OCZ Vertex 3 120GB SSD Repair - U10 Chip Burnt Out

[thewisewizard]

Hello,
I'm currently trying to diagnose several HDDs and SSDs that experienced a power surge in the same PC. Physical inspection shows an obvious issue, the chip in the U10 spot (labeled AMPB L128) is dead with a hole in the top. Inspection under a thermal camera shows the chip heating up fast when power is supplied. I've also purchased an identical drive for comparison/harvesting, the same chip on the good drive does not get hot like the bad chip. My questions are:

1. Does anyone know what the U10 chip is/contains (or have a diagram of this SSD with details on the chip)?
2. Can I swap in the good chip?

I scoured the internet for info on the U10 AMPB L128 chip but have not found anything useful so far.

I've attached images of the donor drive and the U10 chip for reference.

Thanks in advance for any help.

Attachments

Donor Drive PCB U10 Chip

Donor Drive PCB Image

[fzabkar]

"128" would most likely be a YWW date code, ie week 28 of 2011.

If you measure the voltages on C30 and C31 on your donor PCB, that should help us to identify the function of U10. For example, they could be the filter caps for Vin (5V) and Vout. That would then suggest that the part is a converter, either linear or switchmode (with on-chip inductor).

[fzabkar]

EP53A8HQI, Enpirion, Synchronous Buck Regulator with Integrated Inductor, 1A, 2.4V - 5.5Vin, marking AMxx, QFN-16:
https://www.farnell.com/datasheets/1525815.pdf

[Amarbir[CDR-Labs]]

EP53A8HQI, Enpirion, Synchronous Buck Regulator with Integrated Inductor, 1A, 2.4V - 5.5Vin, marking AMxx, QFN-16:
https://www.farnell.com/datasheets/1525815.pdf

Hi,
I have a question ,How do you find this out ... ? , This has Three pin whose state determines the output voltage ,Might be after taking this chip out you can check the voltage on VS0,1,2 , That will help is finding out what voltage was it giving out .I also would like to ask the party that started this thread to please post photos of the SSD from both sides along with top cover so that we know what we are dealing with

[Lardman]

I have a question ,How do you find this out ... ?

I found nothing either (other than LED driver stuff) - Fzabkar has some sort of secret google foo when it comes to finding them.

[Amarbir[CDR-Labs]]

I have a question ,How do you find this out ... ?

I found nothing either (other than LED driver stuff) - Fzabkar has some sort of secret google foo when it comes to finding them.

No ,
He has always been kind enough to explain how he finds it ,We never see what he writes properly .

[Arch Stanton]

I have a question ,How do you find this out ... ?

I found nothing either (other than LED driver stuff) - Fzabkar has some sort of secret google foo when it comes to finding them.

I think he's an AI that taps in to Google directly.

[Lardman]

..... Or he has kidnapped employees of the major chip fabricators and keeps them chained up in his basement, only allowing them home after identifying enough components to balance out their evil doings.

[thewisewizard]

Wow thank you all for the quick responses.

"128" would most likely be a YWW date code, ie week 28 of 2011.

If you measure the voltages on C30 and C31 on your donor PCB, that should help us to identify the function of U10. For example, they could be the filter caps for Vin (5V) and Vout. That would then suggest that the part is a converter, either linear or switchmode (with on-chip inductor).

Ah, yes I see now the bad chip has a different date code. I will measure the voltages of those caps and report back...

EP53A8HQI, Enpirion, Synchronous Buck Regulator with Integrated Inductor, 1A, 2.4V - 5.5Vin, marking AMxx, QFN-16:
https://www.farnell.com/datasheets/1525815.pdf

As others have stated I have no idea how you found this so quickly, but this is very much appreciated.

EP53A8HQI, Enpirion, Synchronous Buck Regulator with Integrated Inductor, 1A, 2.4V - 5.5Vin, marking AMxx, QFN-16:
https://www.farnell.com/datasheets/1525815.pdf

Hi,
I have a question ,How do you find this out ... ? , This has Three pin whose state determines the output voltage ,Might be after taking this chip out you can check the voltage on VS0,1,2 , That will help is finding out what voltage was it giving out .I also would like to ask the party that started this thread to please post photos of the SSD from both sides along with top cover so that we know what we are dealing with

Sorry about that I will attach additional images...

I think I have so info now to start testing. Admittedly I am a novice in terms of diagnosing/testing...are there any specific tests you guys recommend I try first (outside of any already mentioned above)?

Attachments

Donor SSD Front Cover

Donor SSD Back Cover

Donor PCB Front

Donor PCB Back

Dead SSD

Burnt Chip

Burnt Chip Closeup

[fzabkar]

This is my go-to resource for smd part markings:

http://markingcodes.com/

I unsuccessfully tried several variations -- AMPB, AMP, AMPx, AM, AMxx.

An alternative is Google's advance search:

https://www.google.com/advanced_search

I tried the same variations, without results:

https://www.google.com/search?q=filetype%3Adoc+OR+filetype%3Apdf+%22absolute+maximum%22+converter+qfn-16+%22AMPB%22

I use "absolute maximum" to filter out spammy part suppliers.

I examined the nearby components to see whether their context could provide a clue:

https://www.mouser.com/datasheet/2/612/en6337qi_05800-1843177.pdf (EN6337 datasheet)

The EN6337 is a "Step-Down DC-DC Switching Converter with Integrated Inductor". I wondered whether the mystery IC might be a similar component, so I tried a more specific search.

https://ipv4.google.com/search?q=filetype%3Adoc+OR+filetype%3Apdf+%22absolute+maximum%22+converter+qfn-16+%22integrated+inductor%22

This time I saw several Enpirion candidates.

This one (EP5368QI) appeared to have a pinout that matched the OP's photo clip:

https://docs.rs-online.com/68ce/0900766b8126c889.pdf

The part marking on page 2 matched the layout on the OP's IC (AAxx + LYWW).

For visual confirmation, I examined the images at Aliexpress and Google:

https://pl.aliexpress.com/w/wholesale-EP5368QI.html (EP5368QI, AAxx)
https://www.google.com/search?q=EP5388QI&tbs=isz%3Am%2Citp%3Aphoto&tbm=isch (EP5388QI, ABxx)

I noticed that Enpirion's datasheets described the converter as having an "integrated inductor" and a "16-pin QFN" package, so I now refined my Google search:

https://www.google.com/search?q=filetype%3Adoc+OR+filetype%3Apdf+%22absolute+maximum%22+converter+%2216-pin+QFN%22+%22integrated+inductor%22+%22enpirion%22

It was now just a matter of examining each candidate for the desired marking code, AMXX.

Google Images and Aliexpress confirmed the part markings.

In this case I couldn't use "AMxx" as a search term because it appears as a graphic rather than text in Enpirion's datasheets.

[fzabkar]

I think I have so info now to start testing. Admittedly I am a novice in terms of diagnosing/testing...are there any specific tests you guys recommend I try first (outside of any already mentioned above)?

After you have removed U10, measure the supply voltages produced by the other Enpirion ICs (U4 and U8). These will be present on the associated capacitors. Use a screw hole as your ground reference.

U5 (FDS6898A) is a dual MOSFET. It is probably used as a load switch, in which case there would be equal voltages at each of its drain and source pins.

https://www.mouser.com/datasheet/2/149/FDS6898A-1008367.pdf

Once you have replaced all the faulty ICs, you can monitor the UART port (J1, TX, RX) for diagnostic messages. You would need a USB-TTL RS232 adaptor.

[Amarbir[CDR-Labs]]

Hi thewisewizard ,
One thing is that if you are not so confident ,Then do not attemt this give to someone who is good with electronics and soldering desoldering , The outut voltage is controller by the state of three pins vs0 vs1 and vs2 . Plus even if you replace the components might be there is a short before or after that component ,This also would require to be tackled .

[thewisewizard]

Just a quick update before any solder work, continuity testing of the C30 and C31:

Good drive:
C30 - .737
C31 - .864

Bad drive:
C30 - .000 shorted
C31 - .611

Assuming removing that fried chip will eliminate the short…

[fzabkar]

C30 is the output of U10. I would desolder that chip and retest for shorts. If the short persists, then one of the main ICs is probably damaged. Otherwise there could be a shorted capacitor on that rail, but this would be an unlikely consequence of a "power surge".