PCB / Firmware Transfer -- Samsung HD103SI.

[DigiHound]

A few weeks ago, my primary hard drive died. When powered up, the drive would spin up smoothly -- and then emit 12 beeps over six seconds. it sounded just like a POST test. The drive was not detected in BIOS or by any utility. Further investigation revealed that the gold contact points on the back of the PCB were damaged and in one case, almost entirely gone.

I ordered as close a PCB match as I could find online. It arrived today, and I swapped the two. The drive now powers up quietly and in near-total silence, it emits one quiet beep after about 15 seconds. Clearly *something* has changed. It is still not detected by BIOS.

I have been reading that some data may be stored in the firmware / NVRAM of the original drive that must be transferred to the new PCB in order to try to bring the drive back online. I have had no luck finding a utility that's capable of performing this task. Can anyone point me in the right direction?

[harddrivespecialist]

Firmware is stored on a chip and will be required for a new PCB to be compatible with original drive.
You can transfer that chip to new PCB and try if your drive will come back to life.

[DigiHound]

Right.

How?

[labtech]

Soldering tools

[DigiHound]

Sorry, that was a bit unclear. I'm trying to do this in software -- but even if I wanted to do it with soldering tools, I'm not sure which chip is the right one. Here's what the PCB looks like, with one important exception: http://pcb-hdd.com/samsung-hd103si-trin ... mlgoogle.c

My drive does *NOT* have an ESENT chip, or any chip at all in the far-left of the image. There's an attach point for one, but neither PCB has ever had a chip there. There are no chips on the back of the PCB, just the electrical contact points.

Of the various options, the only chip I could conceivably remove by hand would be the eight-pin chip at the left-edge. There's no way I could unsolder the Marvell or the secondary chip at the top left.

So can this sort of job be done via software for a Samsung drive? Does anyone know which chip even stores the data?

[DigiHound]

*nudges this up a bit*

As an added question -- is there any way to fix a bad contact point on the original PCB using conductive ink?

[DigiHound]

So, I don't want to sound like a whiner, but I'm a tad surprised by the lack of responses given the fairly good traffic in the forum. Is this just a strange request? Not much information on Samsung drives?

[jono-ats]

It's not a "strange" request, as these things go on this forum.

However, you are asking for detailed information, and the link you provided is busted, for starters.

The data of interest is located in serial EEPROM, so you have to swap the appropriate IC. Most Samsungs have the adaptives in the MPU; others may have an onboard 8 PIN EEPROM, which is comparatively easy to swap. But if you haven't made a good diagnosis, swapping stuff won't help.

Reading / writing to Samsung EEPROM isn't hard if you have the right gear. Otherwise, you'll need baud rate, parity, stop bits, I/O pinout, etc. and commands to read and write, and be aware of how and when to put the drive in kernel mode. Pages of info to try to tell you how to do it, assuming you don't have the "proper" gear.

And that assumes that you have properly diagnosed the problem.

Maybe your expectation from total strangers is a little high? (PS: whining isn't helpful). Giving as much detailed info as possible, and having a bit of patience are helpful.

You can usually clean up a gold finger and put a layer of solder on it, and wipe off the excess with solder braid. Maybe that's the easiest thing to try, seeing that you aren't sure of the diagnosis?

Good luck!

[DigiHound]

Jono,

http://www.drivestar.biz/samsung-bf4100206b-p-1785.html

That's the equivalent PCB for my drive. I have had no luck finding an external ROM chip and have found comments online (without attribution/proof) that indicate the EEPROM may be inside the drive chassis. Opening the drive is a non-starter, obviously.

Repairing the original PCB with conductive ink (I used a solution that's approximately 50% silver by weight) did not work, insomuch as the drive behaved exactly as before (spins up, 6-8 second pause, 12 beeps, spins back down). There is no visible damage to any other drive component. If there's no external ROM chip to transplant from drive to drive and no way to read/copy the drive's firmware without specialized and expensive tools like PC3000, I'm guessing the only option is to take it to a professional.

I'm just trying to make sure I haven't missed something obvious before I mail it off. Samsung drives aren't as well-documented as Western Digital or Seagates, so I don't want to miss something and then be smacking my head later.

[DigiHound]

D'oh. Apparently my regular message needs human approval, but my thank-you went right through. There's a response to your post coming -- this one was to thank you for your help.

[jono-ats]

The adaptives are embedded in the MPU, which is the large BGA chip on the PCB. There's no EEPROM inside of the enclosure.

Other than making sure the contacts are good, there isn't much you can do yourself with the problem that you describe. Sometimes these are a challenge even with the "right" gear!

[DigiHound]

Jono,

That makes more sense than the EEPROM being inside the unit itself. Thanks for your confirmation.