Vaio notebook intermittent power problems, help
Xs
* Long post *
Hello, folks,
I'm currently attempting to troubleshoot a rather annoying power
problem with an old vaio notebook, pcg-f190. Turns out that this
is not an isolated problem, and there have been reports
of other vaios developing the same symptoms in 1 to 2 years.
See topics such as:
erroneous shutdown
turns itself off
intermittent shutdown
mysterious crash
The symptoms of this failure mode are, as previously
posted sample shows:
> From: no...@ieee.org (no...@ieee.org)
> Subject: Sony VAIO Sudden Erroneous Shutdown
> Newsgroups: comp.sys.laptops
> Date: 2000-12-14 19:00:08 PST
>
>
> My Sony VAIO F290 laptop has developed a habit for randomly shutting
> down. The problem has worsened with time; now my computer shuts off
> within seconds of booting to windows 98.
>
> Basically what happens is I boot to Win98 (the version distributed with
> the machine) and within seconds the entire machine shuts off. That is,
> with the exception of the green power light, which remains
> illuminated. (You can actually audibly hear the computer shut off.)
> To shut off the green power light, thereby completely shutting off the
> machine, I have to perform a hard/"forced" shutdown by holding the
> on/off slider to the "on" position for the requisite 4 seconds.
>
> I have spent an ENORMOUS amount of time researching the problem,
> including several hours on line (dejanews, cnet, sony page, etc),
> general hardware/software debugging, and even contacting the Sony help
> line (which--big surprise--was no help (after spending ~30 min on
> hold)). From my extensive research I have found:
> - The problem only occurs in full Win98, not in Safe mode or DOS
> - Going off memory, the problem seemed to be progressive: it got worse
> with time
> - Reference to forms of this problem show up sporatically in the news
> postings, but no consistent solution is suggested
> - Sony web page and phone service make no mention of the problem and
> have no apparent knowledge of it (actually, I could probably accurately
> omit the "of it" part of that statement)
>
> If you know anything about this problem, preferably a solution, please
> reply.
>
> Thanks in advance for your time and your help!
>
> Mike
>
I can also add some additional symptoms. Sometimes, attempts to power on
only result in a green light (and maybe the fan). The machine is essentially
dead, and a 4 second forced "off" is necessary before trying again. Virus
have been ruled out, as the symptoms don't go away after a clean reload
from recovery CDs. When the machine spontaneously dies, the fan is still
operational.
The best "official" solution to date has been to replace the motherboard.
That's not an acceptable option considering the board, from sony, is
over $1000.
Rather than just junk the notebook, I'd figure I try and disassemble it,
remove the motherboard, power it up on a workbench, and try to find the
failure root cause. This is pretty tough to do without a schematic.
But I'd figure it might be of interest for educational purposes.
Here's the results so far.
The machine spontaneously dies because critical subsystems, such as
the HDD disk drive, lose +5v power. When this happens, the actual +5v
supply is still good and active. Tracing the power path results in
IDing an electronic switch (466 MOSFET) that's shutting off the HDD.
Well, the switch control lead says the MOSFET should be off. So
there's nothing wrong with the switch.
So now, I have to trace back the erroneous control signal. The MOSFET
is controlled through the output of a single chip NAND gate. One input,
in the erroneous state is 0v. It should be 3.3v. Tracing further back
seems to show that the erroneous output is from a VHC273 D-type flip-flop
with clear. This output drives the NAND gate noted earlier, and
some other logic, including a MOSFET gate, and input for a CU345
(SN74CBT3345) Bus Switch (purpose unknown).
I'm uncertain what drives the VHC273. My best guess is it's connected
to the PCIset 82371 (PIIX4) with all it's power management features.
This is a BGA device, so no leads are accessible for testing.
That's as far as I can test. I don't have access to a digital oscilloscope.
It seems that the failure is somehow related to ACPI initialization or
setup operations. Is it possible that the VHC273 operation has degraded
over time? Can something else, via power management control, erroneously
trigger a premature power shutdown of the HDD, and other critical components?
Why is it that sometimes, the initial machine power up does almost nothing?
Has anybody else encountered this problem, and knows what the root
cause is? Any further suggestions on debugging welcome.
Thanks for listening,
Xs
jakdedert
any Power Management features...then go "Start--Accessories--System
Tools--System Information--Tools (in the menu bar)--System Configuration
Utility" on the "Startup" tab, uncheck "Load Power Profile." (I just
noticed my machine has *two* such entries...disabled one.) Possibly Sony
has some other name for it, but that's the default Win 98 entry.
I believe that your problem is not in the hardware, but in software. You've
just trace down the mechanism by which the O.S. performs the task of getting
into standby...maybe not, but it's worth a try. The tip-off is that it
doesn't happen in Safe Mode or DOS.
jak
"Xs" <xs...@comcast.net.removethis> wrote in message
news:3C7B0090...@comcast.net.removethis...
Rawman
together) seems to work except all the plug and play devices are gone
unrecognised.
Tried Jak suggestion with no success. I will probably check the VHC273
inputs with a scope when I have time.
Raw
Xs
jakdedert wrote:
>
> I'd suggest that you go into Safe Mode, open the Control Panel and disable
> any Power Management features...then go "Start--Accessories--System
> Tools--System Information--Tools (in the menu bar)--System Configuration
> Utility" on the "Startup" tab, uncheck "Load Power Profile." (I just
> noticed my machine has *two* such entries...disabled one.) Possibly Sony
> has some other name for it, but that's the default Win 98 entry.
>
> I believe that your problem is not in the hardware, but in software. You've
> just trace down the mechanism by which the O.S. performs the task of getting
> into standby...maybe not, but it's worth a try. The tip-off is that it
> doesn't happen in Safe Mode or DOS.
>
Thanks for the suggestion, and I'll certainly give it a try. I'm also going
to try and disable all power management drivers. Though I think that only
addresses the symptoms, and not the actual problem. The power profile that's
active, according to the control panel, is one that shouldn't shutdown
anything.
The reason I think it is a hardware issue is that the notebook sometimes
doesn't start (does not even begin to boot) at all when powering up. Only
the power light comes on. If it does power up and boot, the failure behavior
is not consistent. It may stay up for seconds, minutes, and maybe not fail
at all after several attempts. The problem also seemed to developed gradually
over time.
Xs
Rawman
You are right, the problem is definitely in the hardware. Mine fail slowly
over time and now, it is a permanent problem.
Let me know if you have more success with the hardware troubleshooting or a
work around.
Good luck, Raw
Xs
Some additional progress.. , still need some assistance..
On the sony vaio pcg-f190 motherboard, I was able to check some of the
VHC273 (near the lithium battery) inputs. It appears that this octal
D flip-flop is some sort of power control/switch register.
When the spontaneous power failure occurs, the master reset input
lead (pin 1) on the VHC273 drops 3.3v to 0, and stays at 0, clearing
this register. The other VHC273 input leads, like Vcc, and the Q0 input
I checked are ok. This rules out the VHC273 to be the problem source.
I guess that the master reset is some sort of emergency power cut
signal. I've tried to locate what is sourcing the master reset, without
much success. The only other PCB trace I found, so far, that may have
connectivity is JR214 near the lower left corner (from top view) of the
neomagic chip.
Xs
Lou K.
"Rawman" <raw...@yahoo.com> wrote in message
news:uHQf8.72735$Ah1.9...@news2-win.server.ntlworld.com...
I was having the same problem but had (incorrectly) assumed that it was
related to the broken power input I had.
Sent it in for repair and upon return there was a mention of them fixing the
problem, but no mention of what the fix was.
$294.
--
Lou K.
exit 7
remove hat to reply
Rawman
signal come from.
Is the VHC273 chip sitting on the main board or the power supply daughter
board?
The master reset must be issued by the ACPI controlled device. When I
disable this the problem is gone but I cannot automatically shut off after
window shut down. This VIAO series could be using Sony own controller chip.
What happen if the master reset is tied high via resistor permanently, we
may get around the problem without replacing the faulty chip.
Raw
Xs
Rawman wrote:
>
> I am gonna open my viao tomorrow and I will trace where the master reset
> signal come from.
>
> Is the VHC273 chip sitting on the main board or the power supply daughter
> board?
On the motherboard I'm working on, the VHC273 (74VHC273), a 20 pin chip,
is on the main board partially covered by the insulating mask around the
lithium coin battery.
Try checking the action of pin 1 (clear, master reset for the D-FF) when power
fails after boot. The HDD power bit is pin 2 (Q0) and pin 3 (D0). So far, not
much luck on my part locating the source of the flip/flop clear signal.
> The master reset must be issued by the ACPI controlled device. When I
> disable this the problem is gone but I cannot automatically shut off after
> window shut down. This VIAO series could be using Sony own controller chip.
Looks like an Intel PCISet FW82371EB (PIIX4). The datasheets are available
at the Intel site.
> What happen if the master reset is tied high via resistor permanently, we
> may get around the problem without replacing the faulty chip.
I will be trying this at some later point, assuming I/we give up locating the
problem's root cause. I'm sorta wondering what functions would "break"
with this workaround.
There is a thermal alert signal as well, sourced from a ADM1021A located
on the processor board. The alert signal is at B86 on the connector
between the processor and main board. This signal does appear to be routed
directly to the PIIX4. It's possible that the ACPI drivers might recognize an
alert during initialization, so I'm going to check for false alerts, later.
However, previous reports about a year ago from someone who tied this signal
high suggest that this wasn't the problem.
Let me know what you find out.
Xs
Rawman
>
> Looks like an Intel PCISet FW82371EB (PIIX4). The datasheets are available
> at the Intel site.
>
batch of viao and this chip has been used extensively elsewhere.
>
> I will be trying this at some later point, assuming I/we give up locating
the
> problem's root cause. I'm sorta wondering what functions would "break"
> with this workaround.
i think we can still get most of the function except the power pff after
wibndow shutdown.
I disable the acpi and still able to get the screen to go off correctly when
inactive.
>
>
> There is a thermal alert signal as well, sourced from a ADM1021A located
> on the processor board. The alert signal is at B86 on the connector
> between the processor and main board. This signal does appear to be routed
> directly to the PIIX4. It's possible that the ACPI drivers might recognize
an
> alert during initialization, so I'm going to check for false alerts,
later.
> However, previous reports about a year ago from someone who tied this
signal
> high suggest that this wasn't the problem.
>
This is interesting since with my viao the fan started to come on few second
after the spontaneous shutdown occur.
Will investigate this evening. will be in touch if i find anything
interesting.
Raw
H. Dziardziel
wrote:
>
>
snip
>
>Try checking the action of pin 1 (clear, master reset for the D-FF) when power
>fails after boot. The HDD power bit is pin 2 (Q0) and pin 3 (D0). So far, not
>much luck on my part locating the source of the flip/flop clear signal.
>
>
Hello, do you have a schematic?
Regards
>
snip
>
Rawman
whole thing is recessed and covered by the screen hingge.
WIll try again at the weekend, does anybody have any suggestion on how to
remove the CPU fan?
Xs
Rawman wrote:
>
> no schematic i am afraid. I could not lift the fan on my viao F360. The
> whole thing is recessed and covered by the screen hingge.
> WIll try again at the weekend, does anybody have any suggestion on how to
> remove the CPU fan?
If you have a copy, the sony Vaio service manual has info on disassembly.
However, only block diagrams for the electronics. No schematic.
Regarding the screen, the screen hinge covers pop out (they are snap in).
The screen connector, with screws removed, must be disconnected. Unscrew all
the hinge mounting hardware on the chassis, and separate the screen with
hinges, from the chassis. That's part of what's necessary to gain access to
the CPU fan.
You practically have to completely disassemble the chassis and remove the
motherboard for the type of low level debugging described earlier.
Almost every accessible, and hidden screws, must be removed to free
the motherboard. It's an effort to do. I had trouble locating all
the screws, especially the one under the docking bay connector cover.
Note, that connector is part of the motherboard.
I'm planning to check for false thermal alerts sometime this weekend.
Your Vaio fails exactly as mine. When the shutdown power cut happens,
the fan starts up a few seconds later.
Best of (continuing) luck,
Xs
Xs
Xs wrote:
>
> There is a thermal alert signal as well, sourced from a ADM1021A located
> on the processor board. The alert signal is at B86 on the connector
No luck with this approach. The alert appears unchanged and remains at logic 1,
3.3v, before and after the spontaneous power failure.
So it's back to trying to locate the source for the vhc273 D-FF clear signal.
Rawman
came in with the CD but my quick search did not reveal it.
have not got much time at the moment, will try to locate the rest signal
source this weekend and hopefully before.
cheers,R
"Xs" <xs...@comcast.net.removethis> wrote in message
news:3C8BA0B7...@comcast.net.removethis...
Rawman
cheers, Rawi
"Rawman" <raw...@yahoo.com> wrote in message
news:Xzaj8.28905$yc2.3...@news2-win.server.ntlworld.com...
Xs
Xs wrote:
>
> So it's back to trying to locate the source for the vhc273 D-FF clear signal.
Here's additional progress on locating the vaio spontaneous power
fail problem.
The vhc273 D-FF clear is also tied to JR214, near the neomagic chip,
JR301, near the ricoh controller, and the output from a single chip
OR gate, IC464 near the fdc37n958fr notebook i/o controller.
The D-FF clear signal source is found. The purpose of this signal appears
to be a form of "global reset" which also cuts power to peripherals.
The OR gate inputs are the i/o controller gpio0 pin 184, and JR408 near
the intel 82371 (piix4). Signal at JR408 is at fault during system power
failure. Best guess is this is PCIRST at the 82371. Looks like the possibility
of false/bad resets. What is the cause?
So the search continues.
Xs
This note is for the benefit of others that have, or may want to
continue to investigate Vaios with this (or similar) faults.
Problem is false system reset assert from the fdc37n958fr notebook
i/o controller (nRESET_OUT pin 106). Signal gets stuck in reset state
upon failure.
Sorry, I wasn't able to find the root cause.