RE: HP Pavilion DV6-2012AX not posting

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
Hi all..
Just discovered this section of internet, News Groups.
Need help in the repair of a HP DV6-2012AX notebook.
Does not post. All lights on but no initial bost screen.
Anyone know what or where I could find any information to start trouble  
shooting the motherboard?
This series is designed and build by Quanta computers in taiwan.

any help will be greatly appreciated.

Re: HP Pavilion DV6-2012AX not posting

On 2/29/2012 11:52 PM, awni wrote:
Quoted text here. Click to load it

Troubleshooting laptops is problematic because so much of it is
integrated.  I would try the obvious stuff first:

1) Disconnect any extraneous devices--USB mouse, external hard drive, etc.

2) Remove the battery.

3) Hook up a known good monitor to the video out of the laptop.  I
suggest using the VGA out (over HDMI) if you have it.

4) Remove memory modules.  Replace with known good modules, in spec, if
you have them.  If not, put in _one_ of your bits of memory.

5) Can you get POST, or any video?

    a) Yes.  If you get video on the monitor, but on the
       screen, you probably have a bad LCD panel, inverter
       or possibly cabling.  If you're getting some other
       sort of error message, or strange behavior, come
       back here and describe what's happening.

    b) No.  If you have extra bits of memory left over from
       #4, swap that in and got to step #5.  If you don't,
       then you probably have a crap motherboard.  Off to
       eBay to find a replacement.

Re: HP Pavilion DV6-2012AX not posting

On 1/03/2012 6:02 PM, Grinder wrote:
Quoted text here. Click to load it
Thank you Grinder,
Much appreciate the suggestion, I Have already done all this.
The reason I'm here is to find a place where I can get more in depth
notbook motherboard troubleshooting, I'm quite confident with the repair
of desktop motherboards and graphics cards, surface mount repair. Most
common problem is under rated electrolytic capacitors.
Can someone please inform what are the main problems with AMD based
notebook motherboards?????
Or al least point me in the direction of a forum, a news feed, a usent
group, etc.
I do have the schematics for this motherboard.
Anyone, please???????

Re: HP Pavilion DV6-2012AX not posting

awni wrote:
Quoted text here. Click to load it

The only thing I might add to Grinders suggestions, would be to try
a PCI port 80 POST debugger card. That will show you whether the processor
is succeeding at fetching instructions or not. When the processor comes out
of RESET (signal is de-asserted), the processor goes to a known address,
which usually maps to the Southbridge spigot with BIOS flash chip connected.
And then code is fetched from there, to initialize the rest of the chipset.

They make PCI port 80 POST cards for PCI and ISA. But when I search on the
web, I also see mini-PCI and mini-Express versions as well. I have no idea,
whether the BIOS writes to all these things during POST, and what restrictions
there might be on which slots get cycles to port 80. Just because some
guy in China, makes a different form factor port 80 card, doesn't mean
any computers actually drive it...

If the laptop had a beeper or piezo, that could also be used to
get some idea whether it was totally dead or not. Pull all the
SODIMMs, and see if it beeps a pattern for "memory error". That
would tell you the processor is able to fetch code from the BIOS chip.


Your debug sequence might be:

1) Examine power block. Remove battery and run off AC. Do the symptom change ?
    (The idea being, to prevent a battery problem from affecting the power
    Measure all major rails. Are any rails flat ? Check for burned switching
    conversion components. The power conversion could make +12V (panel), +5V
    +3.3V (ExpressCard slot or logic chips), 1.8V, 1.5V or lower for DRAM or
    logic. You'd want to check all of those, see whether they have status
    or whether any power bugs are used to detect valid voltage levels or not.
    There is a whole tree of logic feeding into RESET. And the status of the
    RESET signal, tells you a lot about the health of a part of the motherboard.

2) If the system comes out of reset, but is still crazy, then a lot of
    things could be the root cause. Noting how long instruction fetch
    lasts, before the processor goes into a double-bus-fault, might give
    you some idea how far it got. (Back when one of the memory manufacturers
    made DIMMs with activity LEDs on them, you could use the flashing of the
    LEDs as a status indicator. If the LEDs stopped flashing, then you'd know
    the processor was halted or double-bus-faulted.) If you have the schematic,
    you should also be able to check whether there are any run/halt signals
    on the side of the processor or not. Without a CPU data sheet, the signal
    might not be very descriptive of the function. (For example, on one chipset,
    there was a signal called "GHI" and I have no idea what that might be
    derived from.) I mean, try and find an AMD processor datasheet these
    days - good luck with that.

3) For anything more complicated than that, it's "pod and logic analyzer time".
    If the processor has a pin grid array socket, then there is a slim chance
    a company like Agilent makes a $2000 adapter pod for the socket, where a
    $50,000 logic analyzer can be plugged in. We had a couple for usage with
    Motorola 68K processors, way back when. And we could trace execution with
    that. With modern processors with the memory controller integrated into
    the processor, the FSB is no longer exposed so you can watch instruction
    fetches, so you'd have to go looking for whatever debug interface is
    included on the processor (it's probably a "thinner" interface than an
    FSB, and might have single stepping capability, breakpoints or the like,
    so that the equipment connection isn't so fat and onerous - the thinner
    interface also means the CPU operation is "less observable").

The odds of anyone having the gear for step 3 these days, are pretty slim.

With the schematic, you should at least be able to look for the
reset signal pin that feeds the processor, and work your way back
if it is stuck asserted. That's better than nothing. Or work your
way forward from the power converters, through whatever is used for
Power_Good, until you get to the reset signal. To monitor instruction
fetches, is going to be pretty difficult to do, especially if there
is no FSB. You might have to go looking for CS# on an SODIMM and watch
the memory instead. The levels there would be SSTL, and you'd probably
need a high speed scope to look at it. I doubt your garden variety
logic probe could handle it (if they even make logic probes any more).


As for "finding a forum filled with juicy debug info", you're
forgetting the nature of repair. If I know something, and I make
money because of that knowledge, I'm more likely to "keep it
close to the chest". I've noticed that trend, about hard
drive data recovery people around the world. These are people who
change out controller boards, motors, head assemblies, in order
to recover data from hard drives. And they're loathe to give
away information that another of their (competitor) repair
people could use. So if the repair industry is at all
competitive, don't expect any "free lessons". That's an example
of an industry, where part of the knowledge is obtained by
experiment, and takes time. So if I spent a week learning how
to repair an XYZ123 disk drive, and you want to know how I
did it, you'd going to have to trade me for the info, and tell
me something I need to know.


Re: HP Pavilion DV6-2012AX not posting

On Thursday, March 1, 2012 4:52:57 PM UTC+11, awni wrote:
Quoted text here. Click to load it

yes, it is a solder joint problem, by accident I turned on the motherboard =
without the heatsink and fan assembly, within about 10 seconds the motherbo=
ard booted up. I then re-installed the heatsink and fan back on, while stil=
l warm, the motherboard functioned perfect. Turned it off and let it cool d=
own to room temp, then tried to power up, same fault, all lights on and no =
post. after further inspection and testing found the Graphics IC to be at f=
ault, needs reball and reflow. Same problem as all the damn ATI Graphics ca=
rds with large heatsinks.=20
Thank you all that have helped, your support is greatly appreciated.=20

Re: HP Pavilion DV6-2012AX not posting wrote:
Quoted text here. Click to load it
without the heatsink and fan assembly, within about 10 seconds the motherboard
booted up. I then re-installed the heatsink and fan back on, while still warm,
the motherboard functioned perfect. Turned it off and let it cool down to room
temp, then tried to power up, same fault, all lights on and no post. after
further inspection and testing found the Graphics IC to be at fault, needs
reball and reflow. Same problem as all the damn ATI Graphics cards with large
Quoted text here. Click to load it

Weird. There were a bunch of Nvidia GPUs with unreliable
solder joints (needed "toaster oven fix"). I hadn't heard
of any ATI ones going like that (i.e. not bad enough to
require a recall or class action suit).

When a GPU fails, if the motherboard has a beeper, the BIOS code should
sound the beeper. A GPU fault, shouldn't prevent BIOS POST code
from running. (At least, for the architecture shown in the following
picture, with a discrete GPU.) The BIOS POST code can probe things,
and if it detects a problem, start beeping. I don't know if my
laptop has a beeper or not (never had it apart for a look).
In the case of your system, the GPU is on a PCI Express bus
connection, and so wouldn't affect the ability to fetch
BIOS code. I think it's something like this.

      M770---Mobile HD4650
      (Flash chip, BIOS code)

It's possible for things like DIMMs or SODIMMs to be defective
enough, to prevent the CPU in that picture from starting (electrically
destabilize the CPU). DIMM faults with no external symptoms
(not drawing excessive current from memory controller interface),
may allow the CPU to issue a beep code.

A few laptops have diagnostic LEDs, and there is a four LED code
that is displayed during a fault. Some computers even have dual
color LEDs in those positions, which increases the number of
possible codes which can be displayed. But that's not a
very frequent design feature. It gives about as much
information, as a separate port 80 POST card.


Site Timeline