Some time ago I posted a PC problem - random mouse m,ovements and
windows opening and closing all teh time.

Well, it wasn't a virus, and changing the mouse seemed to help...but it
was getting worse and worse..and occasionally the keyboard locked as well.

A bt of a discussion with the local propellor heads elicited the fact
that to do THAT (have the keyboard lock up) was deep Hardware Magic.

My suspicions were already aroused,so I blagged a spare processor fan -
I have seen this sort of random weidrness on CPU's that got too hot before.

Bingo.

On inspection, a moth had died in the fan, and got its corsep stiuck
between the blades and teh heatsink, and teh subsequent build up of dist
fag ash, mouse droppings and who knows waht had clogged theheatsink.

A bit of a puff with the mark I lips pursed plus the MkII finger to
remove the moth, had her all up and running...so far...so good.

Just in case someone else has similar problems.


Must take the fan back. The old one works fine, sans moth...

"The Natural Philosopher" wrote in message
...
On inspection, a moth had died in the fan, and got its corsep stiuck
between the blades and teh heatsink, and teh subsequent build up of dist
fag ash, mouse droppings and who knows waht had clogged theheatsink.

You should ban your mice from smoking in there.

"The Natural Philosopher" wrote in message
...
Some time ago I posted a PC problem - random mouse m,ovements and
windows opening and closing all teh time.


snip fan woes


Must take the fan back. The old one works fine, sans moth...


Failing fans can cause all sorts of problems.

My trusty old linux router/gateway/firewall took upon itself to reboot
spontaneously and frequently. Most un-linux-like behaviour.

Cause was similar - fan blocked with crud and fluff, prolly sucked off the
cat when she decided that power supplies were a useful source of bedtime
warmth.

Quick blast through with Henry, and it's not rebooted once in the three
months that have passed.


cheers
Richard
--
Richard Sampson

email me at
richard at olifant d-ot co do-t uk

"The Natural Philosopher" wrote in message
...
....
Just in case someone else has similar problems.

At the moment my CPU is running at 31C, the main hard drive at 29C, inside
the case is 27C, the PSU is at 35C and all 11 fans are working quite well
thank you.

Colin Bignell

"The Natural Philosopher" wrote
| A bit of a puff with the mark I lips pursed plus the MkII finger

That's pretty advanced. I'm still on MkI for all body parts apart from my
MkII teeth.

Owain

Owain wrote:

"The Natural Philosopher" wrote
| A bit of a puff with the mark I lips pursed plus the MkII finger

That's pretty advanced. I'm still on MkI for all body parts apart from my
MkII teeth.

I get a *lot* of spam offering me various hardware upgrades.

Mary Pegg wrote:

Owain wrote:


"The Natural Philosopher" wrote
| A bit of a puff with the mark I lips pursed plus the MkII finger

That's pretty advanced. I'm still on MkI for all body parts apart from my
MkII teeth.


I get a *lot* of spam offering me various hardware upgrades.


Somehow I feel that penis enlargement techniques might nor suit you..

Andy R wrote:

"The Natural Philosopher" wrote in message
...

Some time ago I posted a PC problem - random mouse m,ovements and
windows opening and closing all teh time.

Well, it wasn't a virus, and changing the mouse seemed to help...but it
was getting worse and worse..and occasionally the keyboard locked as well.

A bt of a discussion with the local propellor heads elicited the fact
that to do THAT (have the keyboard lock up) was deep Hardware Magic.

My suspicions were already aroused,so I blagged a spare processor fan -
I have seen this sort of random weidrness on CPU's that got too hot

before.

Bingo.

On inspection, a moth had died in the fan, and got its corsep stiuck
between the blades and teh heatsink, and teh subsequent build up of dist
fag ash, mouse droppings and who knows waht had clogged theheatsink.

A bit of a puff with the mark I lips pursed plus the MkII finger to
remove the moth, had her all up and running...so far...so good.

Just in case someone else has similar problems.


Must take the fan back. The old one works fine, sans moth...



I've had this prob but resolved by a different means. If your prob recurs
do a KB search on www.microsoft.com for "erratic mouse", it has several
articles.



Sadly its back again, tho much less annoying and I think I may have
damaged the Pentium III beyond repair..I think the timing has driufted
somewhere and teh chip is getting stick in an interrupt service routine
sometimes.


Rgds

Andy R

The Natural Philosopher wrote in message ...
Mary Pegg wrote:

Owain wrote:


"The Natural Philosopher" wrote
| A bit of a puff with the mark I lips pursed plus the MkII finger

That's pretty advanced. I'm still on MkI for all body parts apart from my
MkII teeth.


I get a *lot* of spam offering me various hardware upgrades.


Somehow I feel that penis enlargement techniques might nor suit you..

Depends whose penis it is!

The Natural Philosopher wrote in message ...
Some time ago I posted a PC problem - random mouse m,ovements and
windows opening and closing all teh time.

Hi

Although I didn't quite have that set of problems I once had very
annoying mouse behaviour.

After a bit of analysis I worked out it was because I used the mouse
right next to the window. On sunny days the light was strong enough
to penetrate the case and confuse the opto sensors. A new mouse with
a darker plastic case fixed it.

IanC

The Natural Philosopher wrote:

Hello The

TNP| On inspection, a moth had died in the fan, and got its
TNP| corsep stiuck between the blades and teh heatsink, and teh
TNP| subsequent build up of dist fag ash, mouse droppings and who
TNP| knows waht had clogged theheatsink.

MBM5 is a wonderful tool for diagnostics.

TNP| Must take the fan back. The old one works fine, sans moth...

Keep it. The day after you return it, your existing fan is bound to
die for real.

--
Simon Avery, Dartmoor, UK
uk.d-i-y FAQ: http://www.diyfaq.org.uk/

Ian Clowes wrote:

After a bit of analysis I worked out it was because I used the mouse
right next to the window. On sunny days the light was strong enough
to penetrate the case and confuse the opto sensors. A new mouse with
a darker plastic case fixed it.

I found that once a few years ago with a mouse - it only moved up and
down on sunny days! The solution was to take the lid off it and colour
the inside with a permenent marker pen - put it all back together and it
worked fine after that!

--
Cheers,

John.

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

On Wed, 17 Sep 2003 16:33:03 GMT, (Simon
Avery) wrote:

Keep it. The day after you return it, your existing fan is bound to
die for real.

In fact, having a spare is a good way to ensure you never need one -
makes all fans last forever! Well known that one! ;O)

Take Care,
Gnube
{too thick for linux}

"John Rumm" wrote in message
...
Ian Clowes wrote:

After a bit of analysis I worked out it was because I used the mouse
right next to the window. On sunny days the light was strong enough
to penetrate the case and confuse the opto sensors. A new mouse with
a darker plastic case fixed it.

I found that once a few years ago with a mouse - it only moved up and
down on sunny days! The solution was to take the lid off it and colour
the inside with a permenent marker pen - put it all back together and it
worked fine after that!

That is hardly the case here because the "double-click" is being activated
and it is mechanical rather than optical.

Sadly its back again, tho much less annoying and I think I may have
damaged the Pentium III beyond repair..I think the timing has driufted
somewhere and teh chip is getting stick in an interrupt service routine
sometimes.

A couple of things spring to mind.
Not the mouse, as it happens with more than one. Not the optical sensors as
double clicks are being activated.
Maybe a faulty mouse port? Try a serial mouse, USB mouse - one different to
current one.
A stuck key (or 2) on the keyboard. Have seen this more than once. Alt
stuck on mine, and I caught on when pressing F4 as it tried to close the
program or tab shifted to another one.
Does it only happen when you are working on it, or does it happen when you
aren't touching it? If it happens when you aren't touching it then I'd look
at the mouse port, otherwise the stuck keys. It's always a good idea to
give the keyboard a clean anyway, but if it isn't dirt, just sticky keys
then a good sharp stabbing on the usual offenders may help. Alt -esp if
windows are closing, but also Ctrl, arrow keys, Backspace, tab, shift. Do
you have accessibilty options switched on, which allow keys to control mouse
movements and if one key is stuck may generate the movements etc. Or an
alternative KB if you have one lying about.

If the kb is really bogging a good scrub with with soapy water is all else
fails. Just don't put it up against the radiator to dry out, as a colleague
of hubby's did. Went limp and he ended up with a banana shaped KB.

PS what makes you think its getting stuck in an interrupt service routine?
Thats not easy to diagnose.
Suzanne

In article , Suz
writes

"John Rumm" wrote in message
...
Ian Clowes wrote:

After a bit of analysis I worked out it was because I used the mouse
right next to the window. On sunny days the light was strong enough
to penetrate the case and confuse the opto sensors. A new mouse with
a darker plastic case fixed it.

I found that once a few years ago with a mouse - it only moved up and
down on sunny days! The solution was to take the lid off it and colour
the inside with a permenent marker pen - put it all back together and it
worked fine after that!

That is hardly the case here because the "double-click" is being activated
and it is mechanical rather than optical.

Even "non optical" mice still sense the ball movement optically using
slotted discs.
--
Tim Mitchell

Suz wrote:

I found that once a few years ago with a mouse - it only moved up and
down on sunny days! The solution was to take the lid off it and colour
the inside with a permenent marker pen - put it all back together and it
worked fine after that!


That is hardly the case here because the "double-click" is being activated
and it is mechanical rather than optical.

Sorry you might have got the wrong impression of my answer - I was being
all anecdotal and not attempting to solve the OPs problem at all! ;-)


--
Cheers,

John.

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

The Natural Philosopher wrote:

No. Happens randomly when nothing else is going on. Rebbot (hard) always
fixes it, so cannot be stuick key. Ctrl Alt Del doesn';t work at all,
neither does any key - e.g. NuimLock - cause any keyboard light to
change state. The amount of hardware and software involved in e.g.
getting a keybaord light to change state is very small - probably
tehkeyboard, teh UART and a timney code fragment invoked under interrupt.

Its definitely UART, processor or memory.

Or a corrupted mouse or keyboard driver - have you tried expanding these
again from your windowss CD?

Also check if your BIOS has a fast A20 gate option. If it does, and its
turned off, then turn it on and try again! (Why is a long and complex
story!)

Have you also tried a differnet keyboard?

Its possible i have a corrupted disk but I doubnt it. My giess is that
teh interrup part of the chipset has gone off the edge of its timing
somehow, and/or the UART is faling to respond correctly,

If it's hardware you would probably find a motherboard change would fix it.

The processor should be pretty well isolated from anything nasty
happening to the PS/2 mouse port. So unless you have roasted it good and
proper it ought to be OK.

PS what makes you think its getting stuck in an interrupt service
routine?
Thats not easy to diagnose.



20 years as a hardware engineer and ten as a real time software engineer?


Its very easy to diagnose. Only an ISR turns off enough interrupts to
e.g. make the keyboard caps lock and num lock toggles stop working.

A "CLI" would do it as well! Note also that the keyboard is on IR1 on
the master PIC so it usually has priority over most interrupts except
the system timer (assuming the 82c59 equivilent in the chipset is still
set to "special fully nested" mode, as it is in the BIOS initialisation!
probably a rash assumption these days since the whole lot is virtualised
by protected mode code in the VXDs that deal with mouse and keyboard anyway)

I need new hardware. No sure if a Pentium III is the way to go, or
simply swap the motherboard. Is only 18 months old too.

Socket 370 I hope ;-)


--
Cheers,

John.

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

In message , The Natural Philosopher
writes
Mary Pegg wrote:

Owain wrote:

"The Natural Philosopher" wrote
| A bit of a puff with the mark I lips pursed plus the MkII finger

That's pretty advanced. I'm still on MkI for all body parts apart from my
MkII teeth.

I get a *lot* of spam offering me various hardware upgrades.


Somehow I feel that penis enlargement techniques might nor suit you..

Especially if you get it caught in the fan
--
geoff

geoff wrote:

Somehow I feel that penis enlargement techniques might nor suit you..

Especially if you get it caught in the fan

A friend works for a company that rent out AV kit for corporate do's
etc. Hence quite often he finds himself at various functions driving
some bit of AV gear, and running slide shows etc.

One such occation was a big NHS conference where amoung other things
discussed were some of the strange "accidents" thet the A&E people have
to deal with.

Apparently one of the most common causes of penile shaft injury in the
UK is (or was - this was a few years ago now) - the Dustbuster type of
rechargeable mini vacuum cleaners. Apparently some of them have a
rotating fan blade with sharp edges not that far from the nose of the
machine. I will leave the rest to your collective imaginations!

Salami anyone?


--
Cheers,

John.

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

In message , John Rumm
writes
geoff wrote:

Somehow I feel that penis enlargement techniques might nor suit you..

Especially if you get it caught in the fan

A friend works for a company that rent out AV kit for corporate do's
etc. Hence quite often he finds himself at various functions driving
some bit of AV gear, and running slide shows etc.

One such occation was a big NHS conference where amoung other things
discussed were some of the strange "accidents" thet the A&E people have
to deal with.

Apparently one of the most common causes of penile shaft injury in the
UK is (or was - this was a few years ago now) - the Dustbuster type of
rechargeable mini vacuum cleaners. Apparently some of them have a
rotating fan blade with sharp edges not that far from the nose of the
machine. I will leave the rest to your collective imaginations!

Salami anyone?


Only if he's rolled it in crushed peppers first

--
geoff

"geoff" wrote in message
...
In message , nightjar
writes

"The Natural Philosopher" wrote in message
...
...
Just in case someone else has similar problems.

At the moment my CPU is running at 31C, the main hard drive at 29C,
inside
the case is 27C, the PSU is at 35C and all 11 fans are working quite well
thank you.

11 ? I'd put a few more in if I was you

I would need a different case and it might get a bit noisy if I did.

Colin Bignell

John Rumm wrote:

The Natural Philosopher wrote:

No. Happens randomly when nothing else is going on. Rebbot (hard)
always fixes it, so cannot be stuick key. Ctrl Alt Del doesn';t work
at all, neither does any key - e.g. NuimLock - cause any keyboard
light to change state. The amount of hardware and software involved in
e.g. getting a keybaord light to change state is very small - probably
tehkeyboard, teh UART and a timney code fragment invoked under interrupt.

Its definitely UART, processor or memory.


Or a corrupted mouse or keyboard driver - have you tried expanding these
again from your windowss CD?



not a bad idea.


Also check if your BIOS has a fast A20 gate option. If it does, and its
turned off, then turn it on and try again! (Why is a long and complex
story!)


mmm.


Have you also tried a differnet keyboard?



nope. I don;t think that is th eproblem tho...


Its possible i have a corrupted disk but I doubnt it. My giess is that
teh interrup part of the chipset has gone off the edge of its timing
somehow, and/or the UART is faling to respond correctly,


If it's hardware you would probably find a motherboard change would fix it.

The processor should be pretty well isolated from anything nasty
happening to the PS/2 mouse port. So unless you have roasted it good and
proper it ought to be OK.


I roasted it good and proper - that fan has been stick all summer I reckon.

Heatsink was about 70C with the fan off.


Further data.


The random mouse movements have now almost completely gone, with
restarting of the fan. The current case is that the processor seems to
get 'stuck' - presumably in an ISR - for periods of between a few
milliseconds to permanently. Sometimes the keyboard works, and
Ctrl-Alt=-Del will allow a soft reboot, and sometimes its truly stuck.

The fact that its oinconsistent and seems temperature dependent leads me
to a hardware explantion.

Is the keyboard UART integrated on the pentium? Or is it a separate chip?.

I can change motherboard, processor or m,emory, but don';t want to
change all three - expensiove









PS what makes you think its getting stuck in an interrupt service
routine?
Thats not easy to diagnose.




20 years as a hardware engineer and ten as a real time software engineer?


Its very easy to diagnose. Only an ISR turns off enough interrupts to
e.g. make the keyboard caps lock and num lock toggles stop working.


A "CLI" would do it as well!


Yes, but that would simply freeze everything.

Note also that the keyboard is on IR1 on
the master PIC so it usually has priority over most interrupts except
the system timer (assuming the 82c59 equivilent in the chipset is still
set to "special fully nested" mode, as it is in the BIOS initialisation!


Its windows 98 so its fairly primitive.


probably a rash assumption these days since the whole lot is virtualised
by protected mode code in the VXDs that deal with mouse and keyboard
anyway)



In W98?


I need new hardware. No sure if a Pentium III is the way to go, or
simply swap the motherboard. Is only 18 months old too.


Socket 370 I hope ;-)



Dunno. Lost of pins and a square socket.

Or a corrupted mouse or keyboard driver - have you tried expanding these
again from your windowss CD?

Would it not be more likely to stop working if the driver was corrupted? It
would be a million to one chance of it corrupting and still be usuable part
of the time.


If it's hardware you would probably find a motherboard change would fix
it.

The processor should be pretty well isolated from anything nasty
happening to the PS/2 mouse port. So unless you have roasted it good and
proper it ought to be OK.

I agree.

"Tim Mitchell" wrote in message
...
In article , Suz
writes

"John Rumm" wrote in message
...
Ian Clowes wrote:

After a bit of analysis I worked out it was because I used the mouse
right next to the window. On sunny days the light was strong enough
to penetrate the case and confuse the opto sensors. A new mouse with
a darker plastic case fixed it.

I found that once a few years ago with a mouse - it only moved up and
down on sunny days! The solution was to take the lid off it and colour
the inside with a permenent marker pen - put it all back together and
it
worked fine after that!

That is hardly the case here because the "double-click" is being
activated
and it is mechanical rather than optical.

Even "non optical" mice still sense the ball movement optically using
slotted discs.
--
This is true (but besides the point). But the buttons are mechanical and
not optical, and the OP's problem included windows being opened and closed,
which may mean the double-click signal is being generated somewhere along
the way.

The Natural Philosopher wrote:

Also check if your BIOS has a fast A20 gate option. If it does, and
its turned off, then turn it on and try again! (Why is a long and
complex story!)


mmm.


Have you also tried a differnet keyboard?



nope. I don;t think that is th eproblem tho...

Unlikely granted - but due to a strange IBM design decision many years
ago - potentially related! (A gate introduced into the A20 address line
of the first 286 based PC (the AT - to allow more accurate emulation of
the original 8088 memory architecture) was implemented in the keyboard
controller - since they was the main bit of original design they did -
the rest was lifted from Intel / Shugart data sheets. They also lumped a
few other bits of unrelated functionality in there - like the ability to
reset the processor! Hence on the first PC-AT computers things like
getting the 286 to switch from protected mode back to real mode (which
required a controlled reset) involved sending a message to the keyboard
controller over a serial link!

I roasted it good and proper - that fan has been stick all summer I reckon.

Heatsink was about 70C with the fan off.

70 does not actually sound that bad.... might have done some damage -
but not a dead cert.

The fact that its oinconsistent and seems temperature dependent leads me
to a hardware explantion.

I would tend to agree. There has also been a spate of motherboard
problems recently due to poor quality capacitors that were supplied to a
number of the big OEMs and motherboard makers. Many of these are
starting to show up as failures now they are a couple of years old. This
would also hold with your suspicion of a timing related problem.

Is the keyboard UART integrated on the pentium? Or is it a separate chip?.

It is integrated into the "chip set" on the motherboard - rather than
the processor itself.

I can change motherboard, processor or m,emory, but don';t want to
change all three - expensiove

How many DIMMs do you have fitted?

If more then one, you could try running with only one device fitted for
a bit - then swap to the other and see if that has any impact. If it's a
memory problem then its unlikely to be a fault in both devices.

If only 1 DIMM, then buy another one the same and try that (RAM is cheap
enough at the moment!). Even if it does not fix the problem you at least
have an upgrade for your money!

Next thing to replace would be the motherboard.

the master PIC so it usually has priority over most interrupts except
the system timer (assuming the 82c59 equivilent in the chipset is
still set to "special fully nested" mode, as it is in the BIOS
initialisation!



Its windows 98 so its fairly primitive.

You will find the interrupt handling on W98 is actually pretty
sophisticated! It uses several layers of drivers split into different
processor operating modes. This is mostly to allow virtualisation of the
DOS environment. A DOS application can run, and bang away at the
interrupt hardware as much as it wants, in the full belief that it is
twiddling bits on the actually interrupt controller. Meanwhile there are
several other applications all doing the same in parallel! This is all
handled by a virtualisation of the interrupt controller implemented in a
VXD.

probably a rash assumption these days since the whole lot is
virtualised by protected mode code in the VXDs that deal with mouse
and keyboard anyway)



In W98?

Very much so. In some respects W98 has to pull some quite "clever"
tricks to achieve what it does - not because of its stunning design -
more a case of given the choice you would not have started from DOS!

Socket 370 I hope ;-)



Dunno. Lost of pins and a square socket.

That is socket 370 then - motherboards are getting thin on the ground
but some should still available in this form. (the first PIIIs were
"Slot 1" devices - getting a MB for one of these would be a non starter).

What case style have you got AT or ATX (the latter are easy to recognise
because they have a rectangular section at the back where all the
motherboard mounted IO connectors meet the outside world).

Again if you have an AT style case then getting a motherboard will be a
non starter.

Most new motherboards will be either Socket A (AMD) or Socket 478 (Intel)

If it's any use by way of illustration - I just upgraded a customers
computer with a new fully integrated motherboard and a new Duron 1.3GHz
processor - reused the existing RAM. Total cost was about 80 all in for
the hardware.


--
Cheers,

John.

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

In uk.d-i-y, John Rumm wrote:

How many DIMMs do you have fitted?

If more then one, you could try running with only one device fitted for
a bit - then swap to the other and see if that has any impact. If it's a
memory problem then its unlikely to be a fault in both devices.

If only 1 DIMM, then buy another one the same and try that (RAM is cheap
enough at the moment!). Even if it does not fix the problem you at least
have an upgrade for your money!

Rather less invasive, and usually at least as effective, is to run a
good standalone memory diagnostic: and the greatest of these is memtest86.
See www.memtest86.com - despite the .com it's GPL'd freeware. You download
the entire binary and put it on one floppy or write a bootable CD-ROM
(instructions on the web page), and instead of booting Windows, Linux,
BeOS, or anything else, you boot into the specialised memory torture test.
I've tracked down obscure and 'orrible incompatibilities between RAM
sticks (or even sticks temperemantal about which slot they sit in!) and
Athlon motherboards with memtest86: recommended. Like I say, requires
none of those heart-stopping "put the DIMM in firmly, no not *that*
firmly" moments, and pretty authoritative in tracking down bad RAM -
if you can run 24 hours of memtest86's "thorough" tests without a fault,
your memory system has to all intents and purposes a thoroughly clean bill
of health.

Well (ob. d-i-y), it is fixing-yourself, rather than taking-it-in-to-be-fixed
;-)

HTH, Stefek

On Fri, 19 Sep 2003 09:33:44 +0100, The Natural Philosopher
wrote:

The current case is that the processor seems to
get 'stuck' - presumably in an ISR - for periods of between a few
milliseconds to permanently. Sometimes the keyboard works, and
Ctrl-Alt=-Del will allow a soft reboot, and sometimes its truly stuck.

OK, this is a long shot, but I solved something like that this way, so
it may help.

I was using a microsoft strategic commander. once I wondered about it,
and removed it, and un installed it's software the problem (oddly)
went away, and has stayed away ever since (more than a month now).

That product isn't too bad in use, but it's best if it's not connected
to a machine with too many other input connections, as it does appear
to have more than it's share of clashes with other items. On a machine
with just basic stuff, and it being the only game type controller,
then it's less likely to cause an issue. Since I tinker with joysticks
(build own variations quite a bit) then I can't have it on my main
machine due to the havoc it can bring, but on a single use machine,
then fair enough, it can be great. I had it's keys set up specifically
for news reading, which made that quite a groovy experience. Now found
other ways to automate some of that.

So, the point is this, it may not be the equipment appearing to have
the fault which is actually faulty or is the culprit! Means you may
have to think sideways about whatever you may have attached. For
example, mine was not just causing those lock ups, but also causing
trouble for directx and a number of multimedia apps (used to jam while
playing video or audio files too!) This is completely unrelated, but
the effect clearly wasn't with hindsight!

I note that MS have stopped making their PC range of Sidewinder
products now. I shall miss them as they made great USB donor sticks a
lot of the time, and the more recent software was nowhere near as
borked as the early stuff was, and that early USB stuff was pretty
dire in some cases. I'm not sure which make I'll migrate to as they
all seem a bit wimpy in the hardware dept by comparison. The MS stuff
was nicely over engineered in all the right places for my use!

Saitek seem a good move next, as the hardware in some cases is cheap
to buy due to their software becoming rubbish once win2K and beyond
were released, so that holds the value of the hardware artificially
low, which suits me fine right now! The hardware has similar muscular
properties in places, so it'll last a while! There are other ways to
make drivers available for this stuff, so it's not as mad as it may
look!

Hope you find a culprit nestling in there like I did! Great once
removed!

Take Care,
Gnube
{too thick for linux}

In message , nightjar
writes

"geoff" wrote in message
...
In message , nightjar
writes

"The Natural Philosopher" wrote in message
...
...
Just in case someone else has similar problems.

At the moment my CPU is running at 31C, the main hard drive at 29C,
inside
the case is 27C, the PSU is at 35C and all 11 fans are working quite well
thank you.

11 ? I'd put a few more in if I was you

I would need a different case and it might get a bit noisy if I did.

Might ?

Maybe it's time you went for a water cooled PC
--
geoff

In message , The Natural Philosopher
writes
John Rumm wrote:

Have you also tried a differnet keyboard?



nope. I don;t think that is th eproblem tho...


Sorry, I had to laugh


--
geoff

"The Natural Philosopher" wrote in message
...
Owain wrote:

"The Natural Philosopher" wrote
| Is the keyboard UART integrated on the pentium? Or is it a separate
chip?.
| I can change motherboard, processor or m,emory, but don';t want to
| change all three - expensiove

Could you just try a USB keyboard and mouse instead of PS/2 ones? I
don't
know enough about USB to know how separately it's handled in the chipset
from the standard ports.

Owain


ASadly I am already short of USB ports, and not sure if 98 supports USB
k/b and mouse.


It does. This very mousie is running under 98 and is USB. 95 doesn't. 98
supports USB I and USB II.

Try a serial mouse instead.

Suzanne

"The Natural Philosopher" wrote in message
...
Suz wrote:

Sadly its back again, tho much less annoying and I think I may have
damaged the Pentium III beyond repair..I think the timing has driufted
somewhere and teh chip is getting stick in an interrupt service routine
sometimes.


A couple of things spring to mind.
Not the mouse, as it happens with more than one. Not the optical
sensors as
double clicks are being activated.


Yes. Beebn down that route as well.


Maybe a faulty mouse port? Try a serial mouse, USB mouse - one
different to
current one.


Tried a new mourse
Not that.


A stuck key (or 2) on the keyboard. Have seen this more than once. Alt
stuck on mine, and I caught on when pressing F4 as it tried to close the
program or tab shifted to another one.


No. Happens randomly when nothing else is going on. Rebbot (hard) always
fixes it, so cannot be stuick key. Ctrl Alt Del doesn';t work at all,
neither does any key - e.g. NuimLock - cause any keyboard light to
change state. The amount of hardware and software involved in e.g.
getting a keybaord light to change state is very small - probably
tehkeyboard, teh UART and a timney code fragment invoked under interrupt.

Its definitely UART, processor or memory.


Does it only happen when you are working on it, or does it happen when
you
aren't touching it? If it happens when you aren't touching it then I'd
look
at the mouse port, otherwise the stuck keys.



Seems to be that using the mouse makes it happen NOW, post getting fan
working.

Its entirely consistent with processor getting stuck in mouse interrupt
service routine. Sometimes removing mouse and re-inserting works,
sometimes nothing works.

Its possible i have a corrupted disk but I doubnt it. My giess is that
teh interrup part of the chipset has gone off the edge of its timing
somehow, and/or the UART is faling to respond correctly,


It's always a good idea to
give the keyboard a clean anyway, but if it isn't dirt, just sticky keys
then a good sharp stabbing on the usual offenders may help. Alt -esp if
windows are closing, but also Ctrl, arrow keys, Backspace, tab, shift.
Do
you have accessibilty options switched on, which allow keys to control
mouse
movements and if one key is stuck may generate the movements etc. Or an
alternative KB if you have one lying about.

If the kb is really bogging a good scrub with with soapy water is all
else
fails. Just don't put it up against the radiator to dry out, as a
colleague
of hubby's did. Went limp and he ended up with a banana shaped KB.

PS what makes you think its getting stuck in an interrupt service
routine?
Thats not easy to diagnose.


20 years as a hardware engineer and ten as a real time software engineer?

Bet you remember punched card stuff. ¦^)))

Suz wrote:

"The Natural Philosopher" wrote in message

Bet you remember punched card stuff. ¦^)))



Vaguely. Paper tape more.

I gave up on computers in 1967, until 1980 and the advent of the micro.

"The Natural Philosopher" wrote
| Bet you remember punched card stuff. ¦^)))
| Vaguely. Paper tape more.
| I gave up on computers in 1967, until 1980 and the advent of the micro.

I remember visiting Aberystwyth in the late 1980s and they were using
punched cards and just introducing their New Exciting Time-Share Service.
Stirling were also using punched cards as part of their admissions procedure
at the time.

Owain

On Sat, 20 Sep 2003 10:08:55 +0100, The Natural Philosopher
wrote:

ASadly I am already short of USB ports, and not sure if 98 supports USB
k/b and mouse.

95 doesn't 98 does after a fashion, but I don't think I'd be overly
harsh to suggest it should not hold it's breath waiting for an award
for it's efforts! ;O) Almost everything else can, more or less without
issue.

Take Care,
Gnube
{too thick for linux}

Gnube wrote:

On Sat, 20 Sep 2003 10:08:55 +0100, The Natural Philosopher
wrote:


ASadly I am already short of USB ports, and not sure if 98 supports USB
k/b and mouse.


95 doesn't 98 does after a fashion, but I don't think I'd be overly
harsh to suggest it should not hold it's breath waiting for an award
for it's efforts! ;O) Almost everything else can, more or less without
issue.

Take Care,
Gnube
{too thick for linux}


Hmm. Since the overall room temperatire has come down,. all problems
seem to have vanished.
Hmmm.

Maybe this can wait till next sumer now :-)

On Mon, 22 Sep 2003 20:40:42 +0100, The Natural Philosopher
wrote:

Gnube wrote:

On Sat, 20 Sep 2003 10:08:55 +0100, The Natural Philosopher
wrote:


ASadly I am already short of USB ports, and not sure if 98 supports USB
k/b and mouse.


95 doesn't 98 does after a fashion, but I don't think I'd be overly
harsh to suggest it should not hold it's breath waiting for an award
for it's efforts! ;O) Almost everything else can, more or less without
issue.

Take Care,
Gnube
{too thick for linux}


Hmm. Since the overall room temperatire has come down,. all problems
seem to have vanished.
Hmmm.

Maybe this can wait till next sumer now :-)

We could "park" this thread next to the "air con" ones until then, any
special "mothballing" requirements for it's standdown time? ;O)

Take Care,
Gnube
{too thick for linux}

"Gnube" wrote
| We could "park" this thread next to the "air con" ones until then,
| any special "mothballing" requirements for it's standdown time? ;O)

Wipe with an oily rag and store in a cool dry place out of the reach of pets
and children.

Owain