Asus K53E Laptop with 0x00000124 BSOD

[carwiz]

That looks like a filter capacitor. It's high power so look for a "hole" next to a power transistor, diode or AC line.

The last dump indicates a 0x124 hardware error. More specifically, an L2 Cache Read Error. L2 Cache is located on the CPU chip itself. It's causing page faults so I'm not sure I would rely on maintaining data integrity. The missing capacitor may be associated with power to the CPU.

STACK_COMMAND:  kb
FOLLOWUP_NAME:  MachineOwner
MODULE_NAME: hardware
IMAGE_NAME:  hardware
DEBUG_FLR_IMAGE_TIMESTAMP:  0
FAILURE_BUCKET_ID:  X64_0x124_GenuineIntel_PROCESSOR_CACHE
BUCKET_ID:  X64_0x124_GenuineIntel_PROCESSOR_CACHE
Followup: MachineOwner
---------
1: kd> !errrec fffffa8006688028
===============================================================================
Common Platform Error Record @ fffffa8006688028
-------------------------------------------------------------------------------
Record Id     : 01cf00fd543a6244
Severity      : Fatal (1)
Length        : 928
Creator       : Microsoft
Notify Type   : Machine Check Exception
Timestamp     : 12/24/2013 23:11:15
Flags         : 0x00000000
===============================================================================
Section 0     : Processor Generic
-------------------------------------------------------------------------------
Descriptor    @ fffffa80066880a8
Section       @ fffffa8006688180
Offset        : 344
Length        : 192
Flags         : 0x00000001 Primary
Severity      : Fatal
Proc. Type    : x86/x64
Instr. Set    : x64
Error Type    : Cache error
Operation     : Data Read
Flags         : 0x00
Level         : 2
CPU Version   : 0x00000000000206a7
Processor ID  : 0x0000000000000001
===============================================================================
Section 1     : x86/x64 Processor Specific
-------------------------------------------------------------------------------
Descriptor    @ fffffa80066880f0
Section       @ fffffa8006688240
Offset        : 536
Length        : 128
Flags         : 0x00000000
Severity      : Fatal
Local APIC Id : 0x0000000000000001
CPU Id        : a7 06 02 00 00 08 10 01 - bf e3 ba 1d ff fb eb bf
                00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
                00 00 00 00 00 00 00 00 - 00 00 00 00 00 00 00 00
Proc. Info 0  @ fffffa8006688240
===============================================================================
Section 2     : x86/x64 MCA
-------------------------------------------------------------------------------
Descriptor    @ fffffa8006688138
Section       @ fffffa80066882c0
Offset        : 664
Length        : 264
Flags         : 0x00000000
Severity      : Fatal
Error         : DCACHEL2_DRD_ERR (Proc 1 Bank 6)
  Status      : 0xbe20000000011136
  Address     : 0x0000000137c03180
  Misc.       : 0x00000070c1000086

 

[carwiz]

Just as an update. The other two dumps are also L2 Cache Read errors. The first is a data read and the second is an instruction read from L2 Cache.

 

[Mike7143]

So, in theory, according to the dump, this is a CPU failure? But perhaps it's caused because this filter capacitor is missing? Does it make any sense that I could run a CPU burn-in test like calculating Mersenne primes from a live CD still without L2 cache? I did this the other day for 10 minutes, fan was blowing much hotter air out than when I'm getting the BSODs.

 

[Mike7143]

Also, sorry, but when you say a "hole", do you mean as in a burn through because it is high power? Or do you mean it could have been placed near a purpose hole or pass-through for the MB?

 

[carwiz]

LOL... No, it's a figure of speech. I mean the "hole" would be a location on the MB for the capacitor that isn't there. Its location might be ink screened on the board but not necessarily. I would just look for empty locations near other power components.

As far as L2 cache, I can only tell you what the dump says. I would imagine the CPU test with primes is using mostly the Numeric Processor and little data is moving in or out of cache. It may be one bad area that only gets used occasionally or a voltage spike that getting through because of the missing capacitor.

 

[Mike7143]

Hmm. Okay. Well I literally just combed over both entire sides. I measured the length of the "box" the component is to fit in with digital calipers and then went looking for empty boxes that were appropriately sized like the other 3 I can find on the board (all 3 are same size) and found none... I will say though that the 3 I can find on the board, 2 are on one side and 1 on the other. All 3 are near/around the power port side of the board, but I wouldn't exactly say they are "near" it.

Clicking should make them bigger...

This is the top of the board, here you can see 2 of the 3, near the HDMI out and VGA out:

And here is the bottom side with the 3rd one, near the VGA output:

 

[Britton30]

I thought you meant that Ken, but didn't want to speak for you.

Can it be the missing cap is taking power from the L2 cache? I think the parts of a CPU are allocated their own power?

 

[Mike7143]

And like I said, as far as I can tell, none of the empty boxes are the right/same length as the existing P15 component boxes. And none of the pads look like they've had something soldered to them. You'd think there might be a sharp or jagged edge where the solder broke, but running my fingers over all the empties doesn't yield anything. Then again, if the notebook got hot enough maybe it melted off?

 

[carwiz]

I scanned the pictures but didn't see anything obvious but I can't see the whole board. Those components are applied with solder paste then the board goes through an oven to melt the solder paste and affix the component. It's not likely the board left the factory with a spare piece.

If it is a filter capacitor, the circuit may still work but would carry transients (noise) or spikes in voltage. Either could affect the CPU and memory. Check around the memory modules too. The flow of incoming data (to the CPU) starts at the memory modules then goes to L3, L2, L1 then the CPU. "Data" being information and/or code instructions.

 

[carwiz]

I thought you meant that Ken, but didn't want to speak for you.

Can it be the missing cap is taking power from the L2 cache? I think the parts of a CPU are allocated their own power?

You're right about that! Nearly half of the CPU pins/balls are for power to various circuits. :shock:

 

[Mike7143]

I'll have a look around the memory more... I'll post these just in case too:
Top

Bottom

 

[Mike7143]

Ugh.. Everything that's empty looks like it has nicely melted pillows of solder paste. I can't find a single one that looks like a "cold solder" or looks like it may have been flattened in the past.... This is really frustrating. Maybe I should check the little tiny power button board again just for s&g.

And I know like you said, there doesn't necessarily have to be a box around where it was... but every other tiny little part has one, as well as the other 3 of these things, and they're all the same size border too!

 

[Britton30]

You are indeed persistent, like I would be with something like this. I would suggest getting a fairly decent magnifying glass and go over it very slowly and methodically. You will be able to see what normal solder points look like better and could recognize a separated one easier.

I was testing a new board once and saw a puff of smoke from it. I looked with my naked eye and saw nothing. Using a 8X jeweler's loupe I found the problem. A copper trace on the board had burnt away and burnt off the varnish coating. The burn mark was only about 1mm in length. Looking further I found several burned off places much bigger. It was a used board I was trying to make a useable system with so not much was lost.

 

[Mike7143]

Hold the presses!!

So this is a little aux board, it's got 2 USB ports and headphone/mic jacks. Notice how one USB port has been pushed in, broken?

Look at the back side:

I spy a rectangular outline... and some other broken stuff... This thing connects to the MB with a flat ribbon cable. Any chance that somehow this would cause problems? I mean, USB does handle power...
And look, another P15 chip!

So... what do we think?

 

[Mike7143]

Or, do you think I could put the laptop back together without this board and cable and it'd work?

 

[Britton30]

Are these two things broken too?

View attachment 299624

You could try running it without this sisterboard and see what shakes. You can't break it I guess.

 

[Mike7143]

Yes, those two things are also broken. I've since taken them completely off the board. Unfortunate I already took the HSF and all off. I'll have to get some thermal paste. I'm wondering if it'd do any damage to try without it. That board can be had for about $10 shipped or less. Might be worth replacing it first before I possible mess something else up? I guess I'm just not sure if the system would operate without the sister board. Surely it shouldn't fry anything... unless the MB is using that board for a circuit, but that just seems like a stupid design...

 

[Mike7143]

I guess the real question though... would those damaged chips on the sister board cause all the problems? It doesn't seem like it'd have anything to do with it, but then again, maybe all it takes is a short or something somewhere to maybe introduce noise or spikes elsewhere...

 

[Britton30]

Never run a CPU without paste! It can heat up much more and fast without it.

Well, I don't know how boards are "wired" up but there are so many components there and who knows what depends on another? Those broken part could have shorted something else out, or perhaps just removing the broken part would let it sort of work, it may not have any USB functions. Lotsa maybes to consider and I have no answers to them.

 

[carwiz]

Hey, great detective work! Looks like you found it.

That board looks like it carries the audio too. It might run without it but I wouldn't bet on it. The two USB ports look like they are independent so if you can isolate the broken one, it may work. The critical part is to make sure the port connections and traces aren't shorting. There's 5v at the port and the signal goes straight to the controller (Hub). The hub is probably part of the SouthBridge.

I would push that insulator back through the port and put a piece of tape over the opening. Then move the lose pieces so the contacts aren't shorting. Check the board for torn traces that might short too. Especially at the large vertical pins at the back of the port. (See the one next to it.)