Crucial M4 SSD issue (Micron C400)

[cuebus]

Hi all,

Long time lurker here, first time posting.

I have a crucial m4 ssd that recently died on me. I'm pretty much a beginner at troubleshooting this kind of hardware so I figured I'd take a shot and see if any of you all out there could help me out.

The symptoms:
-Drive does not show up in BIOS
-POST takes a very long time when ssd is connected to mobo, as if it is detecting something

Some things I've already figured out
The 5V rail does seem to be ok as all the different outputs from the LTC3545 regulator are at their correct voltages
All 0 ohm resistors jumping to the SATA connections measure <0.1ohm, so no problem there

I'm not sure what the pad labeled "5VF" means, but when the drive is powered on it jumps to 5V for about 2-3 seconds, and then drops to a few hundred millivolts. During this time there is also a high pitched screeching sound that is heard, like a CRT might make.
I'm curious if the behavior of that rail is normal, and I was curious to know if anyone could identify the components that I point out in the picture, marked AV1 and CC and what their role might play.

I have another one of these drives on the way so I'll probably crack that one open to compare voltages etc.

Any help is appreciated, thanks guys

[poot36]

I would check all capacitors for shorts and that 5vf rail for shorts or leaky semiconductors.  The high pitched sound you hear is a switchmode power supply for one of the chips that is not working correctly or is overloaded.  Most of these drives should have a 3.3V rail for the flash chips and a 1.8V rail for the controler.  Try and find a data sheet for all of the chips and the expected voltages for them.  Also there may be a serial port for debugging as a unpopulated header or as test pads.  If you can find a serial port see if there are any error messages or any data at all coming out of it.  Make sure to use the correct voltage for the serial port (most likely 3.3V but could be 5V, most desktop serial ports are 12V and most laptop serial ports are 5V to save on power).

[cuebus]

Thanks poot36.

I received a second of these drives yesterday and it seems that in fact there is no problem with any of the voltage rails on my unit. The 5VF rail works as it should on the old unit, compared to the new unit. In fact, everything I have probed so far seems to be identical. Nonetheless, the new unit does not give the high pitched screech at startup and, well, it actually works.

The unit seems to have a jtag header on it. Can JTAG read out flash data? I'm new to digital debugging- does anyone have any recommendations on a JTAG unit to purchase? Bus pirate seems like a decent option but I want to make sure.

[poot36]

If you want to try JTAG on the cheap make a Wiggler parallel port cable.  Just a few resistors and diodes.  I have read online that you can try plugging the drive into just power for 20 minutes and then unplugging it for 30 seconds then plugging it with the data cable back in to the computer and see what happens.  You may have to do this a few times.  If you do get it working backup anything that you really need off the drive and then update the firmware.  Good luck finding the JTAG pinouts.

[mariush]

You have test points on the pcb .. put a probe on the ground (black wire from psu)  and the other probe on that pad that says  1v0 (should be 1.0v) , 2v5  (should be 2.5v) or the one that says 1v5  (for 1.5v).

1.0v and 2.5v is probably used by the marvell controller chip, the 1.5v is probably required by the flash memory chips .. I don't know what 5vF would mean, i would guess some code word for "dirty 5v", meaning whatever comes from the power supply through the 5v pins in the sata power connector.

You could use the continuity function of your meter to see if that pad is directly connected to the 5v pins in the sata connector .. the 5v should be coming in through those 3 pins in the center of the power connector which then connect to a copper island through two 0 ohm resistors.

[mikerj]

Just a guess, but the AVI marking on the 5 pin SOT23 is correct for a TPS3801T50DCK supply supervisor.  5VF could be 5v Fail.

[ConnorGames]

I would probably forget about trying to dump the flash. All major SSD vendors use crazy wear-leveling and speed boosting algorithms that make the raw data in the flash chips pretty much gibberish without the right tools. And the chances of you (or anyone outside the biggest data recovery companies and the manufacturers themselves) getting those tools is just about nil.