All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 113 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 16:03 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Quote:
I suspect that the MSP430V952 microcontroller may control the power sequencing for the PCB. It would be interesting to see whether it is getting hot and whether it is powered from the OAH LDO regulator.

It doesn't seem to get hot in the FLIR image. I took another snap and it doesn't seem to generate any heat at all.
In fact, it's not even the regulator that gets hot but the two 10ohm resistors near the Vin and Vout pins do.

I'll remove that cap and organize all the measurements you asked.
Edit: with cap removed the drop remains the same.


Quote:
Edit #2: Measure the resistance between the 0.9V point and ground.

5.2 Ohms.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 16:21 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
That's what I was expecting. :-(

It appears that pin #23 may be the MCU's supply pin, or could it be pins 27, 29 or 30? Are any of these pins shorted to ground? Is the adjacent component also a 10 ohm resistor?

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 16:49 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Quote:
The QY device is a 2.5V regulator. Its input voltage is 0V. Can you check whether the RAU, ODH, and MP5505 parts are also missing their input voltages?
Their Vin is at 5.0V but EN is low at 0V. For the ODH, enable is 0V as well and PVIN/AVIN are ~0.3V.


Quote:
One other thing to do would be to measure the resistance between ground and the outputs of each of the regulators. This will tell us if the flash controller or any of the NANDs are shorted. If any of these ICs are damaged, then there would be no point in continuing.
Good call, seems we're good. Only V1 measures about 50Ohms to ground, the others are much higher.

Quote:
Can you measure resistances Rx and Ry, and voltage Vy, in the MSP430 image?

Attachment:
MSP430.jpg
MSP430.jpg [ 80.63 KiB | Viewed 19179 times ]


Quote:
I suspect that Vy should be +5V. If so, then could you test the resistance between this point and the SATA +5V pins and the Drain pins (output side) of the power MOSFET?
Between Vy and +5V on SATA connector I measure about 1.5M Ohm. Between +5V on SATA connector and PMOS drains is 25.0 Ohm.

Quote:
It appears that pin #23 may be the MCU's supply pin, or could it be pins 27, 29 or 30? Are any of these pins shorted to ground?

Pin 23 is multiple k-Ohm to ground, can't get a stable reading. Pin 27,28 are 0 Ohm to ground, pin 29,30 are 5 Ohm to GND.

Quote:
Is the adjacent component also a 10 ohm resistor?

Which one? If you mean the one next to the 3.3V regulator (Rx in the pic), then yes.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 17:17 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
It looks like pin 29 or 30 of the MCU is the supply pin, in which case the MCU is shorted. :-(

Pin #23 is an I/O pin with an external pullup resistor, so we can ignore it.

The OAH regulator's input voltage of 4.45V must be coming from the +5V supply via an additional component. Perhaps the mysterious Schottky diode is in series with this supply?

The low resistance (50 ohms) at V1 would be normal if V1 were the Vcore supply for the flash controller. I wouldn't be concerned with that.

In short, ISTM that the MSP430 is the power sequencer/controller for the PCB. I expect that each of the regulator EN pins will connect to the MCU's I/O pins. This would be the proof that the MCU is the heart of it all.

Assuming there are no other faults, then ISTM you would need to replace the MCU. Alternatively, assuming that the MCU is a standalone device that does not need to communicate with the flash controller IC, then we may be able to manually EN-able each of the regulators, but we would need to do this in the proper order, otherwise the SSD may not power up correctly.

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 17:48 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
fzabkar wrote:
The OAH regulator's input voltage of 4.45V must be coming from the +5V supply via an additional component. Perhaps the mysterious Schottky diode is in series with this supply?

You're right. Anode measures 0 ohms to +5v SATA supply and cathode is connected to the 4.45V we're seeing in various places.

Quote:
In short, ISTM that the MSP430 is the power sequencer/controller for the PCB. I expect that each of the regulator EN pins will connect to the MCU's I/O pins. This would be the proof that the MCU is the heart of it all.
I've done a quick test and found that at least two of the RAU regulators have their EN pin directly connected to one of the MCU pins. Since MCU and regulators are on different sides of the PCB it's somewhat hard to probe but I'm sufficiently confident to say the MSP430 needs replacement! I'll see if I can source it and report back once I've replaced it.


Edit: actually, I may need some help replacing. Since this isn't a passive component, I probably need a way to put the firmware in the microcontroller don't I?
At least my universal programmer supports this chip.

About device selection, it seems the difference between MSP430x11x2 and MSP430x12x2 is SPI/UART availability in the latter and the difference between 1x22 and 1x32 is flash memory size. I'm guessing the 1232 is the most compatible one and should always work. Does that make sense or would I need the exact replacement?
Can I go wrong with picking the 'largest' device from the device family, i.e.


Last edited by zzattack on August 7th, 2016, 18:00, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:00 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
zzattack wrote:
Edit: actually, I may need some help replacing. Since this isn't a passive component, I probably need the firmware in the microcontroller don't I?

Exactly. :-( It would be programmed with the SSD's power-up sequence. Perhaps it is also involved in subsequent power management (APM), in which case it would need to communicate with the flash controller.

There is potentially one other problem. The power MOSFET is stuck in the ON state with a D-S resistance of 25 ohms. When we start to draw current from it, we may find that its output falls. You should ideally replace this MOSFET as well ... although I'm wondering if we can simply run a shorting link between D and S. Of course there is the question as to why it is not switching off. Is there some component on its output that has been compromised as well?

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:07 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Replacing the MOSFET shouldn't be a problem at all.

According to the datasheet the MSP430 has 'code protection'. The flash memory can be programmed via the JTAG port, the bootstrap loader, or in-system by the CPU. Now I wonder if the code protection applies to both JTAG and the bootstrap loader.. I'm starting to fear we're out of luck.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:16 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
zzattack wrote:
About device selection, it seems the difference between MSP430x11x2 and MSP430x12x2 is SPI/UART availability in the latter and the difference between 1x22 and 1x32 is flash memory size. I'm guessing the 1232 is the most compatible one and should always work. Does that make sense or would I need the exact replacement?
Can I go wrong with picking the 'largest' device from the device family, i.e.

I can't advise you on this. :-( Chip manufacturers often manufacture special versions of their MCUs for OEM customers. I can't tell from the "V592" marking if the chip is compatible with either of those parts.

In fact there are numerous search results for "MSP430" at TI's website:

http://www.ti.com/packaging/docs/partlookup.tsp?searchType=partmark&keyword=&partmarking=MSP430&partNumber=

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:20 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
zzattack wrote:
Replacing the MOSFET shouldn't be a problem at all.

According to the datasheet the MSP430 has 'code protection'. The flash memory can be programmed via the JTAG port, the bootstrap loader, or in-system by the CPU. Now I wonder if the code protection applies to both JTAG and the bootstrap loader.. I'm starting to fear we're out of luck.

Is it worth gambling on a used SSD for spare parts?

Alternatively, at the [probably low] risk of causing destructive latchup, you could remove the MSP430 and allow the EN outputs to float high (or pull them up with resistors). Otherwise, if we knew the correct order for each of V1-V4, then we could manually switch them on, one at a time.

That said, does the flash controller need to see the MSP430, and would it panic if it didn't find it?

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:31 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Was thinking the same. If I can snag one for cheap then it's likely a simple transplant.

I've just removed the MSP430. Now the regulator outputs a nice 3.3V as expected. The EN pins appear to be pulled down though.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:33 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
Could we see the area under the MSP430 chip? Does it look like it might be a standalone MCU?

BTW, if you have a CCD scanner (not CIS), this produces much better images than a typical camera.

_________________
A backup a day keeps DR away.


Last edited by fzabkar on August 7th, 2016, 18:36, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:35 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Absolutely.

Attachment:
IMG_0102.JPG
IMG_0102.JPG [ 2.11 MiB | Viewed 19158 times ]


I can also confirm that all 4 enable pins of the 3x RAU and 1x ODH chips trace directly to one pins for the MSP430.


Quote:
BTW, if you have a CCD scanner (not CIS), this produces much better images than a typical camera.

I have one but it's currently at my parents'. I'm gonna pick it up so I'll be able to get a proper scan on Tuesday.
Each of the 4 scanners I have is CIS... but 3 of those are part of multifunctionals so I didn't expect good quality from it. The other is a recent Canon Lide model but the uneven depth generates absolutely useless pictures so I had to use a poor digital camera for now :(


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:42 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
Thanks. Another thought is that the MCU might be monitoring the actual voltages and generating a PowerGood output for the flash controller.

If all chips use a common enable pin, then that makes it a lot easier. Perhaps we don't need to worry about power sequencing. Would it be OK to add a 1K (?) pullup resistor between the EN pins and the 3.3V supply?

The MP5505 has its own enable pins. Do you think it would matter if this chip were switched off? AIUI, its function is to provide a voltage for the backup capacitors and to keep the SSD alive in the event of a sudden power failure.

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:48 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
If that's the case, would something else enable the regulators? Is the ADC periphal capable of doing this?

Also you've mentioned twice that the order of enabling the regulators is of importance. Is that something common to SSD/HDD chips or is it possible the order may not matter at all?

I could lift the EN pins from those chips and toggle them manually. At 16 combinations finding the correct one shouldn't be too hard. But then again, transplanting from a used replacement sounds like a better option.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:52 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
fzabkar wrote:
If all chips use a common enable pin, then that makes it a lot easier. Perhaps we don't need to worry about power sequencing. Would it be OK to add a 1K (?) pullup resistor between the EN pins and the 3.3V supply?
My bad, I should clarify: the 4 enable pins all trace back to different pins on the MSP430.

Quote:
The MP5505 has its own enable pins. Do you think it would matter if this chip were switched off? AIUI, its function is to provide a voltage for the backup capacitors and to keep the SSD alive in the event of a sudden power failure.
I agree, the 5505 seems only for power storage/backup functionality, so I doubt it's important for normal operation.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 18:58 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
In HDDs which use ARM cores, ISTM that the Vio comes up before Vcore. However, I have seen several examples of failed PCBs where Vio was absent while Vcore was present, and where the board was OK after Vio was reinstated. Computer motherboards often don't reset properly if the supplies aren't sequenced properly, but I don't recall this being destructive.

Would it be worth determining the functions of V2-V4? To this end you could test for continuity between each supply and the filter capacitors adjacent to the NAND chips. This will identify the Vcc and Vccq supplies.

You might also check whether the MSP430 is monitoring the V1-V4 supplies. If so, then we may need to fake a PowerGood signal.

_________________
A backup a day keeps DR away.


Last edited by fzabkar on August 7th, 2016, 19:03, edited 2 times in total.

Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 19:00 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
Unfortunately the 4 enable signals all route to different pins on the MSP. I'm tracing them now and I'll pull them up.
The ODH mode switch signal also traces back to the MSP430.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 19:13 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
There is an 8-pin serial flash memory. You should probably check to see where it gets it supply.

The SDRAM could be a 2.5V or 1.8V device, so it may have its own regulator (LDO?).

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 19:25 
Offline

Joined: March 28th, 2014, 9:11
Posts: 44
Location: Nederland
The RAU output voltages don't trace back to the MSP430 but they do have traces to the BGA of what I think is the micron controller.
I'll try to source a used part tomorrow and if that leads nowhere I'll see what happens if we force-enable the regulators and fake the power good signal.


Top
 Profile  
 
 Post subject: Re: Crucial MX100 repair after PSU failure
PostPosted: August 7th, 2016, 19:29 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
88S9189-BLD2 is the Marvell flash controller. It should receive Vcore (~1V) and Vio voltages.

The D9RLT part is an SDRAM (Google identifies it as MT42L256M16D1).

https://www.micron.com/~/media/document ... it_aat.pdf

_________________
A backup a day keeps DR away.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 113 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 39 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group