Crucial MX300 - No SATA connection

[outbreak198x]

Hello,

need again your opinions

i have this MX300 Crucial drive with 1050GB - it's not a BSY problem... it does not get SATA connection.

it takes 0,27A on 5V - so that looks not so bad

the question for me is now - is this problem of SATA controler? is the controler inside the Marvel Chip, because it looks like that?

is the chip unique, or is it possible to swap, because i have some from 525GB drives.

it looks like there is no short on the pcb, compared with values from 525GB drive it also looks fine. (for the parts i can measure)

should i try to connect terminal, because found 525GB version article before mine for BSY problem - but i think there won't be any output

anyone knows the safemode Pins? tried the classical ones, but nothing happens.

Marvel Chip is getting warm.

[Lardman]

Have you been here yet? SS881074,
https://blog.acelaboratory.com/pc-3000-ssd-sandisk-ultra-ii-and-wd-blue-marvell-cpu.html

[outbreak198x]

Have you been here yet? SS881074,
https://blog.acelaboratory.com/pc-3000-ssd-sandisk-ultra-ii-and-wd-blue-marvell-cpu.html

yes, but my drive does not get BSY and acelab told me, it's not possible, because crucial has his own Firmware on the controller.

[outbreak198x]

just touched the chips, and found out, that one diode is really getting warm instantly, don't know if it's normal, but i think it can burn your fingers, when you let it on...
my IR ThermoMeter can't get the temperature - but i think it's around 60-70 degrees.

diode.png (174.72 KiB) Viewed 10352 times

[Lardman]

fzabkar has posted test points for these a few times - check your voltages but if the diode is getting that hot something is shorted.

[outbreak198x]

i'm a small step closer - on the BIOS ROM Chip is no voltage - maybe a LDO is broken - have to find out

[Lardman]

I have very few SSDs to play with and no Crucial MX's to give you known working reading from I'm afraid.

[outbreak198x]

Looks like the 2,8V missing on the whole board - checked the 525GB PCB - and there are 2,8V spread on the half pcb... not here on the faulty board

[outbreak198x]

and reffering to this - https://forum.hddguru.com/viewtopic.php?f=10&t=36242

i don't have those 2,75V on the NL Chip - there are 1,1V on it...

[fzabkar]

The SSD appears to be stuck in soft start mode.

There are 4 x 100 ohm parallel connected resistors near the 5V SATA power input. Confirm that they measure 25 ohms. Also confirm that the zero-ohm resistors on each side of the PCB are not open.

In soft start mode these 4 resistors limit the supply voltage to the PCB. If you measure 5Vout, it will read maybe 2 or 3V in soft start mode. This voltage is sufficient to power the MSP430 microcontroller. This uC is programmed as a power sequencer and monitor. When this uC senses that all is OK, it switches on the load switch, and this then bypasses the resistor array. The SSD is now in run mode, and there should be 5V at the 5Vout test point.

I haven't measured these voltages myself, and I'm only basing my analysis on another Crucial model, so I could be wrong. If you measure the voltages at the 4 corners of the load switch IC, this will help to confirm or reject my analysis.

As for the diode, I can't see how it is connected. Is one end connected to ground? Is any pin connected to 5Vin or 5Vout?

Attachments

Soft_start_resistors.jpg (153.39 KiB) Viewed 10134 times

[fzabkar]

This is what I believe is happening:

Code:

             4 x 100 ohm
                ___
             .-|___|-.      TLV70028  .-------.
             |  ___  |        .----.  |       |        enable
        5Vin |-|___|-|5Vout   |2.8V|  |       |        pins of
SATA  o--+---+  ___  +---+----|LDO |--| MSP430|------> other
         |   |-|___|-|   |    |reg |  |       |        power
         |   |  ___  |   |    '----'  |       |        regulators
         |   '-|___|-'   |            '-------'
         |               |    "NL"        |
         |               |                |
         |   .-------.   |                |
         |   | load  |   |                |
         |   |switch |   |                |
         '---|---/ --|---'                |
          in |       | out                |
             '-------'                    |
                /|\                       |
                 | On/Off                 |
                 '------------------------'

Something is dragging down the input to the 2.8V LDO regulator ("NL"). The regulator is then only able to produce 1.1V rather than the full 2.8V supply required by the MSP430 uC. The uC never wakes up, and is then not able to turn on all the other regulator ICs.

I expect that 5Vout will measure around 1.5V or less. Try to find what else is connected to this point. If you are willing, you can short 5Vin to 5Vout and see which component gets hot. However it would be better to simply test the resistances. Look for something around 150 ohms.

Maybe the diode is the clue.

[fzabkar]

Look for something around 150 ohms.

Sorry, I meant 150mA or 10 ohms.

[outbreak198x]

Hello, and thx for this great inputs

but 5Vout = 4,95V

also on the back side the 0-Ohm Resistors have around 5V in out

but what looks strange to me is, that the LDO has nearly 5V on all pins - i don't think it should be like that?

[outbreak198x]

and on the Load Switch is 0V

[fzabkar]

The load switch which I have identified appears to be a load switch for the other supplies. It must be used in standby mode.

The SOT23 part you have identified (MP3) is a P-channel power MOSFET which appears to be acting as the load switch for 5Vout.

I believe that the 6-pin part marked "NL" (near the flash controller) is the LDO that powers the MSP430. It should be outputting 2.8V. Its input should be 5Vout. If the input is 5V, then either this LDO is faulty, or the MSP430 is dragging the voltage down, in which case the uC is faulty.

TLV70028, Texas Instruments, marking NL, 2.8V, 200mA, Low IQ, LDO Regulator, WSON, 6-pin:
http://www.ti.com/lit/ds/symlink/tlv700.pdf

[fzabkar]

My bet is that "NL" is faulty, otherwise any large current draw by MSP430 would have appeared as a voltage drop across the 100 ohm resistors.

I'm thinking that you could remove "NL" and inject 3.3V into the Vout pad, or 2.8V if you have a lab PSU. However, first confirm that the Vout pin does in fact connect to the Vcc pin(s) of the MSP430. Also measure the resistance to ground at Vout of "NL" to confirm that there is no overload.

Did you check whether the diode is connected between 5Vin and ground, or 5Vout and ground?

[fzabkar]

BTW, the reason that 5Vout measures 5V is not because the load switch ("MP3") is turned on (it is off, Vgs = -0.2V). Instead it is because the MSP430 is not drawing any current, so the voltage drop across the 100 ohm resistors is zero. I think ...

[outbreak198x]

i love so much input

i'll just change the NL with a new one - i have 1,8V in actually, and 1,2V on the faulty board now - so i think the fault is behind it - maybe really the uC is dead...

but anyway it's changed fast

[fzabkar]

So where is the 1.8V coming from? Something is wrong with my analysis ...

"NL" needs at least 3V to produce a 2.8V output.

Measure the Vcc pin of the MSP430. I'm not sure that "NL" is the source of this supply.

[outbreak198x]

in any case, NL is not dead - working on donorpcb

it's not easy to find out, because both sides...

i think i know where the 1,8V came from

resistor.png (172.92 KiB) Viewed 9852 times

maybe this should be just a 0-Ohm Resistor? should i try to short it - i'm so shy