Dead nvme Intel 600p -> Data Recovery -> now part is missing

[fzabkar]

I think I can see two similar components in the middle of the PCB, above the top left corner of the SK Hynix chip.

[P.F.]

I think I can see two similar components in the middle of the PCB, above the top left corner of the SK Hynix chip.

I've noticed them as well. From the dimensions and colouring, they could be the same.
But I wouldn't know how to measure them, without taking them off.

I can't see the markings on the ICs, but measure the voltages between ground and the junction of each inductor and its associated capacitors.

Thanks for looking at this!!! Just to confirm, since this electronic terminology is foreign to me, would this be the spot to measure, for example?

reg3.jpg (22.2 KiB) Viewed 24662 times

[fzabkar]

Just measure the voltage across the capacitor. It amounts to the same thing.

This is what you are looking for (quicker-and-dirtier switchmode regulator):
http://www.hddoracle.com/download/file.php?id=6446&mode=view

Tutorial - Linear and Switchmode Regulators used in HDDs:
http://www.hddoracle.com/viewtopic.php?f=59&t=231

As for the other components, just measure their resistances (with power off).

[fzabkar]

After you have verified the supply voltages, read this article:

https://blog.acelaboratory.com/how-to-recover-data-and-evidence-from-nvme-drives.html

It may be heavy going for you, but it tells you that your PCB most probably has a "ROM mode" test point at the "technological" end. Normally you would short the "ROM mode" test points during power-on, and then release them after power up. If the SSD then identifies itself, it will prove that the flash controller is not brain dead.

[P.F.]

Just measure the voltage across the capacitor. It amounts to the same thing.

Thank you for the background info! I've done the measurements.
Can you read anything from that?

reg3.jpg (24.13 KiB) Viewed 24638 times

BlueResistors.png (480.56 KiB) Viewed 24638 times

[gold6565]

Resistance to ground

Attachments

[P.F.]

at the points indicated all ~ 3 kOhm, when measuring negative probe at GND

[P.F.]

Resistance at junction inductor/capacitor to GND (image regX.jpg):

• 1 = 1.57 kOhm
• 2 = 0.98 kOhm
• 3 = 5.59 kOhm

[fzabkar]

I wonder if the resistors are of the following type:

High Precision Wraparound - Wide Ohmic Value Range Thin Film Chip Resistors, Sulfur Resistant:
https://www.vishay.com/doc?53017

The SDRAM chip is rated for a supply of 1.35V or 1.5V, so I would think this accounts for the 1.367V measurement.

The Vcore supply for the flash controller would be around 1V, so I would think that would account for the 1.1V supply.

The 1.2V supply and the 3.3V supply would power the NANDs, IIUC.

In short, the power supplies look OK to me.


References:

RT8086B, Richtek, Synchronous Step-Down Converter, 3.5A, 1.2MHz, 2.8V - 5.5Vin, marking 0Yx, UQFN-12L:
http://chip.tomsk.ru/chip/chipfile0-x.nsf/all/34D15D8C9E31816246257F80005444F2/$File/DS8086B-02.pdf
https://www.richtek.com/assets/product_file/RT8086B/DS8086B-01.pdf

RT5784AGQWF, Richtek, ACOT Synchronous Step-Down Converter, 2A, 2.5V - 6Vin, 1.5MHz, 25uA IQ, marking 01x, WDFN-8JL:
https://www.richtek.com/assets/product_file/RT5784A=RT5784B/DS5784AB-03.pdf

MP5087GG, Monolithic Power, 7A, 5.5V load switch, marking BDy, QFN2x2-12P:
https://www.monolithicpower.com/pub/media/document/MP5087_r1.2.pdf

SN74AUP2G07, Texas Instruments, Low-Power Dual Buffer/Driver, 0.8V - 3.6V Vcc, marking H5, SON 6-pin:
http://www.ti.com/lit/ds/symlink/sn74aup2g07.pdf

I suspect that the other 6-pin ICs are dual NPN transistors or dual N-channel MOSFETs.

[P.F.]

@fzabkar: Thank you so much, for identifying all those parts.

I am still trying to get the resistor somehow back.
Of course both companies tell me, they didn’t remove it.

I can‘t judge, if the ssd would be able to communicate at all without the resistor. Shorten the pins or not, drive is not detected in BIOS.

[lcoughey]

I have seen SSDs come in that don't look exactly the same as the stock photo. It is possible that the missing chip was never there in the first place. And, yes, I've seen the solder pads look like there was something there before, without tampering. That isn't to say that someone didn't remove the chip, but there is a chance that they both may be telling the truth.

Without a photo of the device before you sent it out, there is no telling. If you had sent it to me, we likely would have a photo of the device as it arrived to help confirm.

[fzabkar]

@P.F, if @gold6565 measured the resistor in-circuit, then the measurement would have been loaded by the resistance of the circuit. Taking your own measurements into account, the adjusted resistance would be 954 ohms or 825 ohms, depending on how the measurements were taken, but this would still be dependent on the multimeter.

R = 770 x 3990 / (3990 - 770) = 954 ohms
R = 770 x 11460 / (11460 - 770) = 825 ohms

You could solder two short wires to the pads and experiment with various resistance values, eg 820 ohms and 1Kohm.

One other thing you could test is the MP5087GG load switch IC (marking "BDG" ?). Measure the voltages on the adjacent capacitors.

[fzabkar]

@P.F, how are you testing the drive? Are you using an adapter? If so, which one?

[P.F.]

I have seen SSDs come in that don't look exactly the same as the stock photo. It is possible that the missing chip was never there in the first place. And, yes, I've seen the solder pads look like there was something there before, without tampering. That isn't to say that someone didn't remove the chip, but there is a chance that they both may be telling the truth.

Without a photo of the device before you sent it out, there is no telling. If you had sent it to me, we likely would have a photo of the device as it arrived to help confirm.

Well, I've purchased the drive, I've installed it, then it was in use, then it broke, I took it out, and sent it off to Datarecovery.
It got back with visible burns of taking the component out. You are not suggesting, it came with those burns from the factory?
When I bought it? But yes, I was naive, first not to take a picture, second not even to open the package after I got it back from Ontrack and just
handing it straight to my local pc shop. I trusted too much. My mistake.

[P.F.]

@P.F, how are you testing the drive? Are you using an adapter? If so, which one?

I test it with adapter (IB-1817Ma-C31) and directly on an internal NVME Slot.

[fzabkar]

If the drive isn't entering "ROM mode", then it could be that the controller is dead. I would connect a leaded resistor to those suspect pads and try again.

One other thing to check would be the crystal, but that requires a scope.

You could also check whether any IC is heating up excessively.

[P.F.]

If the drive isn't entering "ROM mode", then it could be that the controller is dead. I would connect a leaded resistor to those suspect pads and try again.

One other thing to check would be the crystal, but that requires a scope.

You could also check whether any IC is heating up excessively.

No scope here
Heat is very light, so it seems something is happening, but nothing terrible. For now I am waiting for a similar model to arrive (tomorrow?), with which I can compare and take off a replacement resistor.
@All:
Will keep y'all updated, thanks for the kind support thus far!!!

[P.F.]

Hello Gurus,

sorry for the long wait. The donor drive arrived later than expected and my comparisons took time as well.
Here are the results so far:

1. Restistor Value = 770Ohm, meassured out of circut, so exactly what gold6565 was saying, thank you
2. When the resistor is taken off, drive has the same symptoms, not detected by bios, no rom mode when contacts shortened
3. When resistor is put back, the drive is detected fine, INTEL (or SM2260 in ROM mode) (did this 4 times at least, take off, put back), the donor drive that is, not the one to be rescued
4. When putting the resistor to the patient drive, unfortunatly, nothing
5. BUT (now its getting interresting): I measured the voltages at the resistor pads, with resistor and without, on the working drive and the defunct.

   

   IMGP3635_cut.jpg (46.35 KiB) Viewed 23752 times

   
   

   Code:

     donor  |  patient  |  remarks (all measurements taken using usb adapter, IB-1817Ma-C31, while machine booted fully into windows 10)
   0.45V |    0.000V |  with resistor
   0.30V |    0.005V |  without resistor, laptop powered by battery
   1.00V |    0.005V |  without resistor, laptop powered by power adapter
   0.02V |    0.005V |  without resistor, laptop powered by battery, after ~106s, I noticed some blinking of the led of the usb adapter
         |           |  it seemed that's some powersaving functionality, the blinking and voltage switch did not happen, if only booted to BIOS


So, now I am thinking a lot of probably stupid things, but maybe you can make some sense of those observations.

[fzabkar]

Were the supply voltages (1.1V, 1.2V, 1.35V) the same on your donor SSD?

[P.F.]

Were the supply voltages (1.1V, 1.2V, 1.35V) the same on your donor SSD?

Yes