Hi, I recently found out that the Toshiba MK1032GAX 2,5" HDD I had in an old laptop that I (stupidly) used as a 24/7 SQL server for a project is dead.
After taking the disk out and making sure that it is totally dead (no sound whatsoever), I ordered a donor PCB on the net (100% compatible, as the number indicate), which finally arrived today.

I was tempted to just swap the boards (though I read everywhere that the chance of something like this working is minimal), so I went for it and (thankfully) I heard the drive spinning again, indicating that it is a PCB failure after all. Of course, it only managed to power up and make a clicking sound 4-5 times before I switched if off again.

The point is that I cannot seem to find any visible "ROM" chip on the boards and I don't know what I am supposed to transfer from the old to the new one.
I am attaching a photo (new PCB on top, old one on the bottom):



Can anyone please point me to the right direction?
Thanx in advance

The "ROM" appears to be embedded within the MCU (88i6612-BCJ1).

IMO your best DIY chance would be to repair the PCB. There appear to be fuses at the bottom RHS of each PCB near D1 and E1. Can you zoom up this area of your board?

_________________
A backup a day keeps DR away.

Hi, thanks for the fast response.

Here is a closeup of the old PCB around the area you suggested:



Can you please elaborate a little?

Measure the resistances of the two fuses marked "S". Also measure the zero-ohm resistor marked "000".

D1 appears to be a 5V TVS diode. Measure the resistance between ground and each of its pins.

_________________
A backup a day keeps DR away.

Just measured the resistance of the fuses, all seem to be OK (close to 0 and I get the same readings for the respective fuses on the working board).
As for the TVS diode, are you sure it is actually a TVS diode (it has 3 feet, I thought they had 2)? Measuring resistance between each of the 3 feet and ground (pin 40/43 I suppose?) renders no reading whatsoever on both of the boards...
Any other ideas?

Diodes are sometimes packaged in pairs, ie 2 diodes in one package, with a common anode or cathode. Anway it appears that the fuses are OK, as you say.

My next step would be to measure the voltages at the onboard DC-DC converters. I expect that the 4R7 coil (L7) would be part of the Vcore supply. Measure the voltage at this point.

I'm not sure what IC24 does (below 4R7 coil). Can you tell me its markings? Is it "S78B"?

There should also be a Vio supply. Can you measure the voltages at Q9 (bottom middle of PCB)?

BTW, is it possible that the drive may have been configured to Power Up In Standby (PUIS)? If so, then you could use HDAT2 to disable PUIS. You would first need to wake up the drive as follows ...

HDAT2 /w

_________________
A backup a day keeps DR away.

Hmm, now I am starting to lose you

Can you please elaborate a little on the "measure the voltage" part of the instructions?

As for IC24, yes the marking is "S78B" as you say.

Finally, I seriously doubt that PUIS has anything to do with the issue; the drive was working non-stop for a few years (!) and it suddenly stopped working while the laptop was running. I shut it down and then the disk never came back on.

One more thing: I measured the resistance between the 3-legged diode legs and I got a reading (~140) only when applying positive on the top right leg and ground on the bottom leg (all other combinations produce no readings). The value is the same on both PCBs, so I would assume the problem is not there either.

Can you blow up the areas around L7, IC24 and Q9?

_________________
A backup a day keeps DR away.

Here you go:

I just made a post with the new photos, but it seems that it must be approved by the admin, in the meanwhile:

https://dl.dropboxusercontent.com/u/24277766/IMG_20151203_224400.jpg
https://dl.dropboxusercontent.com/u/24277766/IMG_20151203_224423.jpg

Hmm, just made 2 attempts to post the photos, but both the posts need to be moderated first.
Any ideas about how I could send them to you quicker?

Thank you for the information, but how can this help me?
(Sorry, first time I deal with a failed PCB)

Try to measure the voltages in the highlighted areas.

Set your multimeter on the DC voltage ranges (10VDC or 2VDC, as appropriate). Connect the black probe to the Com terminal and the red probe to the Volt/Ohm terminal. Place the black probe on a circuit ground, eg the a screw hole. Then place the red probe on each of the test points in the highlighted areas. I expect that you should find +5V, +3.3V, and approximately +1.2V.

_________________
A backup a day keeps DR away.

OK, thank you for the update.
Sorry if this sounds stupid, but should I be doing all of the above with the data cable connected to the PCB and powered up? Otherwise I don't see how there will be voltage on any place pf the board. If so, is it safe to connect power to the PCB without it being connected to the disk?

Yes, you need to power up the PCB, and yes it is safe to power up the PCB on its own. Just be sure that your PSU is good and that its output voltage is stable at light loads. You should test your donor board first, just to get the hang of things and to confirm that the supply voltage is sitting close to +5.0V.

_________________
A backup a day keeps DR away.

OK, I finally got some interesting results:

1) All spots marked on the 2nd image (L7 area) read 1,17V on both PCBs, so we are OK there.
2) The 3 spots above Q9 on the 1st image read 5V on both PCBs
3) The rest of the marked spots on the 1st image (bottom legs of Q9 and the 3 spots SW of it) read 3,25V on the new PCB, but just 0,64V on the defective one.

Does that mean that Q9 is the culprit? Please say yes, it looks like one of the few components on the boards I am actually confident I could solder/unsolder

Q9 is probably an NPN BJT transistor. It would be the pass transistor for the Vio supply. The Vio supply powers the 88i6612 MCU and SDRAM (IC21). The control signal comes from the TLS2291A motor controller IC. My fear is that the IC is faulty rather than Q9.

I would now measure the pin at the bottom RH corner of Q9 (with the thin trace). That should be the base (control) pin. Normally it would sit about 0.6V higher than the output, so we should expect around 3.9V. Just be very careful not to short the adjacent pin with your probe.

Next, with the power off, measure the resistance between the base pin of Q9 and ground. This will test for overloads. Use the 200 ohm and 1K ranges.

Now perform a diode test (on the diode range, with power off) between the base pin (red probe) and the 5V and 3.3V pins (black probe). Then reverse the probes and perform the same test. This will test whether Q9 has failed.

_________________
A backup a day keeps DR away.

Any chance you can mark the spots on one of the images?
Sorry for the trouble, but eg. I think we already measured the bottom RH corner of Q9 (it is one of the 2 pins that have the low reading).

I also don't get the last part (diode testing): when you say base pin, I suppose it's the one you are referring to above. An which are the 5V and 3.3V pins?

When I say "5V and 3.3V pins", just use the 5V and 3.25V test points, respectively. I have identified the base pin for you (the square). Can you tell me the markings on Q9?

_________________
A backup a day keeps DR away.