.
I get comparable (within 10%) readings from all the analogous components on the new good version of this hard drive.
HDD DETAILS
MAKE: HITACHI
MODEL: HDS721010CLA332
CAPACITY: 1TB
P/N: 0F10383
MLC: JPT3EA
C P/N: H3D10003272S
MANUFACTURE DATE: MAY-2010
FRONT OF PCB READS ON 4 LINES:
OA72947
BA3321B
89BX
BACK OF PCB READS ON 3 LINES:
220 OA90233 01
88V0B E318699
94V-0
- Attachments
-
- HDD PCB - back side
-
- HDD PCB - front side
Re: Help Needed: Fried Hitachi HDS721010CLA 1TB
August 8th, 2011, 14:18
QUESTIONS
Q1. It's not clear to me whether I take multimeter readings with power applied to the HDD or not?
Q2. In the attached PCB front-side image below, have I correctly circled the TVS diodes?
Q3. If so, which color circle is the 5V and 12V respectively?
A1, Apply power to the hdd using ext. PSU
A2, ur right those are the 2
A3 doesn`t matter, the multimeter should read both
FYI, when you power the hdd if @ 1st. no reading in ur multimeter then i suggest to change both diodes and try again reading (should know soldering/de-soldering)
also chk (suggest to replace) the one in the middle of the circle as well with the other 2
The Best Place for this is here: http://www.users.on.net/~fzabkar/HDD/
chk the HDSxxxx files there many thnx 2 the site owner (member here) for those tips
good luck
Re: Help Needed: Fried Hitachi HDS721010CLA 1TB
August 8th, 2011, 14:33
recurve7 wrote:Q1. It's not clear to me whether I take multimeter readings with power applied to the HDD or not?
Sounds like there is a short to ground on one of the voltage lines (probably 5V that accidentally got 12V), so you should not plug power to the PCB until that is resolved.
Re: Help Needed: Fried Hitachi HDS721010CLA 1TB
August 8th, 2011, 15:34
@recurve7: A few comments...
Do you really want to take the risks of DIY with "critical" data?
The answer depends on the type of readings you're taking. Voltage readings = yes, power would need to be applied to the circuit, for you to measure voltages. Resistance readings = no, power must not be applied to the circuit under test when taking resistance readings, as the multimeter is supplying its own voltage in order to perform a resistance measurement.
Look at the SATA power pinout, compare that to your PCB, and you'll see the answer becomes clear - on your photo, green circle = +5V TVS, red circle = +12V TVS.
It isn't. Uni-directional TVS diodes (like these) are used so they are normally reverse-biased and don't conduct (until their breakdown voltage is reached), so the cathode is connected to the supply voltage (at the top of your image), and the anode is connected to 0V.
That's odd, especially if your comment applies to the readings around the +5V TVS.
recurve7 wrote:The result is that a hard drive that has tons of critical data will no longer spin up, power on, or be recognized by the motherboard.
Do you really want to take the risks of DIY with "critical" data?
recurve7 wrote:Q1. It's not clear to me whether I take multimeter readings with power applied to the HDD or not?
The answer depends on the type of readings you're taking. Voltage readings = yes, power would need to be applied to the circuit, for you to measure voltages. Resistance readings = no, power must not be applied to the circuit under test when taking resistance readings, as the multimeter is supplying its own voltage in order to perform a resistance measurement.
recurve7 wrote:Q3. If so, which color circle is the 5V and 12V respectively?
Look at the SATA power pinout, compare that to your PCB, and you'll see the answer becomes clear - on your photo, green circle = +5V TVS, red circle = +12V TVS.
recurve7 wrote:assuming the top of the image is the anode
It isn't. Uni-directional TVS diodes (like these) are used so they are normally reverse-biased and don't conduct (until their breakdown voltage is reached), so the cathode is connected to the supply voltage (at the top of your image), and the anode is connected to 0V.
recurve7 wrote:I get comparable (within 10%) readings from all the analogous components on the new good version of this hard drive.
That's odd, especially if your comment applies to the readings around the +5V TVS.
Re: Help Needed: Fried Hitachi HDS721010CLA 1TB
August 8th, 2011, 16:49
Thanks for the replies and information, guys.
Vulcan makes a good point about what steps I'm willing to try 'DIY' with this critical drive.
I was comfortable taking multimeter readings of various components, and comfortable trying to swap the same-model new PCB with the bad one (which didn't work), and even comfortable putting a tiny drop of solder from that fried ground pinout (12) to its PCB hole, but I'm not really brave or stupid enough to remove any of the diodes and continue applying power to it.
I've dropped it off (and its new good clone) at a recovery service. I was really really hoping to avoid the $500 - $1000 hit, but I don't think there's much more I'm comfortable trying on my own.
Vulcan makes a good point about what steps I'm willing to try 'DIY' with this critical drive.
I was comfortable taking multimeter readings of various components, and comfortable trying to swap the same-model new PCB with the bad one (which didn't work), and even comfortable putting a tiny drop of solder from that fried ground pinout (12) to its PCB hole, but I'm not really brave or stupid enough to remove any of the diodes and continue applying power to it.
I've dropped it off (and its new good clone) at a recovery service. I was really really hoping to avoid the $500 - $1000 hit, but I don't think there's much more I'm comfortable trying on my own.
Powered by phpBB © phpBB Group.