Toshiba PCB Problems

[scotts]

I have a Toshiba MK8032GSX drive that won't spin up. I sent it off to a small DR company that said they could fix it, but after replacing several chips w/o success, they claimed the CPU (Marvel?) chip was bad and cannot be replaced. I had sent an identical working drive along with the bad drive so they could swap PCB boards, but the owner explained that Toshiba HDs only work with the original PCB. He referred me to another (bigger) DR company that might be able to fix it.

The second DR company said no problem they can fix it, but shortly after receiving the HD they claim they cannot proceed because the PCB is not the original board. The tech said he read the serial # off the PCB with his software and it does not match the drive. So I contacted the first DR firm to get the original PCB back, but the owner insists he did not swap the PCB because he knew it would not work, plus he does not have any spares PCBs for my drive anyway. He also questions how the second company was able to get the s/n off the PCB because the CPU was fried which prevents the board from powering up.

My first DR guy (who seems extremely knowledgeable) now recommends I get the drive back and send it to the Seagate DR division so they can take a look at it. He thinks they can swap PCB boards and reset the s/n on the ROM to get it working.

Although I have excellent computer skills, I'm not a DR expert like some of you guys - the advice being shared in this forum is impressive. If anyone has any experience or advice they can share I would greatly appreciate it!

- Scott

[msurgeon]

If you could send a photo of the PCB that is on the drive that would be extremely helpful. It is true that the original PCB contains a flash ROM with adaptive data which is unique to the drive; another PCB will not work unless the original ROM is transferred to the new board. It would be helpful to see the photo to ascertain if such a ROM chip exists on this board since newer models are continuously being modified from earlier ones. To be honest, I don't see how any DR company without the original board, could solve the problem. If indeed they could do this, then my hat is off to them. Re-setting the serial number would have no effect on accessing the data.

Last edited by msurgeon on September 5th, 2008, 10:57, edited 1 time in total.

[beto]

Could u post us here the PCB photo?

[scotts]

Thanks for the prompt replies! I attached a photo of the PCB. Sorry it is so blurry, my camera is messed up. Will try to get a better photo later if needed.

So it sounds like a PCB swap is out of the question. Apparently the drive was hit with a power surge (the motherboard & LCD of the Sony laptop it was in was also fried) is it possible the PCB is too damaged for DR?

Attachments

[msurgeon]

It appears that your board does have an external flash ROM chip. That's good. Apparently,
your original DR guy tried to repair the board, but when that failed he gave up. The fact is a donor PCB with a transfer of the original ROM contents (or the chip itself) will solve the problem (assuming other problems haven't occurred within the drive itself). The critical thing here is that you must have the original board. Neither Seagate nor any other DR company can do anything without that. In that instance the only hope would be to get enough donor PCBs and hope that one of them has adaptive data close enough to the original to work. This is obviously not the preferred method. With the original board, most reputable DR companies can effect the recovery.

[pcimage]

Doesn't look like a regular external ROM chip to me.

But I'm willing to be stood corrected!

[msurgeon]

Blurry photo makes it difficult to see if what I think is ROM actually is. If I'm wrong, then the ROM is embedded in the main chip. If that chip is indeed bad, then the situation is not hopeful.

[Doomer]

ROM on this drive is inside CPU (Big Marvell chip)

BTW if drive doesn't spin it is not always fried CPU
sometimes it is possible to read ROM through COM port even if drive does not spin and not coming ready normally
Maybe the second DR company did read ROM through COM port, compared SN with SN on the drive's lid and made a conclusion about non-original PCB

W/o original PCB DR is almost impossible (not because of SN but because of unique adaptives)

[badsector7]

I attached a photo of the PCB. Sorry it is so blurry, my camera is messed up.

It's from a scanner. Right?

[scotts]

The drive definitely has the original PCB, I sure of that. My first DR guy thought the CPU might be fried only because he had replaced 3 bad chips and the drive and it still would not spin.

Doomer thanks for confirming that the s/n can be read from the ROM of a non-functioning PCB board

Yes the original image was from my scanner which cannot focus on a 3D object. I borrowed a camera and managed to get a better image of the board.

Attachments

[msurgeon]

Better photo quality indicates that there is in fact no external ROM on this board; adaptive data is contained within the large Marvell CPU. However, I notice that there is a fuse missing at position F2 on the board. Short across the two contacts and see if the drive spins up. Don't know if your original guy already tried this but it's certainly not going to work without a jumper in place.

[Doomer]

The drive definitely has the original PCB, I sure of that. My first DR guy thought the CPU might be fried only because he had replaced 3 bad chips and the drive and it still would not spin.

Well the PCB maybe original but the ROM is located inside CPU. And those guys replaced chip 3 times. They might just screwed up a little with chips and forgot to place original one back on the place (I really don't think they did it on purpose, just you know sh*t happens)

Last edited by Doomer on September 5th, 2008, 15:33, edited 1 time in total.

[Doomer]

However, I notice that there is a fuse missing at position F2 on the board. Short across the two contacts and see if the drive spins up.

F2 has to be missing nothing wrong with it or around it

[cgallery]

You think they took that Marvel off? It is a BGA. Do they have the right tools for BGA work?

And here I was just the other day saying hard drives don't use BGA's.

[scotts]

msurgeon:
This is a photo of my identical working PCB not the bad one (DR firm still has my bad one), so the fuse missing must be normal.

Doomer:
You misunderstood, the CPU wasn't removed from PCB 3 times, 3 other fried chips (possibly fuses?) were replaced on the PCB but he did never replaced CPU.

Sorry for confusion over the identity of the PCB in photo guys, but it was the only one I had here at the moment. Like I said it is from an identical drive thought, and the serial # is only a few digits off from the bad drive.

[msurgeon]

However, I notice that there is a fuse missing at position F2 on the board. Short across the two contacts and see if the drive spins up.

F2 has to be missing nothing wrong with it or around it

Thanks for the tip. I hadn't seen that before on older Toshiba boards.

[Doomer]

You think they took that Marvel off? It is a BGA. Do they have the right tools for BGA work?

I'll tell you a secret
In Russia some guys are using frying pan and powerfull light bulb (no kidding) to replace BGA chips and they succeed but it really depends of the skills
or you can have less skills and expensive tool

[dick]

You think they took that Marvel off? It is a BGA. Do they have the right tools for BGA work?

I'll tell you a secret
In Russia some guys are using frying pan and powerfull light bulb (no kidding) to replace BGA chips and they succeed but it really depends of the skills
or you can have less skills and expensive tool

Not only in Russia......it is possible to do this work even with a Black & Decker hot air paint stripper. I know I have done it but the skill level to do it is VERY high. It is not for the faint hearted and definately not to be done on a customer job. I have only tried and achieved this on an experimental basis.
So what would be the right tools? $$$$$$$$$$$$$$

[N.C.]

@ scotts

I am not 100% sure about the configure and adaptives are in the mcu. (marvell, cpu)
Can you write down for as the numbers from IC6 and IC10?

This drive only use 5V and have fuse in the circuit.
Usually the first chip is the VCD wich is being shorted.
The next moment, the fuse is blown.
So, you have a great chance about the other parts are all good. (maybe)

Regards,
Janos

[scotts]

Wow - I could not see any numbers on those chips until I put a 10x loope on it.

IC6: TLS229 1A
58A7FLJBC

IC10: G31 00A3R
531AD