Shorted 9040VM509 chip from Sandisk SSD

[michael chiklis]

Hello everyone,

i've this 480GB Sandisk Ultra II SSD which doesn't power on.
I've found a lot of shorts on caps around 9040VM509 chip (glass package), or with very low resistance (as 4 or 10 ohms).
This chip (located on U11 position) gets pretty hot, i also did the isopropyl alcohol test which evaporates faster on that chip, this means that most likely 9040VM509 chip is shorted.

Problem is that i can't find this chip anywhere, even if i search ACT9040VM509-T as product number.
Do you know where can i find it?
Maybe the only way how to get it is buying donor ssd, but if i do so i will need to reball all those very tiny bga pads

P.S.
I've also found cracked KG8 diode located on D2!

cracked diode.jpg (34.88 KiB) Viewed 13048 times

[fzabkar]

KG8 looks like a 5V TVS diode, so it appears that the SSD was clobbered by an overvoltage. Since you have voltage at the PMIC, then the diode must have been blown open circuit. The overvoltage would then have hit the PMIC.

If the shorts at the PMIC are on the output side (ie between inductors and ground), then one would have to be concerned about the main ICs (controller and NAND). Be aware that the Vcore supply normally measures in the low ohms range.

[michael chiklis]

Hi Frank,
is it ok to replace the KG8 diode with a common 5V TVS diode from a 2.5" drive?

The diode is not shorted but it is 15 Kohm on both directions measured on pcb.

[fzabkar]

IMHO there is no point in replacing the diode, assuming it is a TVS diode.

AISI your problem is to determine whether the damage to the PMIC was confined to that chip. The easiest way to do this is to remove each inductor (L7, L8, L9, L10) and measure the resistance between ground and each of the pads of the inductor. Don't mix up the inductors. Also measure the resistance between each inductor and the 5V supply. That will tell us if the surge passed through the switching element and onto the load.

[michael chiklis]

I removed L7, L8, L9, L10 but all those shorts on caps are still present.

This are the resistance test i did on inductors pads
L7: 0 to 155 KOhm (the range must be due to some cap charge in parallel)
L8: 0 Ohm
L9: 0 Ohm
L10: 0 Ohm

Resistance between inductors pads and GND
L7: 0 Ohm (pad 2) - 0 to 155 KOhm (pad 1)
L8: 0 Ohm (both pads)
L9: 0 Ohm (both pads)
L10: 0 Ohm (both pads)


Resistance between inductors pads and 5V line (multimeter red probe on 5V connector)
L7: 1,5 MOhm (both pads)
L8: 1,5 MOhm (both pads)
L9: 1,5 MOhm (both pads)
L10: 1,5 MOhm (both pads)

Resistance between inductors pads and 5V line (multimeter COM probe on 5V connector)
L7: 1 MOhm (both pads)
L8: 1 MOhm (both pads)
L9: 1 MOhm (both pads)
L10: 1 MOhm (both pads)

[fzabkar]

Your measurements indicate that the outputs of each of the buck converters is shorted to ground inside the IC. The loads are probably also shorted, although it is not clear if this could be due to an internal short to ground via the FB (feedback) pin.

Since the PMIC is bad, you have nothing to lose by removing it. I would then retest the capacitors. If the shorts remain, then I would put my bets on the other ICs being bad rather than the capacitors.

Attachments

[michael chiklis]

I thought since when i started the diagnostic the PMIC was faulty.
PMIC is a wafer-scale ic, they're very sensitive. The alcohol drying test near to never fails indicating where is the short (it is a simple and inexpensive test in the absence of a thermal cam).
Tomorrow i'll try to remove the pmic and let u know.

Thank you.

[fzabkar]

A PMIC can become hot if it is driving into a shorted load, so a heat test is not always conclusive.

[michael chiklis]

After talking with customer we decided to stop here, so i will not even try to unsolder the pmic.
Data are very important for him so i suggested to contact a pro which has a lot more experience with bga reworks.

Thank you for your support Frank