This particular external hard drive was left running over night. In the morning, I immediately noticed the room was unusually quiet and also, there was a weird smell.

I removed the HDD (Hitachi HUA723020ALA641 2TB 3.5" 7200rpm) from the enclosure. This is the SATA-USB bridge they used. Where the transistor marked "122A" used to be, there is now just a black pit on the PCB.





If I connect the HDD to an identical known-good SATA-USB bridge, nothing happens. Not even the LED on the SATA-USB bridge comes on (which it does when the HDD is not connected).

The 12V PSU still measures fine and is working normally with another similar external hard drive.

Looking at the PCB of the HDD itself, there are some signs of overheating in the area of the motor control IC.





There may be some non-backed up data on this drive that I'd like to try to rescue. I have quite decent SMD soldering skills. I'd be thankful for any advice regarding troubleshooting and repairing this drive.

Post a clear picture of your drive PCB.

Here's an attempt at some better pictures.

The 12V TVS diode ("Z4") appears to have been clobbered by an overvoltage (black spot on body).

There is also a small surface mounted fuse between Z4 and the SATA connector, so that would most likely be open circuit.

LE25FS406 is the "ROM" that must be transferred to a replacement PCB, if the motor controller (0A73080) has been damaged.

You could remove the shorted diode and replace the fuse. The fuse rating would be 4A. A suitable replacement diode, should you so choose, would be an SMBJ12A.

_________________
A backup a day keeps DR away.

Many thanks for replying so quickly! Will a failed TVS diode usually behave basically like a 0 Ohm resistor? The two small dark green fuses can of course easily be checked for continuity. I will try to take some measurements.

I suppose the LE25FS406 should be pretty much identical to e.g. W25Q40EW and SST25WF040 (1.8V) and should be readable with a TL866ii+ programmer?

I think I may have a proper donor PCB for the HDD. I will check that also.

Yes.


https://www.reveltronics.com/downloads/datasheet/le25fs406.pdf

_________________
A backup a day keeps DR away.

Well, I don't know if this is good news or bad news.

- The mentioned fuses are fine.
- The 5V TVS diode measures (in circuit) 5.76 kOhm with the black probe on the cathode and 7.27 kOhm with the red probe on the cathode.
- The 12V TVS diode measures (in circuit) 1.16 kOhm with the black probe on the cathode and 21.89 kOhm with the red probe on the cathode.

For a known-good donor board, I got these results:

- The mentioned fuses are fine.
- The 5V TVS diode measures (in circuit) 11.72 kOhm with the black probe on the cathode and 11.80 kOhm with the red probe on the cathode.
- The 12V TVS diode measures (in circuit) 1.17 kOhm with the black probe on the cathode and 20.28 kOhm with the red probe on the cathode.

Video showing how the measurements were made: https://www.youtube.com/watch?v=rKGH-QOFgGw

Measure these voltage test points.





Could you upload a photo of the damaged bridge PCB? Hopefully this will give us a clue as to what happened.

_________________
A backup a day keeps DR away.

So that would be a total of 7 voltage measurements; +5V, V_core, V_io, V+, V-, V_boost, +12V. Can I use the same ground reference point for all 7 measurements, e.g. one of the screw holes?

And all 7 measurements should be made with the PCB loose from the HDD, while providing power to the PCB e.g. through a confirmed good SATA-USB bridge?

This is what the burnt SATA-USB bridge looks like. I cleaned off the soot with IPA and a brush. (I also desoldered the LED).

Yes to all, except that you can power the PCB from your computer's power supply via a SATA power cable, if you prefer.

Hires photos of bridge PCB:

https://upload.wikimedia.org/wikipedia/commons/d/df/Airy_by_CnMemory%2C_external_hard_disk_-_SATA_2_USB_adapter_EBS-J29-10-93239.jpg
https://upload.wikimedia.org/wikipedia/commons/7/78/Airy_by_CnMemory%2C_external_hard_disk_-_SATA_2_USB_adapter_EBS-J29-10-93238.jpg

The blown transistor is the main power switch. It's failure is inconclusive. You can bypass this switch by linking the D and S pins, but you would need to check for shorts before doing this. You would only do this to find out what else was damaged, if anything, so I wouldn't bother.

_________________
A backup a day keeps DR away.

* Bad PCB:

+5V : keeps rapidly circulating between voltages in approx. the 0.1V .. 1.0V range.
V+ : keeps rapidly circulating between voltages in approx. the 0.1V .. 1.0V range.
V- : -0.00V
V_io : -0.00V
V_core : -0.00V
+12V : keeps rapidly circulating between voltages in approx. the 0.2V .. 1.2V range.
V_boost : keeps rapidly circulating between voltages in approx. the 0.5V .. 2.5V range.

* Good PCB:

+5V : 5.07V
V+ : 5.07V
V- : -2.992V
V_io : 1.803V
V_core : 0.938V
+12V : 12.25V
V_boost : 12.18V

It looks like the 5V fuse may be open.

What voltage do you measure at the 5V TVS diode?

BTW, the boost voltage is low because it is switched off when the HDD is not running.

_________________
A backup a day keeps DR away.

Is something making either the SATA-USB bridge or the 12V PSU shut down immediately, each time the HDD PCB tries to power on? (Same bridge and PSU as I used for powering the good HDD PCB while taking the above mentioned readings).

The LED on the SATA-USB bridge doesn't come on properly; it just blinks every now and then.

Both fuses still have continuity, and I'm getting similar circulating voltage readings on both sides of either fuse.

Sorry, I didn't notice your 12V reading.

Yes, it does appear that something is overloading your power adapter. I think that a PCB swap and "ROM" transfer is the most expedient solution.

You could measure the resistance between ground and each of the voltage test points. That might tell us which chip, if any, has failed.

_________________
A backup a day keeps DR away.

It feels a bit strange that those fuses are still ok, even though something might be drawing excessive amounts of current at power-up. But in any case, apparently something on the HDD PCB is toast. If it's e.g. the motor control IC, surely a board swap and ROM transfer would be the easiest (attempt at a) solution.

I guess there's no easy way of telling if I might have e.g. a blown preamp or seized motor?

Resistance to ground from test point;

* Bad PCB:

+5V : 11.2 kOhm
V+ : 11.2 kOhm
V- : 0.60 kOhm
V_io : 1.4 kOhm
V_core : 59.2 Ohm
+12V : 22.0 kOhm
V_boost : 92.6 kOhm

* Good PCB:

+5V : 11.76 kOhm
V+ : 11.76 kOhm
V- : 0.58 kOhm
V_io : 1.46 kOhm
V_core : 38.0 Ohm
+12V : 20.47 kOhm
V_boost : 86.8 kOhm

I'm thinking that the MOSFET drivers for the 3 motor phases may be shorted. You won't see the short until the MOSFETs are switched on by the motor controller. That said, these kinds of shorts usually result in visible blistering of the IC, so I'm confused.

You can see the MOSFETs here:

http://www.hddoracle.com/viewtopic.php?p=484#p484

If you are concerned about preamp damage, locate the V+ and V- test points on the other side of the PCB and measure the resistance to ground with the PCB mounted on the drive.

Testing for a damaged preamp:
http://www.hddoracle.com/viewtopic.php?p=23821#p23821

_________________
A backup a day keeps DR away.

This is actually quite fascinating stuff. Well, in any case, I decided to go ahead and perform the board swap and ROM transfer. Unfortunately, the results are not great. I've tried to power up the drive exactly once now, and only got 6 x BUZZZZ and then a faint bump+knock sound, only once. Then nothing more, and after a few seconds, I switched off the power.

Video of this power up attempt: https://www.youtube.com/watch?v=R3YOi5YWPQk

Are these MOSFETs that have been integrated into the motor controller chip? And when they blow, the motor controller will usually look burnt (which it doesn't in this case)?

ROM chip removed from bad PCB:


...and installed on good PCB. Cosmetics worsened a bit, as I managed to fill a couple of vias with solder.



Resistance to ground with good PCB installed on bad HDD;
V+ : 11.75 kOhm
V- : 0.596 kOhm

It sounds like the spindle motor has a problem, or it may be a stiction issue (heads stuck to platters). In any case, your ROM transfer seems to be OK. And yes, the MOSFETs are integrated into the motor controller.

You could measure the resistance between each pair of motor windings. I expect that the phase-to-common resistance would be of the order of 1 ohm, while the phase-to-phase resistance would be twice this figure.

_________________
A backup a day keeps DR away.

* Bad HDD with PCB installed:

Pin 1 to pin 2: 0.4 Ohm
Pin 1 to pin 3: 0.4 Ohm
Pin 1 to pin 4: 0.4 Ohm

Pin 2 to pin 3: 0.4 Ohm
Pin 2 to pin 4: 0.5 Ohm

Pin 3 to pin 4: 0.5 Ohm

* Good HDD without PCB:

Pin 1 to pin 2: 0.9 Ohm
Pin 1 to pin 3: 0.9 Ohm
Pin 1 to pin 4: 0.9 Ohm

Pin 2 to pin 3: 2.2 Ohm
Pin 2 to pin 4: 2.2 Ohm

Pin 3 to pin 4: 2.2 Ohm

This is not looking so great. Does it matter if the drive has the PCB installed or not?

Would the drive be behaving even worse if there was a problem with the ROM?

Btw, did my earlier results indicate that the preamps are ok?