Hi,

I came back to my Windows 10 box to a blue screen, and a dead cold non-system SATA disc. Swapped cables, etc. It doesn't spin up.

Nothing (I thought) could see it. But then I noticed that on system startup, during the initial BIOS enumeration process the disc IS listed and flagged with BAD SMART before that screen disappears. Seconds later, the BIOS setup screen fails to list the disk.

Can I assume from this that:
(1) the SMART data for this disk is on the PCB?
(2) the SMART data is not "simply bad" as in there are serious, perhaps fatal problems with the disk, but in fact the SMART data itself is corrupt, missing or otherwise unreadable?

And if so, does a PCB replacement with a BIOS transplant still make sense as a first recovery effort?

Thanks for any advice.

----------
Disk is WD1502FYPS. Motherboard is a Gigabyte GA-970A-UD3 with an Award BIOS F1 05/17/2011 under an AMD Phenom II X4 955

SMART is stored on the platters and a PCB swap won't fix the issue.

How valuable is your data? If it is worth about $300USD or more, send it to a pro.

If it isn't with that, and you accept the risks that you could kill the drive, move forward with DIY.

_________________
Luke
Recovery Force Data Recovery

What does CrystalDiskInfo tell you, if anything?

_________________
A backup a day keeps DR away.

Fair enough... just to be clear though, I was referring to two different BIOS screens. The 1st is just after the POST when the BIOS looks for and displays attached devices. It shows both the system ("S.M.A.R.T. OK") and data ("BAD S.M.A.R.T." or "S.M.A.R.T. ERROR") discs. By the time I enter the CMOS/BIOS Setup screen, only the system disc shows up. And I assure you the data disc isn't spinning, ever.

BTW I should also point out that these are the same S.M.A.R.T results I get if I detach the PCB from the disc and power up the computer without the physical disc.

Is this what I'm seeing then? The system BIOS finds an attached device, showing that power is getting to the controller/disc subsystem. The BIOS then issues ATA commands to read the ID (which succeeds; but this only involves the PCB) and next to read the S.M.A.R.T. info (which fails).

But if the S.M.A.R.T. data is stored on the disk itself then that failure report can't be from a literal reference to the data (btw, the Crystal Disk logs on the system disc don't show any S.M.A.R.T. anomalies) since the platters never spin up. So doesn't that suggest that:

1. The system/PCB BIOS is trying to spin up the disk, failing to do so and reporting the bad news as a generic S.M.A.R.T. failure
or
2. The system/PCB BIOS is reading the sensors directly and reporting the bad news as a generic S.M.A.R.T. failure

And while I don't want to be optimistic, doesn't this suggest that the problem might well be with a PCB component (e.g., motor driver, HDA connectors...)? In other words, in the generic case of a disc not spinning up, isn't the S.M.A.R.T. failure report (possibly) totally spurious?

@fzabkar Sorry, we cross-posted. As I mention above, CrystalDiskInfo shows nothing alarming.

Does CrystalDiskInfo identify the drive's model number and capacity? Could we see the output?

If it turns out to be a PCB fault, then a replacement PCB and firmware transfer should cost you US$50 or less.

https://www.hdd-parts.com/13021879.html

If you could upload clear photos of the PCB, I could help you measure the voltage test points.

_________________
A backup a day keeps DR away.

Did you find my attempt at analysis illogical or naive? Of course you don't have to respond, and I do see that you want to step away, but I am curious as to why you seem "shocked".

Anyways, I do thank for your time.

@fzabkar Many thanks, I'd appreciate the guidance and the learning opportunity. I'll post the info and photos later today.

@fzabkar The disc is a 1.5TB Western Digital WD1502FYPS. (More specs are here if you want them.)

I've attached a pic of the PCB (2060-771642-003 REV A).

Hmmm.. I decided to do the little I know of in terms of obvious suspects before sending this, but D3 and D4 check out ok. But checking the other diodes I see D1 at about 200mV but reading ~1V when reverse-biased. Not good, I think.

Thanks again for any pointers you can offer.

If your drive doesn't spin, then it could be because your drive is configured to Power Up In Standby (PUIS), as has already been suggested. In such cases the drive may still be detected by BIOS, although sometimes the model number may be blank. CrystalDiskInfo would show some ID info, but not the SMART data. What does it show?

I doubt that the PCB is at fault, but measure voltages V1, V2, Vneg, and +3.3V. Don't short adjacent pins, as catastrophic damage may result.

Note that Vneg may come up very briefly to -5V after power-on, but the MCU will probably switch it off when it fails to detect the preamp. That's normal behaviour for some PCBs.

_________________
A backup a day keeps DR away.

I hope I'm giving you all what you need without overloading...

1. Sorry if there's been a misunderstanding wrt the CrystalDiskInfo issue. To be explicit, I was referring to the historical "Crystal Disk logs on the system disc [that] don't show any S.M.A.R.T. anomalies"; CrystalDiskInfo can't of course access the current state of the data drive that isn't working.

But I should have attached the logs anyways, so here they are as a zipped folder.

2. I also attached an image showing the data disc "disappearing" during the BIOS process.

3. As for PUIS, I didn't follow up because it's an unknown to me, and after a bit of reading I (incorrectly?) made these assumptions:
o "Power Up" was not the same as "Spin Up", and since the BIOS reports the disc and its ID when it complains about bad S.M.A.R.T. it was getting power so PUIS/PM2 must still be off. (No jumpers are set on the disc, btw.)
o I thought it unlikely PUIS could have been turned on AND malfunction in mid-session

Both of these assumptions now seem perhaps dubious, and if either is wrong I'll happily try ensuring PUIS is off or turn it off if it's on. Can you suggest the best method for doing so for a disc I can't access?

(You'll note from the 1st screenshot in the attached image that he BIOS does show the model. As for CrystalDiskInfo , as mentioned it doesn't detect the drive.)

4. Yes, I know the donor board ROM would have to be replaced with the (hopefully intact) original to get at the the existing data.

5. Thanks very much for the clear instructions on what to measure - that's great. (Also, since you didn't mention the D1 readings, could I ask for a brief explanation of why it wouldn't be relevant? To this amateur, voltage drops in both directions signifies a failed diode.)

In-circuit measurements are not always conclusive. Not only are you measuring the resistance of a particular component, but you are measuring everything it is connected to. In this case your meter is seeing the resistance of the load at V2 via inductor L2. That is, you are measuring the resistance of the MCU, not just D1.

Does CrystalDiskInfo see your drive if you install the PM2 jumper? Is there any difference if you connect the PCB on its own, including the jumper?

I suspect that the drive may have a head or preamp fault (assuming that the firmware tests these components before spinning up the drive). It could also be a bad motor controller.

_________________
A backup a day keeps DR away.

As far as I remember, these drives won’t spin up if they detect a bad head.

_________________
PC Image Data Recovery
http://www.pcimage.co.uk

New!! HDD-PCB.COM for all your PCB and donor HDD requirements!

Thanks for the explanation

Good idea. I'll let you know.

Understood.

Likely, but I'm not sure about these 7 year old Reds. I'll do some research.

Understood. Of course, I'm hoping it's not the heads.

Yeah, I'm not feeling confident but to me it's probably worth the $50 risk.

Oh, I know that - I'm not planning on going anywhere past the board. Thanks.

It's black.

_________________
PC Image Data Recovery
http://www.pcimage.co.uk

New!! HDD-PCB.COM for all your PCB and donor HDD requirements!

Youch! I typed RE4s and spellmangler "fixed" it for me!