Samsung HD501LJ 500GB Not Starting

[fzabkar]

"Autorange" just means that you don't have to set the scale. You still need to select DC Volts, AC Volts, Ohms, DC Amps, AC Amps, etc.

[odedshankar]

Its a new $50.00 Multimeter off ebay UNI-TC something like that.
I have it for 2 weeks so didn't work with it much.
I measured resistors with it - autorange jumps automatically between OM to microOM but measures correctly.
It test Diodes and continuity quite easy.

But with the Voltage, i did set it up for DC on autorange and i tried also Range manual mode but it wouldn't allow me to change anything - looked at the manual but its useless Chinese manual - i'm surely missing something here, i'll try and contact the seller maybe he knows.
Thanks

[Vulcan]

Note: All suggestions below of taking measurements are at your own risk - one slip of the meter probes, lack of ESD precautions etc. and you could cause a bigger problem.

I tested the diodes, the 12V was defected.

Do you mean the 12V TVS diode or something else?

How exactly what it defective - what value did you measure, using what setting on the DVM (i.e. multimeter)?

Its a new $50.00 Multimeter off ebay UNI-TC something like that.

If you supply a link to that item on ebay, or supply a photo of your DVM, a reader here might already be using something similar and be able to give you some advice...

But with the Voltage, i did set it up for DC on autorange

It's difficult to understand what your meter is actually showing, when trying to take those readings which fzabkar requested. A DVM showing an AUTO symbol is normal, when it's set to autorange; it's the actual measured value which is needed (of course) but I can't see where you have reported whether that is all zeros, or fluctuating wildly, or totally blank, or something else, when trying to take those measurements which he suggested. More details from you about exactly what you are seeing, may be helpful.

For what it's worth, with some kinds of PCB fault then DVMs can show unusual and difficult-to-interpret voltage readings. An appropriate oscilloscope can be a much better diagnostic tool for these situations, although obviously I'm not suggesting you go out and buy one right now. I'm just explaining that DVMs have their limitations. However if you were confident with your equipment (and in this case I sense that you are not) then the inability to get a stable, valid reading is itself a diagnostic result, suggesting that the measured voltage isn't stable.

One thing you could do is to remove the PCB (to minimise the risks) from any other working (but unimportant) drive and measure the voltages from the similar coils on that PCB. Can you successfully measure those voltages, and get some stable, sensible DVM readings? If so, then that indicates something unusual (i.e. bad) with the behaviour of your "problem" PCB.

FYI I'm assuming that you have already measured some voltages successfully (not just resistances & continuity) with this DVM (e.g. SATA power +5V and +12V) and so have some confidence about its ability to display those.

All comments are just my $0.02, in an attempt to suggest tests / details / information for you to supply (at your risk), which would help others to understand what problem you have, because I'm not able to get enough understanding from what you have said so far. It's your choice whether or not to follow those suggestions, of course. Good luck

P.S. I won't be surprised if the PCB is beyond easy DIY repair, based on what you've said so far - but I can't be sure whether the details so far are due to problems with your meter, or your technique, or the PCB.

[odedshankar]

Thanks Volcan, appreciate your detailed answer.
The 12V diode showed 0 value on both directions on diode test.
My measures of the coils went wild because i was silly and didn't supplied power to the PCB, I'll check it again on Monday when i'm back and office and will be really careful with the probes.
As you can obviously see my knowledge with electronics is quite limited, but I'm trying to catch up, another stupid question if i may, what is "An appropriate oscilloscope" for data recovery needs, what does it provide that DVM doesn't?
I've seen they have some scopes what does it mean?

[odedshankar]

I finally managed to get the Voltage readings, please find attached image with the reading values.

I don't know if these values seems bad or not, but readings were stable every time.
The only thing that looked suspicious is the 2 Diodes measuring marked in blue:
Both diodes are exactly the same based on the part number which is identical, but the Multimeter Diode tests values are different:
1. the top one red probe on + showed 0V, other way showed 0.198V.
2. the bottom one red probe on + showed 0.070V, other way showed 0.130V.

Is it possible that one of the Diodes is causing all the problems?

Attachments

HD501LJ_regs.jpg (27.74 KiB) Viewed 7823 times

[fzabkar]

The diodes are not important. The coils are where the output voltage appears.

You have confirmed that the Vio voltage (+3.3V) is OK.

The 4R7 coil is associated with the Vcore supply. This supply appears to be OK, but you should measure the bottom terminal. I have given you 3 equivalent test points in the attached diagram. Choose whichever is more convenient for you.

The -5V supply for the preamp appears at the bottom terminal of the lower 100uH coil. You need to measure that as well.

At the moment it is looking as if the -5V supply is missing while the Vcore supply is present. This would suggest that the motor controller is working, at least partially, and that the MCU is faulty. I'm assuming that the MCU switches on the -5V supply.

Attachments

HD501LJ_regs_2.jpg (87.92 KiB) Viewed 7819 times

[odedshankar]

I actually measured the other terminal of each coil too.
In all of them the values were slightly lower than the upper terminal.
The vCore as you describe it was also ~1.397V on the bottom terminal.
The Bottom Terminal of the middle coil, the one you marked with -5V was also showing a read of 113microV.

So does it indicated the Marvell MCU is gone?
In that case this case is not doable ad the adaptives are in the controller.

Cheers

[fzabkar]

The motor controller generates the Vcore (+1.4V) and Vio (+3.3V) supplies for the MCU. These appear to be OK. Once these supplies are up, then the MCU comes alive and performs a POST. The -5V supply is also generated by the motor controller, but it isn't clear to me whether the MCU needs to turn it on. In some WD designs, for example, the negative supply remains switched off if the board is dismounted from the drive. I suspect this may be because the MCU cannot detect the preamp. If the MCU were dead, then the result would be the same.

In short, I suspect that the MCU may be dead, but I can't be certain. In any case there doesn't seem to be any easy repair option. What puzzles me is that you observed a shorted 12V TVS diode. This kind of fault would impact on the motor controller, if anything. It would affect the spindle motor and VCM sections of the chip, but perhaps the +5V sections would survive. I don't know.

Have you tried to communicate with the MCU via its serial terminal interface?

[odedshankar]

I checked it with terminal
and no response so i assume the MCU is gone

[BlackST]

It's pointless to try to repair these PCBs unless you have everything like combo ICs and spare MCUs and all discrete components and want to spend some time. I do it only in two case : emergency and mass repair.

In my opinion it is a total waste of time, if the purpose is to get data (and ca$h) as soon as possible. Just do a quick test if there is activity and power at MCU / CIC / port, if everything is flatlined and you don't see any evident problem in 5 minutes, get another PCB /donor and make it work with your HDA (if it is only the PCB).

This was my opinion posted same day (May, 9) and now is May,15.
Note (and that's what I have noticed in this field) : the faster you give an answer to customer, the more the succe$$ ( and this will avoid other problems, too ).

P.S. if this is the scenario, you have either to find a PCB that works OR make a new PCB work , all this assuming there is no other problem on HDA or firmware. Cross fingers.

[odedshankar]

You are totally right about these kind of cases being a waste of time, but with this specific case i identify that the customer didn't plan to pay to much so it was pretty much for research and not for the $$$$.
Anyway i missed your line at the time which is marked in Red, that is a very good diagnostics tip.

With these kind of cases of Samsung are there unique parameters stored in the MCU?
is it possible to regenerate them for a new PCB, using PC3000 or other tool?

[odedshankar]

I replaced the PCB with a donor one and managed to recover it all