www.deadharddrive.com

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

Reasonably intelligent, yet woefully ignorant would be data recovery hack here. I'm in no hurry, but would like to retreive data from this disk. No, it's not worth the cost of professional service.

what happened was... flat desktop computer on a shelf in a desk - jar to desk - out of the ordinary fan noises - no disk on reboot. No clicking or grinding noises heard before or after the reboot.

Disk does not spin - Disk is not recognized by bios

There was immediate shut down upon start up on that pc... indicating a short?

Upon placing the drive in another computer, there was no immediate shut down of that system... I still have to check out the original pc it was in. It may have a fried piece or two.

In the newer pc, the drive still does not spin, nor is recognized by bios.

Reading here, I tried the TVS route and lifted one end of 2 tvs near the power input and tried it on battery back-up power to hopefully avoid any additional problems - the ps in that computer is stable.

I don't have a multi meter. but the resistance reading on the tvs were the same in both directions - around 10 ohms or less.

Lifting the tvs didn't help matters - no change.

I'm assuming the next step is to get a donor drive and replace the pcb.

Is there a need to swap the rom if one can get a donor board manufactured the same month?

If so, then it might help if I knew which chip on this board was the rom, if there is a rom on this board. I do have a heat gun and can handle a swap easily enough.

Here is a pic I found here at this site... my board is identical - same numbers: bf41-00086a
http://dl.dropbox.com/u/8286013/img2101zh6.bmp

I've circled the tvs I lifted. I used an iron, but do have a heat gun and can handle a chip swap easily enough.

Your thoughts?

is there anything else I should check / try first?

fzabkar · Deadharddrive regular +3 Joined: 02 Apr 2010 Posts: 210

If you don't have a multimeter, then how did you measure the resistance?

The ROM is internal to the Marvell MCU (big "M").

As for what to check, buy a digital multimeter (US$10) and measure the resistances of each diode and each of the zero-ohm resistors marked "000".

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

another question concerns the microcode... would this drive shut down, not spin and not be recognized by bios if there was corruption of the info?

There was a second ata drive in the computer that was shocked. bios did recognize it, but it took the windows recovery console and fixmbr to bring it back.

I'm mainly focused on the samsung board as it's not spinning and not recognized by bios. Could it be that the board is fine and it's the microcode? ... would pinning safe mode bring it up? I didn't know this was something to try until recently, but haven't searched for the pin settings for it yet.

I'm still woefully ignorant.. but learning.

fzabkar · Deadharddrive regular +3 Joined: 02 Apr 2010 Posts: 210

Does the drive feel or sound like it is trying to spin up? If so, then you could have a stiction problem.

I can't answer your question about microcode, but I do know that the bulk of the firmware is stored in a reserved area on the platters. The MCU contains a small amount of boot code and "adaptive" data. I suppose if this code and data were corrupt, then the drive wouldn't wake up.

If I were troubleshooting your board, I would measure the voltages at the DC-DC converters and at the TVS diodes. Useful test points are at the pins of the coils (100, 100, 4R7), pin #1 of the Samsung SDRAM, jumper pins, and the 3rd and 4th pins from the left on the bottom row of the 20-pin preamp connector.

There is also a transistor at the top left corner of the Marvell MCU, and another below the motor controller chip. Be very careful not to slip with the probes, as major damage could result.

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

fzabkar · Deadharddrive regular +3 Joined: 02 Apr 2010 Posts: 210

The voltages should be measured with respect to ground. That is, the black common lead of the meter should go to ground.

Use the SATA power ground as the reference:
http://pinouts.ru/Power/sata-power_pinout.shtml

For example, measure the voltage between pin 3 of the preamp connector and ground, and between pin 4 of the preamp connector and ground. Also measure the voltage between each coil and ground. Same thing with each pin of the transistors.

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

new board arrived - manufactured the same month...

swaping it was good news - bad news... the board spun up and looks to the SA - making a clicking pattern - 2 - 2 - 1 - repeat.

so the heads are going to the SA, but it's still not reading and not recognized by bios... which tells me something, but not enough for a diagnosis.

the marvel chip... the wording of "for that drive" I read online has me confused... does each individual drive have specific info for that individual drive on it? Is particular info for the platters written to the rom during low level formatting or something?

Question is - with a donor board with the right date and numbers (4th character on the board sticker), do I still need to swap the marvel chip?

or should I start assuming the problem is further down the line - the pre-amp / heads / platter?

fzabkar · Deadharddrive regular +3 Joined: 02 Apr 2010 Posts: 210

I could be wrong in my initial statement. It appears that not all, if any, Samsung drives have the same requirements for "adaptive" transfers as other brands. The professionals who are watching this thread could confirm this one way or the other. However, in my experience, errors that are to the benefit of the data recovery profession tend to remain uncorrected.

I still think the best approach is to determine which components failed on your original board.

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

again, thanks for the reply.

Clearly, I was hoping for an easy plug and play solution.

Digging into the board and replacing parts may be the path I need to take. The transistor near the M is gone... as is the one near the motor controller.

at this point, I'm still thinking of swapping the marvel chip before using the donor board for parts to fix the original... or replacing anything inside.

We'll see.

fzabkar · Deadharddrive regular +3 Joined: 02 Apr 2010 Posts: 210

I think you will find that one or both of those transistors constitute linear regulators. I suspect that one supplies the +3.3V for the SDRAM and the MCU's Vio rail. That said, if the 3.3V regulator is faulty, then this doesn't explain how you could still be measuring 3V at pin #1 of the SDRAM.

Anyway, you now have the advantage of having a working board to compare voltages. I'd be very interested in your feedback.

Casual · Active contributor Joined: 26 Jan 2011 Posts: 7

I shall amend that to "at one time, the sdram had 3.3v" Embarassed

I didn't see smoke, but I must've...

If and when I go to the new board, it'll be with a sewing needle taped to the end of the probe.

Kit · Deadharddrive regular +2 Joined: 10 Nov 2008 Posts: 56

In your case you need find exact board from drive with same capacity and same firmware. And after that drive have start to working. Your donor pcb possible is from drive with bigger capacity or different firmware that's why drive clicking. Sometimes samsung write information about firmware on label, sometimes not. If you will change mcu it can help in your case. And you can compare voltage between good and bad pcb to find bad transistors or resistors or may be problem in chip which control spin up of drive (in the corner of board)

corbingravely · Posted: Wed Jul 01, 2015 9:31 am Post subject:

The best thing that you can do is to connect this hard drive parallel to a working hard drive in a machine. Then you can actually retrieve it with some software like retrieving data from USB. I hope this helps.

-------------------------------------------
CORBIN GRAVELY
how to setup microsoft outlook
-------------------------------------------