Dead Hitachi 1Tb HDS721010CLA332

[yello]

I have a totally dead Hitachi 1Tb HDS721010CLA332. It was working fine for a few year, no strange noises. Then the MB went bad and the HDD was out of service for 2 weeks or so. Now I wanted to get some data and the HDD is totally dead, no sound, no small, nothing.

I sto.... I mean borrowed a picture from another user here added the voltages that I measured with a cheap MM. One voltage I could only measure in AC mode. My oscilloscope is not on hand right now.

The three TVS seem ok, they are not shorted. Is there anything else I can try? Any idea what it may could be?

Attachments

hita.jpg (82.71 KiB) Viewed 17330 times

[yello]

Please let me add a piece of new information: I checked the the two (current sense?) resistors marked R100 and their resistance is high (out of range). I wonder if that is normal. The seems to be connected to the TVS and part of the protection circuit.

[HaQue]

I doubt the resistors will measure properly in circuit

[fzabkar]

The R100 resistors should be measuring 0.1 ohm.

The Vcore supply should be around +1.2V IIRC.

The Vneg supply should be -5V.

Could you post a link to the larger photo?

[Vulcan]

I checked the the two (current sense?) resistors marked R100 and their resistance is high (out of range). I wonder if that is normal.

Were you doing that measurement with power applied to the board, or with power removed?

[fzabkar]

I doubt the resistors will measure properly in circuit

If a 100 milliohm resistor measures high, then it is open. It doesn't matter whether it is in or out of circuit. It's only when high resistances measure lower than their nominal value that you need to consider circuit loading.

[HaQue]

Thanks for the explanation, Ive done stacks of hooking things up and hardware hacking but no real electronics theory. Its good to learn some where ever possible

[yello]

The R100 resistors should be measuring 0.1 ohm.

The Vcore supply should be around +1.2V IIRC.

The Vneg supply should be -5V.

Could you post a link to the larger photo?

Where are the measure points for what voltage?

All voltages were measured against GND. The two R100 were measured with power off. I could replace them, but have no similar resistors on hand, I guess I could simply measure out a piece of wire and solder them in?

The picture I posted wasn't mine, but my board is exactly the same. I will take a better picture late.

[fzabkar]

No, I wouldn't bypass these resistors. There must be a short circuit somewhere on the PCB. We need to locate that first.

[yello]

No, I wouldn't bypass these resistors. There must be a short circuit somewhere on the PCB. We need to locate that first.

I added two pictures of MY board. I should disclose at this point that I work in circuit assembly and have a vague idea about circuits. I can not see any mechanical damage, or shorts.

I am not clear about the function of the R100, they seem to work together with the TVS - but the TVS have no function in normal operation and only work when there is overvoltate (same way that a a fuse has no actual function). A circuit diagram would be helpful at this point.

BTW, the HDD specific ROM seems to be connected the HDD drive and not on the PCB. So it should be quite possible to simply swap the PCB, or not?

Attachments

[yello]

Adding:
When I short the R100 with an A-meter I get on both R100 zero, which is strange. On the right R100 I have 0.33V (TVS side) an 0V. On the left R100 I have an unclean signal and need a scope to look at it.

[fzabkar]

I assumed that the resistors you were referring to were the R050 resistor and the R470 resistor in the regulator circuits. Instead it appears that those two R100 resistors are indeed fusible, and they appear to be fusing the grounded side of each TVS diode. If they're open, then I would have thought that both diodes (Z2 and L14) would have been shorted.

I notice that the motor controller has a black spot under the "5" in "OA5". If so, then that would explain the missing voltages.

Your best bet may be to try moving the 25FU206 chip to a donor PCB.

I would measure the resistance between the "10mV" coil and ground, with the PCB installed on the drive. That will tell us if there is a short circuit within the preamp.

You might find the following tutorial useful.

Tutorial - Linear and Switchmode Regulators used in HDDs:
http://malthus.zapto.org/viewtopic.php?f=59&t=231&p=597

The Vneg regulator uses a Cuk topology whereas the others appear to be buck regulators.

[yello]

I notice that the motor controller has a black spot under the "5" in "OA5". If so, then that would explain the missing voltages.

The black spot is actually a hole, I didn't notice that. Can I just change the OA56511 (I saw I can buy that chip online), or change the whole PCB?

If I change the whole PCB, I need to carry the 25FU6 to the new board?

[Vulcan]

I notice that the motor controller has a black spot under the "5" in "OA5". If so, then that would explain the missing voltages.

Indeed - that black spot looks to be an actual hole in the chip.

[Edited to add: I see that yello just confirmed that. Just shows why having photos of the actual PCB can be helpful ]

[yello]

I notice that the motor controller has a black spot under the "5" in "OA5". If so, then that would explain the missing voltages.

Indeed - that black spot looks to be an actual hole in the chip.

[Edited to add: I see that yello just confirmed that. Just shows why having photos of the actual PCB can be helpful ]

I totally see your point!

I took two more pix, sorry for the quality, it's the best I can get.

Is my plan realistic:

1. I get a new PCB AND the OA56511 IC
2. I solder the OA56511 to the old PCB (I got hot air and QFP nozzles)
3. If that doesn't work I use the new PCB and move the 25F IC to the new PCB

Can somebody confirm if I MUST move the 25F IC or not please.

PS: Does the PCB need to be the same revision? Mine is 110 OA90233 01

Attachments

[fzabkar]

Just move the 25FU206 IC to a compatible donor. Alternatively, if you have access to an SOIC/SOP chip clip and a device programmer, then reading and writing the chip in-circuit would be an easier option.

BTW, if you transfer the motor controller, there may be other problems (apart from the two R100 resistors).

I'd check the -5V supply for a short before proceeding. If the preamp is dead, then DIY will be a waste of time.

Just out of curiosity, I'd measure the resistances between each end of the R100 resistors and the ground pins at the SATA power connector. I would also determine where the diode ends of each resistor go. That is, to which pins of the motor controller, if any, do these resistors connect. Does the motor controller have separate analogue and digital ground pins, and are these connected together via the R100 resistors? Do the +5V ground pins and the +12V ground pins connect together at the SATA power connector?

If these grounds are open, then the motor controller must be providing some kind of internal (?) path in its present fault mode to explain the 0.33V reading at the TVS diode.

[yello]

When I measure the R100 against GND (no power cable connected) I get left to right: 0 - 0 & 0 - 1M8

The 2 large TVS are 'open' but the on the small TVS I measure 26K

What is the IC that is under the PCB and fixed to the housing?

Is there a chance that that is defect, or is there a chance that the motor spin motor or the arms are defect too? (I guess there is always a chance, but how likely is it?)

[Vulcan]

What is the IC that is under the PCB and fixed to the housing?

Please supply a photo, showing what you mean.

Just FYI, I doubt the MB went bad independently of this drive being damaged, as your original story seemed to suggest. More likely is the PSU became faulty (or allowed through an external power event e.g. lightning) which damaged both the MB and this drive. However due to the failed MB, you didn't know that the drive had been damaged as well.

[BlackST]

To cut what is starting to be a long story short, can be that the headstack was killed by this failure (50/50). It is NEARLY IMPOSSIBLE unless you have industrial equipment to successfully move the chip , and usually when it breaks there are other failures.
P.s. Depending on what is broken, if you swap the pcb you will fry the new one at power up.
So said, you can 'experiment' with it.

[Vulcan]

can be that the headstack was killed by this failure (50/50).

Indeed, the OP has been warned, but he keeps asking more and more questions...