SSD Samsung PM991a fail

[giulio93]

Hi everyone, I have this ssd nvme that has suddenly stop to works. It was in a laptop. I unplug it from the laptop and connected via usb adapter to another PC. The drive is not detect, nothing. I tried with Win10 and Lubuntu but same.
Then I grab my thermal camera and noticed the top side of the nand chip went hot immediatly, raising the temp from 50 to 85 deg in a minute. I inspect the board and take some resistance measurments to ground but no short detected. I suppose the nand is shorted internally. Can you help me diagnose this drive and eventually fix it?
I can take also voltage measurments but I don't know what should I aspect.
Thanks you all

Attachments

[fzabkar]

Measure voltages V1, V2, V3. Also measure the resistances between ground and each of these test points.

Attachments

[giulio93]

Measure voltages V1, V2, V3. Also measure the resistances between ground and each of these test points.

Resistance at Inductor1 = 147 Ohm
Inductor 2 = 4.7 KOhm
Inductor3= 1MOhm (ish)
The Ind3 resistance is decreasing during the measurments

V1= 0,9 V on top and 0,93 V on bottom on the ind1
V2= 1,8 V
V3 = 2.5 V

[fzabkar]

There are no shorts, but I don't know what to make of V3. I would measure the voltages on the small bypass capacitors surrounding the NANDs.

[giulio93]

There are no shorts, but I don't know what to make of V3. I would measure the voltages on the small bypass capacitors surrounding the NANDs.

Sorry i did not understand the V3 thing. It has suppose to be different than 2.5V?
Those are the tensions of all the strips on the right of the nand chip.
0.3 on both sides
0
0
0,9
0,9
0,9
1,8
1,8
1,8
1,8
1,8
1,2
1,2
2,5
2,5
2,5
Anyway, I was looking at the cap next to V1, am I wrong or that tiny line could be a "crack" in it? I don't know if I'm just overlooking at it.
Tell my yours opinion please.
Thanks

P.s. if you need the voltages i took put in the images to understand where are they, just tell me

[fzabkar]

Is KLFEG4R4AB an SSD-on-a-Chip? Does it incorporate both the flash controller and NAND flash on a single IC?

[giulio93]

There are no shorts, but I don't know what to make of V3. I would measure the voltages on the small bypass capacitors surrounding the NANDs.

I don't understand the V3 things, the voltage has suppose to be different?
Here the tensions from top to bottom
0,3 (each side)
0,3
0 each side
0,9
0,9
0,9
1,8 each
1,8 each

1,8
1,8
1,8
1,2
1,2
2,5
2,5
2,5

If you need that I put those into the image just tell me.
I also notice 2 things:
1) my adapter has a cable that one side has type C (the one goes into the adapter) and on the other side 2 USB A at Y config. The drive seeams to be powerd just with one of those, the blue one usb 3.0, with the other no power (withe one).
2) Is it just me or in the cap next to Inductor1 i see a little crack?

Attachments

[ccc]

Deleted

[Pi0tR]

Wouldn't it be faster to buy a new drive and solder the memory and controller?

[fzabkar]

It does look like a crack on that MLCC. There are some hairline marks on the PCB, too. ??

[giulio93]

Is KLFEG4R4AB an SSD-on-a-Chip? Does it incorporate both the flash controller and NAND flash on a single IC?

I don't know

[giulio93]

It does look like a crack on that MLCC. There are some hairline marks on the PCB, too. ??

I can try to remove it. Do I have to replace it?
For the voltage I wrote do you see something strange? Also the V1, the last message you told me something about it but I don't get it if the tension is wrong or not.

[fzabkar]

If there is only one large IC on the PCB, then it's an all-in-one package.

[giulio93]

If there is only one large IC on the PCB, then it's an all-in-one package.

the other side of the board is empty.
What about the black IC? Is it not the controller?

[fzabkar]

The DCDxyz IC is not a flash controller. It is a PMIC.

[giulio93]

It does look like a crack on that MLCC. There are some hairline marks on the PCB, too. ??

I can try to remove it and measure the resistence. If it is good i will have a OL, isn't it?
The black IC can't cause the problem?
And pls tell me what is wrong with the V3

[giulio93]

It does look like a crack on that MLCC. There are some hairline marks on the PCB, too. ??

I remove that cap but it measure OL and the resistance on the board is like before.
So it's the nand or the controller. Isn't it?

[Hardcore Games]

Looking at the images I noticed some possible cracked caps which can lead to short circuits

[MDJ]

I also have an dead Samsung PM991a after 4 years of working in my laptop and 1.5 year of game development. It draws a lot of power and also the monolith ic (yes it is one and incorporates both NAND and Samsung Pablo S4LR033 Controller also found on t7 and 980 non pro), and the little ic in the bottom part of the main ic is an PMIC with 3 coils. Mine is still visible on internal PCIe (the controller is seen by BIOS and Windows with the correct model and capacity) and DMA off stuck in D3Cold power state. The monolith heats after ~1 minute of power after the PMIC is heating a milisecond after getting power. I also have an working donor and I will try to transfer the monolith to this one, maybe you should also get one and try this, don't loose your time with those small caps near the monolith because they are more hard to remove than the monolith itself and a donor is very cheap. Mine suddently stopped working after short time an 30GB download was finished on my external HDD and I was just running chrome in background. A day before I checked the NAND life with CrystalDiskInfo and was still 100%. Maybe my diagnosis will help you (and also me) on fixing this PM991a.