Shorted Monolith = No Money :-(

[jeremyb]

AFAIK there are two variates on this monolithic. The Phison version and SSS version used by Kingston. I have no problem recovering the Phison version but for some reason every SSS version fails the same way, controller overheats, sometimes the controller logic works and it will boot into flash mode, sometimes not. The NAND chip isn't responding to any logic sent from the controller and the +3.3v rail is shorted to ground @ 5-20ohms. Voltages from logic signals are all low 0.9v - 2.4v.

I've tried removing the controllers silicon die, incrementally cutting the exposed 3.3v rail trying to find the short (which made no sense because cut segments still shorted to ground).. Perhaps the NAND IC is physically damaged. IDK..

Any ideas where this phantom short is originating from?

Attachments

[arvika]

From my experience if VCC and GND is shorted - usually it means that nands module is damaged (crack, overheat etc). Maybe it is common failure for this model - like factory bad designed production.

[jeremyb]

From my experience if VCC and GND is shorted - usually it means that nands module is damaged (crack, overheat etc). Maybe it is common failure for this model - like factory bad designed production.

I think you are correct, I isolated the NAND chip and got the same results. PCB is .4mm, NAND die is .2mm, easy to break NAND....

[einstein9]

Honestly jeremyb & arvika

I Admire your work (both of you)
in Addition to Sasha of course

Good to know you guys really

all the best in 2015 :>

[Matiw]

jeremyb,

Did you split the PCB from the edge ? For my eyes, it seems sort of a physical impossibility ! That is amazing.

And the PCB trace picture you posted, it is kind of very clean, did you use a chemical ? Do you mind eh eh... telling ?

Thanks

From my experience if VCC and GND is shorted - usually it means that nands module is damaged (crack, overheat etc). Maybe it is common failure for this model - like factory bad designed production.

I think you are correct, I isolated the NAND chip and got the same results. PCB is .4mm, NAND die is .2mm, easy to break NAND....

[jeremyb]

I had a cracked identical donor and took it apart. I took pictures of the inside (PCB, wire bonds, controller, etc). I'll write an article, might be fun eyecandy...
I looked at my notes, minor correction: PCB height is 0.47mm, NAND height 0.15mm.

[jeremyb]

http://www.recovermyflashdrive.com/crac ... nolithics/

You can click on an image and then click full screen on the top right to see a higher resolution version...
Let me know what you think

Opacity and Angle are pretty cool

[arvika]

Good, so it confirms my suspicions

[HaQue]

Nice write-up Jeremy.

One thing I learnt is that it won't help stripping back a monolithic to find out the pinout., as they don't seem to use any recognisable pin(bond) layout.

[jeremyb]

Nice write-up Jeremy.

One thing I learnt is that it won't help stripping back a monolithic to find out the pinout., as they don't seem to use any recognisable pin(bond) layout.

I figured as much, in my case I meant +3.3v is connected to the contact point above it on SSS, thats why I thought the pin-out may be different.

[networkpc3000]

this type of with very bad quality. i meet many with short internal. some i can recovery. some is flash dead. 50% for this type of monolith success with short condition. but i think for your case may no chance. it's SSS6697. i have same. have been try
two version :
1,it's SSS669X controller short almost no chance.
PS. i meet 3.4 cases SSS6696~98 with dead TSOP48 NAND flash(no monoith) i think it may something wrong with these controller
2, PS2251 has big chance. i think 95% or above. i never see dead NAND with it even monolith short

[jeremyb]

I see the same, Phison version has a high success rate, SSS low. All the SSS I see come from Kingston, the one I looked at has an internal crack where the connector changes from metal to plastic which isn't visible externally only internally.

[networkpc3000]

good stuff of your website. this design really big chance make monolith crack. but i am thinking invisible externally monolith may no internal crack for SSS. i think it may controller kill it? do you meet SSS6696~8 normal usb flash drive(tsop48) with dead flash? anybody meet?

[networkpc3000]

the one I looked at has an internal crack where the connector changes from metal to plastic which isn't visible externally only internally.

agree some cases like this.