I shut my system down to plug in internal DVD drive. When powering up, I got a chirp from the PSU and nothing happened. Did that a few more times and figured something was up so I disconnected all of my HDDs.

When powering back up, this HDD was not recognized, determined that it was not spinning up (mind you, the drive was working before I shut the system down). So the platters are probably perfect. The CPU on the drive was warm, but it did not spin up.

I have another identical HD103UJ purchased at the same time, so I backed up my data on that one and started comparing PCBs. There are no burn spots on the drive, so I got out my multimeter. The diodes next to the power connector appeared to be fine (no open path going in the wrong direction).

I started comparing the two PCBs with multimeter. Found an open circuit that shouldn't be there (see attachment), as near as I can tell its inside inside the MCC, a Ti SH6125B. Would this explain the HDD not spinning up at all? How can I be sure the MCC is the culprit?

Based on what I've read here, this drive has an internal ROM so PCB swap would not work. Is there a way to hook up a TTL console to dump the ROM (since the CPU seems to work, it is warm after all), replace the PCB, and flash it with the original ROM?

Otherwise, is an MCC swap the only reasonable alternative? This probably near impossible without precision tools.

Maybe this datasheet for a representative spindle motor and VCM controller chip will help you:

http://wandrew.regruppa.ru/PCInfo/TechDoc/L7250(Smooth).pdf

The parallel connected 1R00 resistors are the current sense resistors for the spindle motor (3-phase?) and voice coil motor. I'd check for a shorted phase winding in the spindle motor before bothering to transplant a PCB.

I don't know if these motor circuits are fused, but you may want to check for that, too.

Did you check the zero ohm link (fusible resistor ?) near the 12V TVS diode?

BTW, here is a photo of your HD103UJ PCB:

http://mobile.beareyes.com.cn/2/lib/200 ... C_3701.jpg

As an aside, I notice that it has two shock sensors whereas the sibling in your photo appears to have none.

The 5V fuse (for sata 5V connector) appears to be intact as do the two diodes there (one is ground to 12v, other is 5V to ground).

See attached file with further annotations. Except for pin 3 on the MCC, everything matches the GOOD drive that I am using as a reference.

On pin 3, I get 5.5 ohm on the bad drive, 750 ohm on the good drive. Is that a 750 ohm resistor to the left of PIN 3? If so, I believe I may have confirmed that the MCC is burned out.

Sorry about my 5V/12V error.

It does look like an internal short. Perhaps one of the drivers for the three phases is shorted. Does pin 3 connect to the chip's supply rail?

I suppose you could prove that the chip is faulty beyond doubt by disconnecting the motor flex cable from the PCB. If the low resistance goes away, then that would suggest that at least one of the phases is involved.

Thanks for the advise everyone. I have removed the PCB from the donor and original drive and no changes to the readings. Also the readings on the spindle and head stack contacts were the same on donor and bad drive. Can someone guide me on what is involved in a PCB swap (or where I can learn more), or even if it is possible?

The problem is elsewhere.

What do you mean by elsewhere? Facts:

- drive died due to power supply malfunction (died at power-on of the system)
- drive will not spin up
- diodes have been checked and work
- another drive that works shows different readings at the locations noted
- removal of the PCB confirms that the problem is not in the spindle or head stack

What am I missing?

I have been reading more about PCB swaps. Does anyone know if a PCB swap w/EEPROM swap will work with this model? There appears to be an 8-bit serial eeprom on this PCB.

Bad news:

The 8 pin SOIC is a power component, not an EEPROM.

Your EEPROM data is inside the MCU.
MCU is a big "M" chip with a lot of ball pins under the top....

Janos

Wally, to make it straight ( I have read the thread since the beginning) : your drive needs to be serviced by a pro, and - this is really bad news - few people know how samsung drive works, especially new arch like F1 .
This drive has NO eeprom / nvram / flash , ROM is inside the Marvell chip and this thing it's "game over" for non-really-experts. So beware !
In any case the drive needs to be serviced. There are valuable members of this forum with the necessary knowledge relatively near you, if you want professional help (check the "members" option and make a search using the location) , I call me out because I am at the opposite corner of the world.
It won't be a 10$ quote for sure, but won't cost you an arm and a leg ( I am quite sure about what the problem is but don't ask me anything, I can't. Period.)

P.S. don't open it.

"What have you missing?"

Maybe a lot of things.
The samsung don't want to spin up if the ROM checksum is incorrect, or the content are corrupted, or the pcb "senses" some strange about the preamp or heads.
(The preamp can test the heads without spinning up, and the pcb can test the preamp's functions as well.)
So the problem could be anywhere.
Proper diagnose is needed for better advise.

Janos

BlackST, thanks for clarifying. I can live without the data, but it would be nice to have. I am approaching this from a purely academic standpoint, if the PCB fries (further) so be it.

Since I will not be opening the drive, I don't think there is too much chance of doing more damage at this point - in case I ever want to drop the cash to take it to a pro (spent $3500 at work once for this) that will remain an option.

Since the VCU seems to be the problem I will look into what's needed to desolder that and put on a new one.

MCU is the last thing to check. Don't loose your time.

I don't mean to offend, but I'm wondering if you understand the difference between an open circuit and a short circuit.

In your opening post you say that "the diodes next to the power connector appeared to be fine (no open path going in the wrong direction)". Elsewhere you say that you "found an open circuit that shouldn't be there (see attachment), as near as I can tell its inside inside the MCC, a Ti SH6125B". You say that this O/C measures 5.5 ohms, yet a good drive measures 750 ohms.

What is the resistance of the +12V TVS diode?

The plated thru hole next to the motor IC -- where does the trace go?

No offense, feel free to assume I'm a total idiot. Seriously.

The 12V TVS diode (furthest one to the left in the attached pic) read 150 K in one direction and 225 K in the other (it takes about 5 seconds to settle on 225 K, weird). This matches the donor drive.

The other TVS diode (for ground?) reads 2.5 K in each direction.

The suspect pin on the motor IC is the 4th one down from the top-left in this pic (I called this "pin 3" in my last post, it is actually the 4th one). I read 5 ohm on this (between pin 4 and ground) on the bad drive and 750 ohm on the good one.

The through-hole next to pin 4 doesn't seem to go anywhere, is not 12v, 5v, 3.3v or ground.

The through-hole on the left of the resistor on that trace is to ground.

No offense taken, I will accept any and all suggestions

The 12V TVS reads 150K/225K in each direction.

The through-hole next to pin 4 (see pic) looks like a test point (doesn't go to 12v, ground or 3.3v)

The through-hole just on the other side of the resistor off of pin 4 is to ground.

wally9000,
Hi, would like to know, how you end up with your Samsung HD 103UJ.
I have a similiar problem with that drive.
Thank you, greetings
Urug13

Once for all and for everybody for HD103 and co. (the times of old VG11+ are over)

- When this PCB fails (no, not the protection devices) it's usually game over as there are semicustom and custom parts not documented and not readily available.
- Probability for a pcb swap to work straight away are 1 out of 50 due to differences and variants.
- It is almost useless to make measurements near the SH combo chip as IT FAILS frequently when power surges especially on +12V rail occurred. There's only one option if you know ONLY THIS CHIP is at fault : move a good one from donor board to faulty board. Or have them
- Keep in mind that these HDDs are not a stereo boombox easy to diagnose.
- There's code in MCU, on external ROM and on the platters and everything must match or the drive won't start.
- Even the most expensive tools on the market have VERY LIMITED CAPABILITIES at present for these drives regarding firmware (had to build custom tools for working 100% on them).
- These Samsung drives are undocumented and very few of us know how to work 100% on them (luckily).

This is an old thread but still shows up in search engines. Apologies for resurrecting it but I just wanted to let everyone know that I had a similar failure with my Samsung HD103UJ and there was even a small hole in the motor controller chip indicating failure in the internal FET array.

First I tried to swap the whole board from another HD103UJ drive with very close serial number. It spun up the drive but otherwise refused to start up. Then I purchased a new motor controller chip, carefully removed the old one (the middle of the chip is soldered to the board to transfer heat so it is not possible to remove it without breaking it), soldered the new chip and voilá -- it works again.

Congratulations. Nice work.

_________________
A backup a day keeps DR away.