Dave Plowman (News) wrote:
In article ,
Rob wrote:
Just disable suspect software, and/or try to replicate the fault in safe
mode (F7 on boot IIRC).

Happens in safe mode too. Should have mentioned that.

PCs are basically a few components - you have to
work through and isolate. My guess from what you say would be MB on a
new PC - the trickiest component to test alas.

I've fairly extensive experience of electronic faults but this one doesn't
follow the norm.

My Mac has been doing the same. POS.

Anyway, yesterday it screwed the whole system disk to the point it
couldn't even verify t.

New disk went in, but ..oh dear..better, but still not fixed.

I am ****ing around pulling out one ram chip after another to see if one
removal makes it happy.

So far..the smallest one is out..and it hasn't crashed..

Look, everything goes on the bus in a computer. CPU, peripherals, RAM
chips- the lot. One of those doing silly buggers can cause random memory
corruption. Sometimes one thing you plug in works fine. And another, but
two together..no way.

Then it escalates from there. You gate a couple of bytes corrupted.No
big deal. Then maybe that bit gets swapped out, or swapped into another
bit of Ram. so thats two bits of corruption..and on it goes, and
eventually the OS itself tries to access some memory it shouldn't, and
it's showtime. Or you send an impossible command to the disk controller,
and the disk ****s up. A call to memory location zero, will reboot the
machine. Or is it 0000FFFF? I forget.


All you can do is remove as much as possible and try for stability, and
then put a piece back at a time..

Anyway, sorry for rambling on. I am waiting for this ******* machine to
crash. So far it hasn't. Oh, could it be the 64Mbyte chip? I only put it
in because it was the last spare chip and there was a slot available...

In message , Clint Sharp
writes
In message , geoff
writes
It was, I've repaired several Mobos with this problem, It's not a
problem with 12 month old Asus boards
Ah, but then you made a rather sweeping statement in the first instance
and didn't specify Asus only.

I was specifically referring to his problem - some one "bet" that was
what HIS problem was, I bet that it wasn't

I see probably 30-40 systems a month for repair and I can say it's
definitely still a problem that affects quite a few big name box
manufacturers. I usually only get to see 'clone' stuff when family or
friends phone me so I couldn't comment on Asus specifically.

London Pride, if you please ...
Tell you what, I'll buy mine, you buy yours and we can share a pub...

OK - just as soon as I've finished my diet (a month so far since I had a
beer, almost a stone lost ! ... five to go)

--
geoff

On Thu, 07 Feb 2008 02:38:01 +0000, The Natural Philosopher wrote:

Dave Plowman (News) wrote:
In article ,
Rob wrote:
Just disable suspect software, and/or try to replicate the fault in
safe mode (F7 on boot IIRC).

Happens in safe mode too. Should have mentioned that.

PCs are basically a few components - you have to work through and
isolate. My guess from what you say would be MB on a new PC - the
trickiest component to test alas.

I've fairly extensive experience of electronic faults but this one
doesn't follow the norm.

My Mac has been doing the same. POS.

Anyway, yesterday it screwed the whole system disk to the point it
couldn't even verify t.

New disk went in, but ..oh dear..better, but still not fixed.

I am ****ing around pulling out one ram chip after another to see if one
removal makes it happy.

So far..the smallest one is out..and it hasn't crashed..

Look, everything goes on the bus in a computer. CPU, peripherals, RAM
chips- the lot. One of those doing silly buggers can cause random memory
corruption. Sometimes one thing you plug in works fine. And another, but
two together..no way.

Then it escalates from there. You gate a couple of bytes corrupted.No
big deal. Then maybe that bit gets swapped out, or swapped into another
bit of Ram. so thats two bits of corruption..and on it goes, and
eventually the OS itself tries to access some memory it shouldn't, and
it's showtime. Or you send an impossible command to the disk controller,
and the disk ****s up. A call to memory location zero, will reboot the
machine. Or is it 0000FFFF? I forget.


All you can do is remove as much as possible and try for stability, and
then put a piece back at a time..

Anyway, sorry for rambling on. I am waiting for this ******* machine to
crash. So far it hasn't. Oh, could it be the 64Mbyte chip? I only put it
in because it was the last spare chip and there was a slot available...

You can also try booting a standalone OS such as Unbuntu or RedHat, and
see if the system holds up. Since the standalone systems are very memory
intensive it could show up the problem.

--
Ed Sirett - Property maintainer and registered gas fitter.
The FAQ for uk.diy is at http://www.diyfaq.org.uk
Gas fitting FAQ http://www.makewrite.demon.co.uk/GasFitting.html
Sealed CH FAQ http://www.makewrite.demon.co.uk/SealedCH.html
Choosing a Boiler FAQ http://www.makewrite.demon.co.uk/BoilerChoice.html

In message , writes
On 7 Feb,
geoff wrote:

OK - just as soon as I've finished my diet (a month so far since I had a
beer, almost a stone lost ! ... five to go)

I've cut back (still enjoy a pint!) after discovering I'd put on two stone in
the last two years, most, I think over last christmas. I'm hardly noticing
the diet, and have already (in a fortnight) lost a quarter of my target.

Eighteen 'kin stone

That's what comes of spending evenings in UK.d-i-y with half a dozen
cans / night

"The road is Loooong, ..."

Summer will be here before I have another pint

--
geoff

In article ,
geoff wrote:
In message , "Dave Plowman (News)"
writes
In article , Rob
wrote:
Just disable suspect software, and/or try to replicate the fault in
safe mode (F7 on boot IIRC).

Happens in safe mode too. Should have mentioned that.

PCs are basically a few components - you have to work through and
isolate. My guess from what you say would be MB on a new PC - the
trickiest component to test alas.

I've fairly extensive experience of electronic faults but this one
doesn't follow the norm.

pop over here with it, we can fart around swapping bits over, chuck
another HD in etc, get my free pint(s) by eliminating the capacitors as
being the problem,

Thanks for the offer. I haven't yet eliminated the simple things like
removing all non essential hardware - of which there's a lot. But if I get
stuck I'll hold you to that. ;-)

a process of elimination, which is much easier when you have everything
to hand

Yup. Of course it's been rock solid since I first posted. Left it on for
24 hours too.

--
*If you're not part of the solution, you're part of the precipitate *

Dave Plowman London SW
To e-mail, change noise into sound.

Dave Plowman (News) wrote:

I've got a home assembled PC based around an Asus motherboard. Which shuts
off at random. Not a controlled shut down - it just dies. Thinking PS
problems I've replaced it. Same thing. Sometimes it will boot up ok and
continue for as long as I want. Sometimes it switches off half way through
the boot sequence - but never at the same time. Any guesses? I've got AVG
free installed and that shuts it down after doing its tests. That sequence
always works ok.

This is slightly less usual than one that reboots at random. The dodgy
cap problem I have seen many times - but this is often a random reboot
cause.

Overheating processor is often a reboot or more likely a lockup or blue
screen.

Dodgy reset circuitry is a possibility - but only slim.

I would suggest downloading a copy of the "ultimate boot CD" iso image
and booting from that. Then leave memtest 86 running for a good few
hours and see what happens.

If that is ok then grab a copy of Prime95, and leave it running the
torture test - that is very good at shaking out subtle memory and
processor timing issues and faults.

Check your power saving options as well - make sure you have not got
some unusual trigger set up for powering the machine off.



--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

Eighteen 'kin stone

Ditto here - possibly even more now, but the missus decided she'd feel
better throwing the scales away !

Ed Sirett wrote:
On Thu, 07 Feb 2008 02:38:01 +0000, The Natural Philosopher wrote:

Dave Plowman (News) wrote:
In article ,
Rob wrote:
Just disable suspect software, and/or try to replicate the fault in
safe mode (F7 on boot IIRC).
Happens in safe mode too. Should have mentioned that.

PCs are basically a few components - you have to work through and
isolate. My guess from what you say would be MB on a new PC - the
trickiest component to test alas.
I've fairly extensive experience of electronic faults but this one
doesn't follow the norm.

My Mac has been doing the same. POS.

Anyway, yesterday it screwed the whole system disk to the point it
couldn't even verify t.

New disk went in, but ..oh dear..better, but still not fixed.

I am ****ing around pulling out one ram chip after another to see if one
removal makes it happy.

So far..the smallest one is out..and it hasn't crashed..

Look, everything goes on the bus in a computer. CPU, peripherals, RAM
chips- the lot. One of those doing silly buggers can cause random memory
corruption. Sometimes one thing you plug in works fine. And another, but
two together..no way.

Then it escalates from there. You gate a couple of bytes corrupted.No
big deal. Then maybe that bit gets swapped out, or swapped into another
bit of Ram. so thats two bits of corruption..and on it goes, and
eventually the OS itself tries to access some memory it shouldn't, and
it's showtime. Or you send an impossible command to the disk controller,
and the disk ****s up. A call to memory location zero, will reboot the
machine. Or is it 0000FFFF? I forget.


All you can do is remove as much as possible and try for stability, and
then put a piece back at a time..

Anyway, sorry for rambling on. I am waiting for this ******* machine to
crash. So far it hasn't. Oh, could it be the 64Mbyte chip? I only put it
in because it was the last spare chip and there was a slot available...

You can also try booting a standalone OS such as Unbuntu or RedHat, and
see if the system holds up. Since the standalone systems are very memory
intensive it could show up the problem.

Not in my case I cant. PowerPC chip. Not Iintel. Linux is possible, but
not from CD.

Anyway its been stable since I removed the third RAM chip, so anyone
with any 512Mbyte SDRAM non ECC 168pin PC100 or 133, that they dont
need....;-)

"Dave Plowman (News)" wrote in message
...
I've got a home assembled PC based around an Asus motherboard. Which shuts
off at random. Not a controlled shut down - it just dies. Thinking PS
problems I've replaced it. Same thing. Sometimes it will boot up ok and
continue for as long as I want. Sometimes it switches off half way through
the boot sequence - but never at the same time. Any guesses? I've got AVG
free installed and that shuts it down after doing its tests. That sequence
always works ok.

Something I've just found out about AVG Free Edition. My pc used to report a
Commit Charge (memory being used basically) in Task Manager of about 133mb
after a clean reboot and would go back down to that if I closed all open
applications. I noticed the other day that it was running at nearly twice
that but with nothing apparent to account for it in the Processes section of
Task Manager. I'd recently installed AVG after picking up a virus and that
turned out to be the culprit. Even with it shut down there was over 100mb of
something or other memory resident. I uninstalled it and things are back to
normal. In future I'll just install it for the occasional disk check and
then take it out again or try something else. Most viruses I pick up I can
deal with manually anyway by hunting them down and deleting the files and
registry entries. I actually quite enjoy that rather than letting a prog do
it. AVG didn't fully deal with what I'd picked up anyway. The virus had
created a couple of hidden directory entries starting with a ? and I had to
remove those in safe mode after AVG missed them.

My old machine only has 512mb installed so this extra 100mb was causing it
to max out a lot faster with other apps running and start using pagefile
which then drastically slows things down. Might be part of your problem too
if you have a dodgy memory stick which would be getting fully utilised
faster.
--
Dave Baker
Puma Race Engines

On Fri, 08 Feb 2008 11:02:16 +0000, The Natural Philosopher wrote:

Ed Sirett wrote:
On Thu, 07 Feb 2008 02:38:01 +0000, The Natural Philosopher wrote:

Dave Plowman (News) wrote:
In article ,
Rob wrote:
Just disable suspect software, and/or try to replicate the fault in
safe mode (F7 on boot IIRC).
Happens in safe mode too. Should have mentioned that.

PCs are basically a few components - you have to work through and
isolate. My guess from what you say would be MB on a new PC - the
trickiest component to test alas.
I've fairly extensive experience of electronic faults but this one
doesn't follow the norm.

My Mac has been doing the same. POS.

Anyway, yesterday it screwed the whole system disk to the point it
couldn't even verify t.

New disk went in, but ..oh dear..better, but still not fixed.

I am ****ing around pulling out one ram chip after another to see if
one removal makes it happy.

So far..the smallest one is out..and it hasn't crashed..

Look, everything goes on the bus in a computer. CPU, peripherals, RAM
chips- the lot. One of those doing silly buggers can cause random
memory corruption. Sometimes one thing you plug in works fine. And
another, but two together..no way.

Then it escalates from there. You gate a couple of bytes corrupted.No
big deal. Then maybe that bit gets swapped out, or swapped into
another bit of Ram. so thats two bits of corruption..and on it goes,
and eventually the OS itself tries to access some memory it shouldn't,
and it's showtime. Or you send an impossible command to the disk
controller, and the disk ****s up. A call to memory location zero,
will reboot the machine. Or is it 0000FFFF? I forget.


All you can do is remove as much as possible and try for stability,
and then put a piece back at a time..

Anyway, sorry for rambling on. I am waiting for this ******* machine
to crash. So far it hasn't. Oh, could it be the 64Mbyte chip? I only
put it in because it was the last spare chip and there was a slot
available...

You can also try booting a standalone OS such as Unbuntu or RedHat, and
see if the system holds up. Since the standalone systems are very
memory intensive it could show up the problem.

Not in my case I cant. PowerPC chip. Not Iintel. Linux is possible, but
not from CD.

Anyway its been stable since I removed the third RAM chip, so anyone
with any 512Mbyte SDRAM non ECC 168pin PC100 or 133, that they dont
need....;-)

I think there is a version of Ubuntu that runs on a PPC. It is bootable
and standalone.


--
Ed Sirett - Property maintainer and registered gas fitter.
The FAQ for uk.diy is at http://www.diyfaq.org.uk
Gas fitting FAQ http://www.makewrite.demon.co.uk/GasFitting.html
Sealed CH FAQ http://www.makewrite.demon.co.uk/SealedCH.html
Choosing a Boiler FAQ http://www.makewrite.demon.co.uk/BoilerChoice.html