HDD GURU FORUMS

Here it is...

pcimage wrote:Here it is...

Sean ,
What Are You Trying to Do In It [ Research or Data Recovery ] . Its Interesting To Know How In So Many NAND Flash Memories You Find The Sequence of Chips like 1...2...3... .

Drive is dead.

Voltages normal coming in...

But on the back-side of the PCB's the capacitors C166,167 and 168 get very hot.

Voltages measured from ground for the capacitors and the transistor.

I'm no component-level electronic expert, so any help/advice gratefully received

If the caps are getting hot its probably a ripple coming from the VRM..

get a scope on the VRM outputs to see whats going on..


thing is the ripple could be the result of 1 of 2 scenarios..

1. the VRM is faulty and outputting a ripply mess..
2. the VRM could be overloaded by a faulty component taking it outside of its drive specification causing the ripple.

hope that helps


Marko

pcimage wrote:Drive is dead.

Voltages normal coming in...

U37 most probably (90%)

If the capacitors are getting hot it's means that they are short circuited on the ground via other elements. As Marco said, first need to check voltage convertors and fuses. If it's OK, so CPU is damaged. In this case the only way is to unsold memory IC's and work without CPU. There is the chance to have success because previous Indilinx CPU's (like IDX110M00) have no hardware encryption (like Sandforce), so you CPU (IDX210M00-FC) probably has no hardware encryption as well.

AFAICT, C166, C167, and C168 are connected across the incoming +5V supply via inductor L1. Together with L1 they comprise a low pass LC filter which is intended to remove high frequency noise from the supply. That said, the voltage at L1 is 5V, yet the voltage at the caps is 4.5V and 0.5V. ISTM that the voltage at the caps should be 5.0V and 0V, in which case you are using the wrong ground reference for your measurements.

It would be very helpful to know the part markings on U37, but AFAICT it consists of 3 buck regulators. I would measure the following voltages (using the correct ground reference):

U37 is the Power IC
as far as i know (from older cases)

fzabkar wrote:That said, the voltage at L1 is 5V, yet the voltage at the caps is 4.5V and 0.5V. ISTM that the voltage at the caps should be 5.0V and 0V, in which case you are using the wrong ground reference for your measurements.

You're right, using a different ground point, the caps are read 5.0 & 0V.

Sorry for the error

Suitsof wrote:If the capacitors are getting hot it's means that they are short circuited on the ground via other elements. As Marco said, first need to check voltage convertors and fuses. If it's OK, so CPU is damaged. In this case the only way is to unsold memory IC's and work without CPU. There is the chance to have success because previous Indilinx CPU's (like IDX110M00) have no hardware encryption (like Sandforce), so you CPU (IDX210M00-FC) probably has no hardware encryption as well.

I don't really want to go down the route of "chip off" if I can help it, as neither PC3K or SC appear to have support (yet) for this controller, though as you say it's likely to be similar to IDX1100M00.

The markings on U37 are:

RT9991
GQV
EG94B

Which appears to be a "Richtek" PMIC

fzabkar wrote:
It would be very helpful to know the part markings on U37, but AFAICT it consists of 3 buck regulators. I would measure the following voltages (using the correct ground reference):

U37.jpg

V1 reads 0.45V average, it does waver between 0.4V & 0.5V
V2 reads 1.8V
V3 reads 1.0V
+5V reads as it should.. 5V

V1 is the NAND supply. Something wrong there ...

V2 appears to be the supply for the SDRAM, while V3 is the Vcore for the flash controller. ISTM that these two are OK.

I would start by confirming that pin 7 is high.

If pin 7 is low, then the SSD may be powering up in standby mode. Windows 10 again?

fzabkar wrote:If pin 7 is low, then the SSD may be powering up in standby mode. Windows 10 again?

Powering up via PC3K (which is showing the ATA registers greyed out), so no Windows to get in the way!

Pin 7 is reading 5V

Whereas pins 2,3 & 4 (LX3) are fluctuating around 0.5V

The NAND supply is being enabled, so either the chip is bad, or there is an overload on its output. Measure the resistance between V1 and ground. If the resistance is low, then I would suspect one or more of the flash chips.

Got it working

C48 was the culprit, shorting the NAND side.

Cloning now....

Many thanks to all that contributed, most helpful!!!

Super stuff