On Fri, 9 Dec 2005 19:59:40 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
wrote:

> have an os/2 system with 3 os/2 operating system partitions, only one
>
> of which is currently working. The good system has gotten flakey
> lately
> so I have been trying to reinstall new 4.5.2 systems in the backup
> partitions but with no success. I probably have a hardware problem
> but have
> not been able to determine what it is.

Try memtest86 for 12 hours or so. There's also a program called something like
k7burn to test Athlon processors. Should check out that side of the hardware for
you.

> Upon insertion of the last disk, I get a message that neither my hard
> drive
> nor my floppy can be operated.

This really means "I can't find my own boot drive" so if booting from floppy,
check you have the right drivers. If booting from CD, check the drivers for that
etc.

> It is possible that my SCSI controller is having problems in that
> sometimes it does not see the second SCSI hard drive or the Plextor
> SCSI
> CDORM, but a reboot usually fixes that and they then work fine.

But this...

> However, I did have to change
> the
> transfer rate from 40 to 20 mbps.

....and this both seem to show that the majority of your problems are to do with
your SCSI bus. Maybe you should check all your cabling is snug and tight and not
creased or broken. If using multiple devices on the same controller then there
are some length restrictions on cabling - especially if mixing LVD and non-LVD
devices on the same bus. Might help to post how all your SCSI devices are
connected. Also, I'm not sure if the 390U2B/W controller has a switching device
that arbitrates between different sections of the bus and allows you to put LVD
devices on one cable and non-LVD devices on another.

--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com

Trevor - Thanks for your input.

"Trevor Hemsley" wrote:

>
> Try memtest86 for 12 hours or so. There's also a program called something like
k7burn to test Athlon processors.

I ran memtest86 for more than 12 hours with no errors. I have not yet
done a burn-in test but have run thousands of seti-at-home processing
units with this MB.

Some progress was made today however.

By REMing out all unneeded drivers in config.sys and unneeded .snp
calls in snoop.lst (I'll post those files if anyone is interested). I
did get past loading my SCSI driver and past the point of
reformatting the target partition and starting to load files from the
SCSI CDROM. However I get a SYS0039 mesage (L: device not ready)
wherre the L: device is my SCSI CD. Retry yields a few more files
read but then the file transfer eventually hangs.

This means that either the SCSI CDROM or the SCSI controller has a
problem, even though the CDROM appears to work fine when booting in my
good system partition. I have an older slower SCSI CDROM (somewhere
around the house) I can try and will do that. If the controller is
bad, I'll try to replace it with the same model so that I have a good
chance of reading my existing SCSI drrives.

I also tried to install using my IDE DVD/CD by disconnecting the SCSI
CD. This failed because OS2CDROM.DMD would not load. I don't know
why it would not load but I got a message that the line in config.sys
calling that driver was ignored. This was true whether I used the
DANI drivers (daniatapi.flt, danidasd.dmd, and danis506) instead of
the IBM equivalents or not. Any ideas on this one??

Thanks for your help

Nate Liskov

--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

I achieved a successful installation with my older SCSI CDROM. Thus
the current CDORM will be replaced.

However, I had to delete the statement

basedev=ibmint13.i13

to prevent a hang during the reboot that occurs in the middle of the
installation process and had to delete the statement

basedev=sym8xx.add

to prevent a hang during the first reboot after installation. Finally
there was a hang (trap 8) that occurred after the next boot, I think
due to some network requisition process that auto-started after the
next reboot. I stopped that process after the following reboot and
the installation seems stable.




"Nathan Liskov" wrote:

> However I get a SYS0039 mesage (L: device not ready)
> wherre the L: device is my SCSI CD. Retry yields a few more files
> read but then the file transfer eventually hangs.
>
> This means that either the SCSI CDROM or the SCSI controller has a
> problem, even though the CDROM appears to work fine when booting in my
> good system partition. I have an older slower SCSI CDROM (somewhere
> around the house) I can try and will do that. If the controller is
> bad, I'll try to replace it with the same model so that I have a good
> chance of reading my existing SCSI drrives.
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

In spite of the installation success that I reported, my system is
still unstable.

I used a simple diagnostic technique that I should have though of
earlier. I set the BIOS to run my processor at about half the normal
speed and the system became relatively stable.

The tech at my computer store thought that the processor or the MB
might be at fault. I then replaced the processor, but that did not do
it. Note that no problems showed up when I ran memtest86 or 386test
programs (at full speed), but then OS/2 itself is probably a more
severe memory/processor test than these programs.

The next step is to try a new MB. He has recommended the ASUS AV8
Deluxe with an athlon 64 at 3500 MHz.

Pretty much all the new boards are for Athlon 64 (other socket 939
boards they have are by Abit, Gigabyte, and MSI). (my current board
is an ASUS S7V8X-X, an athlon socket 460A board)

I am worried about either new installation using my PA warp 4.5.2 or
operation with the existing working 4.5.2 partitions with an athlon 64
MB. The last time I tried I had no success using my existing
installation, but I did not try a new installation.

Most postings I have seen with Athlon 64 and os/2 have indicated
complicated installation problems.

There is a recent thread on an AMD Athlon 64 X2 with success using ECS
1.2 with SMP, but I am not using ECS.

If there are any success stories with the AV8 Deluxe MB (or any other
socket 939 MB) with warp 4.5.2, I would like to hear them.

Thanks

Nate Liskov

ps My previous diagnosis of a bad SCSI CDROM reader is probably wrong.


"Nathan Liskov" wrote:

> I probably have a hardware problem
> but have
> not been able to determine what it is.
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

On Thu, 15 Dec 2005 07:20:45 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
wrote:

> My previous diagnosis of a bad SCSI CDROM reader is probably wrong.

It *still* sounds like a SCSI cabling/termination problem to me. Why not post
how you have everything cabled up and see if htat helps someone to help you?

--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com

OK.

My controller has two 68 pin SCSI connectors and one 50 pin
connector. I have recently reseated the controller and all
connectors with no effect on my problems.

On the 50 pin connect I nominally have a 620 mb Fujitsu MO drive and
a Plextor 12/4/32 CD writer/reader and a passive terminator on the
last connector. This cable is not an issue as problems occur with
this cable disconnected from the controller.

On one 68 pin connector I have a generic 68 pin cable with 5
connectors. The controller is on connector 1, drive 1 is on
connector 3, drive 0 on connector 4 and a passive terminator on
connector 5. This cable is just over one meter long (42 inches).
The other controller 68 pin connector is not used.

In my old system, I had a custom made premier teflon SCSI cable custom
made with 8 connectors. It is about 5 ft long and has an active
termination.

I replaced the generic cable with my teflon cable. On the first
reboot (at full speed) I got the following message:

Error reading disk 4. Press any key

Disk 4 is one of my IDE drives. After pressing a key I got a Boot
Manager screen with nonsense in the boot options (not my normal drive
E:, G: and H: options).

The next reboot was successful. However, I should note that it
takes almost 3 minutes for HPFS.IFS to load on every successful
reboot. I don't think this is normal.

The following reboot resulted in the disk 4 read error again followed
by a system stop with messages SYS01475 (OS/2 boot cannot be found)
and SYS02027 (insert a disk and restart the system).

The next reboot worked. So far it is working OK (no crashes yet), so
I'll monitor for stability.

I should also note that with the new cable I was successful in
setting the SCSI controller BIOS to enable both wide and synchronous
negotiation and now get 40 mbyte/sec transfers versus the 20 I was
getting with the original cable. So I guess a better cable is called
for anyway.

A little history: I have two IDE 32Gb Quantum Fireball LCT15 drives
in my system. They were "free" drives as they came from my TIVO DVR,
when I upgraded my TIVO with bigger hard drives. I did not use them
much and did not notice when they stopped being available in my
system. However I noticed when one day I started to get messages on
failure to read either disk 3 or disk 4 and complete failure to boot.
After disconnecting these drives the system booted, but that may be
when my stability problems started. Both drives are good and are
working in my system now.

Drive 1 is a new Seagate Cheeta (32 gB) that I just put in to replace
my noisy older IBM 16 gB drive. Drive 0 is an older IBM 16 gB drive,
which still seems to be OK.

It is possible the problems are in the SCSI controller (a Tekram
DC-390U2B/W). I have a new controller on order (the Tekram U3 model
that some users have reported success with in OS/2), which will
double the transfer speed for the new drive. Hopefully, the new
controller will be able to read the existing SCSI drives.

Since the controller may be the problem, I will wait until I put in
the new controller and see how things go before attempting to replace
the motherboard.

Thanks for any ideas.

Nate Liskov

"Trevor Hemsley" wrote:

> On Thu, 15 Dec 2005 07:20:45 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
> wrote:
>
> > My previous diagnosis of a bad SCSI CDROM reader is probably wrong.
>
> It *still* sounds like a SCSI cabling/termination problem to me. Why not post
> how you have everything cabled up and see if htat helps someone to help you?
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

A few more points I forgot in my recent long post:

I flashed the latest BIOS for my MB.

My old processor was a Athlon Barton 3200 and there were BIOS speed
choices from 2200 to 1100. The newer chip is a slower 2800 but the
four BIOS speed choices range from 2500 to 1250. I got nothing to
work at 2500 (why is it there) and consider full speed for this chip
at 2083, the second highest rate choice.

I have mucked with the BIOS setting for OS2 using more than 64 Mb.
Currently it is disabled. My understanding is that this setting was
relevant only for OS/2 version 2.11 and is a no-op for current Warp
versions. I assume that this had no effect when I did have it
enabled.





nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

On Thu, 15 Dec 2005 18:18:07 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
wrote:

> On one 68 pin connector I have a generic 68 pin cable with 5
> connectors. The controller is on connector 1, drive 1 is on
> connector 3, drive 0 on connector 4 and a passive terminator on
> connector 5.

If this is an LVD bus then it requires active termination. Did your problems
start shortly after upgrading to the Seagate Cheetah? That's most likely also an
LVD drive - maybe the controller is being confused and trying to negotiate
ultra2 speeds? I'd try to find an active terminator in any case - they're
recommended for UW buses and required for LVD. I'd also check the termination
settings within the Tekram BIOS and make sure that it's set right - I think the
390U2B does not have an isolator that separates each segment of the bus so you'd
either need it on automatic (which is sometimes unreliable anyway) or set to
terminate the upper 18 lines of the bus but not the lower 50 (since you have a
terminator on the 50 pin cable). Oh, and check the jumpers on all your devices
and make sure that none of them have termination enabled by mistake - too many
sources of termination is just as bad as too few!

--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com

On Thu, 15 Dec 2005 07:20:45 UTC, "Nathan Liskov"
wrote:

> In spite of the installation success that I reported, my system is
> still unstable.
>

That's too bad....



> There is a recent thread on an AMD Athlon 64 X2 with success using ECS
> 1.2 with SMP, but I am not using ECS.
>
> If there are any success stories with the AV8 Deluxe MB (or any other
> socket 939 MB) with warp 4.5.2, I would like to hear them.

The thread I started on the A8V Deluxe (socket 939) with an X2 3800+
was a success for eComstation 1.2 and for WSeB (ACP) as well. I have
an A8V Deluxe running a single CPU AMD 64 (a 3000+) under eComstation
1.2 and another A8V Deluxe running the X2 3800+ under WSeB (ACP) using
SMP.

The X2 box was set up by swapping out my old dual PIII 750 motherboard
and putting in the A8V - no installation was done, I just deinstalled
the old sound card drivers (Aureal) and the LAN card driver and
installed UNIAUD and the Marvell NIC driver (yukon.zip from
http://hobbes.nmsu.edu).

I tested the X2 with an SMP install of eComstation 1.2 that was
running on a Dual PII 400 motherboard and the disk booted up without a
problem.

I have no SCSI equipment on the X2 box so I cannot vouch for moding
over a SCSI installation.

--
Lorne Sunley

Trevor - thanks for the informed posting.

I would not say that my problems started when I installed the Seagate
Cheetah, but they did get worse. I have now ordered a proper SCSI
Ultra320 cable with terminator. Note that my old SCSI hard drive (IBM
DRVS18V) was an ultra-2 LVD drive that did work well for years.

I am not sure whether the 50 pin and 68 pin busses are isolated or
not. In my old machine, they were not (an adaptec controller) and my
active terminator, I think took this into accound. In any case on
that system I used only the 68 pin bus and one 68-pin cable with
68-to-50 pin adapters for each 50 pin device.

At the present moment, I can operate with my old (3200 Barton)
processor only at about 1500 MHz and with the new processor (2800
Barton) at about 2083 MHz, when the machine boots correctly (not every
time).

To confirm that SCSI cabling is the souce of my problems, I will do
the following ( while waiting for the new cable and possibly the new
controller):

1. Remove the SCSI controller board so the system has no SCSI
devices connected.

2. Install OS/2 on the first IDE drive (which will then be seen as
drive C) or if I have to (I could not get OS2CDROM.DMD to load
previously) install windows XP.

3. Test that the system is stable at the full processor speed.

I will defer a MB change until doing this, getting a good SCSI cable
and perhaps the upgraded SCSI controller.

I'll report back when results of these tests are in.

Thanks again

Nate Liskov

"Trevor Hemsley" wrote:

> On Thu, 15 Dec 2005 18:18:07 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
> wrote:
>
> > On one 68 pin connector I have a generic 68 pin cable with 5
> > connectors. The controller is on connector 1, drive 1 is on
> > connector 3, drive 0 on connector 4 and a passive terminator on
> > connector 5.
>
> If this is an LVD bus then it requires active termination. Did your problems
> start shortly after upgrading to the Seagate Cheetah? That's most likely also an
> LVD drive - maybe the controller is being confused and trying to negotiate
> ultra2 speeds? I'd try to find an active terminator in any case - they're
> recommended for UW buses and required for LVD. I'd also check the termination
> settings within the Tekram BIOS and make sure that it's set right - I think the
> 390U2B does not have an isolator that separates each segment of the bus so you'd
> either need it on automatic (which is sometimes unreliable anyway) or set to
> terminate the upper 18 lines of the bus but not the lower 50 (since you have a
> terminator on the 50 pin cable). Oh, and check the jumpers on all your devices
> and make sure that none of them have termination enabled by mistake - too many
> sources of termination is just as bad as too few!
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

On Thu, 15 Dec 2005 21:14:21 -0600, Nathan Liskov wrote:

>At the present moment, I can operate with my old (3200 Barton)
>processor only at about 1500 MHz and with the new processor (2800
>Barton) at about 2083 MHz, when the machine boots correctly (not every
>time).

Doesn't the 2800 have a clock speed of about 1600Mhz? The number is the
equivalent speed for a P4 and not the bus speed of the AMD so if you
are trying to run a 2800+ at 2083Mhz you are likely to have problems.
What happens if you run at 1600Mhz?

--
Regards - Rodney Pont
The from address exists but is mostly dumped,
please send any emails to the address below
e-mail ngpsm4 (at) infohitsystems (dot) ltd (dot) uk

I can only select from the four frequencies offerred by my BIOS which
were (2200, 1833, 1467, 1100) for the 3200 processor and (2500, 2083,
1666 and 1250) for the 2800 processor. These rates are 1X, 5/6X, 2/3X
and 1/2X, where X is the highest rate.

I find that stability increases as the speed is reduced. This makes
sense to mee, assuming I am having SCSI cabling reflection problems.

Nate Liskov

"Rodney Pont" wrote:

> On Thu, 15 Dec 2005 21:14:21 -0600, Nathan Liskov wrote:
>
> >At the present moment, I can operate with my old (3200 Barton)
> >processor only at about 1500 MHz and with the new processor (2800
> >Barton) at about 2083 MHz, when the machine boots correctly (not every
> >time).
>
> Doesn't the 2800 have a clock speed of about 1600Mhz? The number is the
> equivalent speed for a P4 and not the bus speed of the AMD so if you
> are trying to run a 2800+ at 2083Mhz you are likely to have problems.
> What happens if you run at 1600Mhz?
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

A friend of mine just bought ECS and told me that the manual claims
that ECS 1.2R is the only os/2 based system that will install on
Athlon 64 CPU based boards. If I have to upgrade my MB, I will
definitely give that a try.

Nate Liskov

"Lorne Sunley" wrote:

> On Thu, 15 Dec 2005 07:20:45 UTC, "Nathan Liskov"
> wrote:
>
> > In spite of the installation success that I reported, my system is
> > still unstable.
> >
>
> That's too bad....
>
>
>
> > There is a recent thread on an AMD Athlon 64 X2 with success using ECS
> > 1.2 with SMP, but I am not using ECS.
> >
> > If there are any success stories with the AV8 Deluxe MB (or any other
> > socket 939 MB) with warp 4.5.2, I would like to hear them.
>
> The thread I started on the A8V Deluxe (socket 939) with an X2 3800+
> was a success for eComstation 1.2 and for WSeB (ACP) as well. I have
> an A8V Deluxe running a single CPU AMD 64 (a 3000+) under eComstation
> 1.2 and another A8V Deluxe running the X2 3800+ under WSeB (ACP) using
> SMP.
>
> The X2 box was set up by swapping out my old dual PIII 750 motherboard
> and putting in the A8V - no installation was done, I just deinstalled
> the old sound card drivers (Aureal) and the LAN card driver and
> installed UNIAUD and the Marvell NIC driver (yukon.zip from
> http://hobbes.nmsu.edu).
>
> I tested the X2 with an SMP install of eComstation 1.2 that was
> running on a Dual PII 400 motherboard and the disk booted up without a
> problem.
>
> I have no SCSI equipment on the X2 box so I cannot vouch for moding
> over a SCSI installation.
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

On Fri, 16 Dec 2005 08:48:34 -0600, Nathan Liskov wrote:

>I can only select from the four frequencies offerred by my BIOS which
>were (2200, 1833, 1467, 1100) for the 3200 processor and (2500, 2083,
>1666 and 1250) for the 2800 processor. These rates are 1X, 5/6X, 2/3X
>and 1/2X, where X is the highest rate.

Can you select auto somewhere?

>I find that stability increases as the speed is reduced. This makes
>sense to mee, assuming I am having SCSI cabling reflection problems.

SCSI cabling problems shouldn't make any difference when you change the
processor speed, the cabling is on the SCSI side of the controller
which is isolated from the processor side.

--
Regards - Rodney Pont
The from address exists but is mostly dumped,
please send any emails to the address below
e-mail ngpsm4 (at) infohitsystems (dot) ltd (dot) uk

On Fri, 16 Dec 2005 14:48:34 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
wrote:

> I find that stability increases as the speed is reduced. This makes
> sense to mee, assuming I am having SCSI cabling reflection problems.

I change my mind then! This doesn't sound like SCSI problems though you may have
those too. As far as I can see from the AMD web site, a 3200+ is designed to run
at 2200MHz and if it won't then you have a problem there. How hot does it get?
Checked your temperatures with some sort of monitor?

--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com

I removed all drives from the system but the number one IDE hard disk
and the IDE DVD/CD drive. The IDE 32 Gb drive was partitioned for 2
gb for the operating system and 30 gb for whatever, with the 2 gb
partition made primary. There is no SCSI stuff in this test system
(other than the controller itself which I was too lazy to take out).

I installed a new os/2 system on this partition (processor operating
at 1100 mhz). I got beyond the refusal of OS2CDROM.DMD refusing to
load by putting IBMDECD.FLT back into CONFIG.SYS.

The installation went normally and seemed solid. However, I could not
boot with the processor set to 2200 mHz. The machine would hang at
the os/2 blob or just before it. I did get it to boot at speeds up to
1833 mHz but not at 2200 mHz.

Thus I come to the conclusion that the motherboard has developed a
problem that manifests itself at higher operating speed.

I did this test with my old (3200) processor, but since I had problems
with higher speed with the 2800 processor, I think the MB is much more
likely the problem.

The only anomaly that occurred is that I got DOSCALL1.DLL error
causing boot to fail (at any speed) when I added my SCSI driver
(TMSCSIW.ADD) to CONFIG.SYS. I wanted to try to access the SCSI
drives when booting from the IDE. I'll leave that problem for another
day.

I will still improve my SCSI cabling and controller to get full
transfer speed, but that does not seem to be the underlying problem.

I will get ECS 1.2 before changing the motherboard. In the meantime,
I'll keep operating at lower speed.

Thanks to everyone's help.

Nate Liskov



"Nathan Liskov" wrote:

> To confirm that SCSI cabling is the souce of my problems, I will do
> the following ( while waiting for the new cable and possibly the new
> controller):
>
> 1. Remove the SCSI controller board so the system has no SCSI
> devices connected.
>
> 2. Install OS/2 on the first IDE drive (which will then be seen as
> drive C) or if I have to (I could not get OS2CDROM.DMD to load
> previously) install windows XP.
>
> 3. Test that the system is stable at the full processor speed.
>
> I will defer a MB change until doing this, getting a good SCSI cable
> and perhaps the upgraded SCSI controller.
>
> I'll report back when results of these tests are in.
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

I have concluded that the MB is at fault as a new os/2 installation on
an IDE-only version of my system will not boot at full speed. See a
separate posting for details.

I do not have a temp monitor, however I have a Zalman fan which is
supposedly the best you can get.
When I switched processors, I should have refreshed the thermal paste
at the interface between the fan and the processor chip (I did not
have any), but I still think the MB is the problem.

Thanks for all your help.

Replacing the MB will be a fun winter project (good thing golf season
is over so I have the time).

Nate Liskov

"Trevor Hemsley" wrote:

> On Fri, 16 Dec 2005 14:48:34 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
> wrote:
>
> > I find that stability increases as the speed is reduced. This makes
> > sense to mee, assuming I am having SCSI cabling reflection problems.
>
> I change my mind then! This doesn't sound like SCSI problems though you may have
> those too. As far as I can see from the AMD web site, a 3200+ is designed to run
> at 2200MHz and if it won't then you have a problem there. How hot does it get?
> Checked your temperatures with some sort of monitor?
>


--
nate_NOSPAM@lcs.mit.edu http://nateliskov.ne.client2.attbi.com
or http://home.comcast.net/~nateliskov

On Fri, 16 Dec 2005 19:20:16 UTC in comp.os.os2.setup.misc, "Nathan Liskov"
wrote:

>
> I do not have a temp monitor, however I have a Zalman fan which is
> supposedly the best you can get.

You could try hobbes for pmp2l97 or something equivalent.

--
Trevor Hemsley, Brighton, UK.
Trevor-Hemsley at dsl dot pipex dot com

On Fri, 16 Dec 2005 13:11:30 -0600, Nathan Liskov wrote:

>The installation went normally and seemed solid. However, I could not
>boot with the processor set to 2200 mHz. The machine would hang at
>the os/2 blob or just before it. I did get it to boot at speeds up to
>1833 mHz but not at 2200 mHz.

The AMD 64 2800+ runs at 1800Mhz according to the AMD website
http://www.amdcompare.com/us-en/desktop/Default.aspx

>Thus I come to the conclusion that the motherboard has developed a
>problem that manifests itself at higher operating speed.

No, you are just trying to run the processor too fast.

--
Regards - Rodney Pont
The from address exists but is mostly dumped,
please send any emails to the address below
e-mail ngpsm4 (at) infohitsystems (dot) ltd (dot) uk