I just thought I'd start a little thread for people to discuss what te
subject line might mean.

The reason being I was chasing down some unrelated issues when I noticed
some 'reduce CPU speed - over temperature' messages in the log files.

Since SWMBO was out dog-walking, the chance to power down and open the
case was taken. Sheesh. The finned CPU heatsink was solid fluff. I tried
vacuuming, but no joy, and lacking and airline all I could think off was
a household paintbrush.

That did get into all the nooks an crannies and I was able to clean the
filters and the fans and the heatsinks (one PSU fan and grille, one CPU
fan and finned heatsink, one GPU fan abnd heatsink ) and tip the crud
out onto the floor, where SWMBO didnt notice it next time she vacuumed..

I dunno if the machine is faster, but the warnings have gone..and its
quieter too.


Another thing I do on the server, is monitor it for internal disk errors.

http://www.linuxjournal.com/magazine...rd-disks-smart

However a word of caution. When a drive started making bad noises on the
server, it showed nothing. The disk died irrecoverably a week later. It
was fortunately only the backup disk, so a simple 1:1 replacement lead
to no data loss at all.

teh final tutrine thing is to clean te compoutercase externally and
songe eberything down with a slightly damp cloth.

You may also want to replace mice and keyboards periodically as they do
suffer death - although mice are generally better now they aren't trackball.

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years, and I would say that 5 years is the limit for a hard
disk drive too. Yes, they may last longer, but that is when the chance
of catastrophic failure starts to rise steeply. Motherboards have been
usasble up to 10 yearas IME, and I only replace them whn I can get more
performance for less or the same money - that used to be three years,
notw it's well over 5..CPU development has it seems slowed
substantially, with all the effort going into either servers with
millions of cores, or portable devices using lower power.

Feel free to add any other things you consider are useful under the
general heading.

I've left out software upgrades, because these days most people do those
on a regular basis as they become avaliable, esp,.. if they are using Linux.

I've left out backups as well, because those have been extensively
discussed before.

Suffice to say that irreplaceable data should be held in at least three
places if possible on at least two machines one of which if you REALLY
want to survive a house fire etc, ought not to be in the same house..



--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

teh final tutrine thing is to clean te compoutercase externally and
songe eberything down with a slightly damp cloth.

If you can't be arsed to expend the slightest effort on writing your
posts, why should the rest of us be arsed to expend any on reading them?

On 01/10/13 12:42, Adrian wrote:
On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

teh final tutrine thing is to clean te compoutercase externally and
songe eberything down with a slightly damp cloth.

If you can't be arsed to expend the slightest effort on writing your
posts, why should the rest of us be arsed to expend any on reading them?

just **** off.


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On 01/10/2013 12:29, The Natural Philosopher wrote:

I dunno if the machine is faster, but the warnings have gone..and its
quieter too.

I had one the other day where one of the complaints was its very
"noisy". I was expecting a failing fan with noisy bearings. What I found
was an AMD fan on a Athlon 6000+ 3GHz chip that was just exceptionally
powerful when on full speed (we are talking hair drier loud here). It
was running in temperature sensitive mode - so once the machine was up
to around 50% CPU loading it was screaming at you.

Speedfan showed that it was running in the high 40s when unloaded and
was pushing up into the 75 deg C+ ranges when loaded.

I took the whole cooler off the CPU and blew it clean with an airline,
but also noticed it had a dry contact with the CPU - there may have been
some form of conductive pad there at one point but there was not much in
evidence now. Gave it a good blob of thermal paste, and reassembled.

Retesting showed unloaded temperatures back down to the mid 20s, and
peak rising to mid 50s. Much quieter!


--
Cheers,

John.

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

At least you can easily clean the fans and sinks on a desktop, some
laptops are not so straightforward

On 01/10/2013 12:29, The Natural Philosopher wrote:
I just thought I'd start a little thread for people to discuss what te
subject line might mean.

The reason being I was chasing down some unrelated issues when I noticed
some 'reduce CPU speed - over temperature' messages in the log files.

Since SWMBO was out dog-walking, the chance to power down and open the
case was taken. Sheesh. The finned CPU heatsink was solid fluff. I tried
vacuuming, but no joy, and lacking and airline all I could think off was
a household paintbrush.

That did get into all the nooks an crannies and I was able to clean the
filters and the fans and the heatsinks (one PSU fan and grille, one CPU
fan and finned heatsink, one GPU fan abnd heatsink ) and tip the crud
out onto the floor, where SWMBO didnt notice it next time she vacuumed..

I dunno if the machine is faster, but the warnings have gone..and its
quieter too.


I was recently prompted to do a major fluff clear-out by the noise when
the fan ramped up. Also replaced the CPU thermal compound.

On 01/10/13 15:50, newshound wrote:
On 01/10/2013 12:29, The Natural Philosopher wrote:
I just thought I'd start a little thread for people to discuss what te
subject line might mean.

The reason being I was chasing down some unrelated issues when I noticed
some 'reduce CPU speed - over temperature' messages in the log files.

Since SWMBO was out dog-walking, the chance to power down and open the
case was taken. Sheesh. The finned CPU heatsink was solid fluff. I tried
vacuuming, but no joy, and lacking and airline all I could think off was
a household paintbrush.

That did get into all the nooks an crannies and I was able to clean the
filters and the fans and the heatsinks (one PSU fan and grille, one CPU
fan and finned heatsink, one GPU fan abnd heatsink ) and tip the crud
out onto the floor, where SWMBO didnt notice it next time she vacuumed..

I dunno if the machine is faster, but the warnings have gone..and its
quieter too.


I was recently prompted to do a major fluff clear-out by the noise when
the fan ramped up. Also replaced the CPU thermal compound.

replacing thermal comopund is one to note, and I reckon its worth doing.


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On 01/10/2013 14:12, Lee wrote:
At least you can easily clean the fans and sinks on a desktop, some
laptops are not so straightforward

Well to be fair they are easy enough to clean once you can get to them!
Getting there however is not always straightforward.

--
Cheers,

John.

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

The Natural Philosopher wrote:
On 01/10/13 12:42, Adrian wrote:
On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

teh final tutrine thing is to clean te compoutercase externally and
songe eberything down with a slightly damp cloth.

If you can't be arsed to expend the slightest effort on writing your
posts, why should the rest of us be arsed to expend any on reading
them?
just **** off.

TNP,

I regularly clean the innards of several computers in the household using on
of those small air-pumps used to inflate plastic paddling pools etc for the
grand-kids and purchased for that grand sum of around a fiver from Aldi's
and beats spending a fiver or more a time for canned air. Getting rid of the
dust build-up greatly removes the cause of heat build-up inside the case
which can be a great 'killer' of some of the components.

I a have also made some flexible attachments using small bore tube inserted
into the air pump nozzle to get into some of the more inaccessible places -
along with a small, clean good-quality 1" paint brush to remove any gunge
that has hardened.

I do that little job about twice a year and that keeps everything spic and
span and occasionally getting rid of the odd whine cause by dust sticking to
the CPU and PSU fans.

If I'm really in a workish mood, during the cleaning period I will also
remove and re-seat the memory and connectors to the discs, motherboard,
optical drives, graphic and sound cards (removes the possibility of 'heat
creep').

As for the outside of the case, I simply wipe over with a dry rag as part of
the above cleaning process.

Some will agree and others disagree with my cleaning procedure, but I still
have fifteen year-old computer running as sweet as a nut with only minimal
replacement of parts (battery and a couple of cables) - which I use to test
any new Application Downloads first.

Cash.

On Tuesday, October 1, 2013 5:02:49 PM UTC+1, John Rumm wrote:
On 01/10/2013 14:12, Lee wrote:

At least you can easily clean the fans and sinks on a desktop, some
laptops are not so straightforward

Well to be fair they are easy enough to clean once you can get to them!
Getting there however is not always straightforward.

This is crude but normally works well enough: put a domestic hoover snout to the air in & outlets alternately repeatedly, 20 times. It sure saves time.


NT

He is obviously using one of those nearly dead keyboards he was talking
about.
actually the sighted do not spot errors the way we blind do as they see
what they expect to see.
One other thing of course is that not everyone is so pedantic as yourself..
grin.

Brian

--
From the Sofa of Brian Gaff Reply address is active
"Adrian" wrote in message
...
On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

teh final tutrine thing is to clean te compoutercase externally and
songe eberything down with a slightly damp cloth.

If you can't be arsed to expend the slightest effort on writing your
posts, why should the rest of us be arsed to expend any on reading them?

On 01/10/13 20:16, Brian Gaff wrote:
He is obviously using one of those nearly dead keyboards he was talking
about.
actually the sighted do not spot errors the way we blind do as they see
what they expect to see.
One other thing of course is that not everyone is so pedantic as yourself..
grin.

Brian

Half th problem is the 'low energy' lightbulb conveniently situated so
the keybaord is in permanent shadow.


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

I just thought I'd start a little thread for people to discuss what te
subject line might mean.

Since SWMBO was out dog-walking, the chance to power down and open the
case was taken. Sheesh. The finned CPU heatsink was solid fluff. I tried
vacuuming, but no joy, and lacking and airline all I could think off was
a household paintbrush.

I clean all of them out annually. Probably a hangover from the
maintenance on our mainframes years ago, when they did regular vacuuming.

Another thing I do on the server, is monitor it for internal disk
errors.

http://www.linuxjournal.com/magazine...rd-disks-smart

Oh yes. All the time.

You may also want to replace mice and keyboards periodically as they do
suffer death - although mice are generally better now they aren't
trackball.

Usually the slidey feet fall off but I have a pack of replacements!

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years

Yup. Do that.

and I would say that 5 years is the limit for a hard
disk drive too. Yes, they may last longer, but that is when the chance
of catastrophic failure starts to rise steeply. Motherboards have been
usasble up to 10 yearas IME, and I only replace them whn I can get more
performance for less or the same money - that used to be three years,
notw it's well over 5..CPU development has it seems slowed
substantially, with all the effort going into either servers with
millions of cores, or portable devices using lower power.

My rule is 50,000 hours.

Suffice to say that irreplaceable data should be held in at least three
places if possible on at least two machines one of which if you REALLY
want to survive a house fire etc, ought not to be in the same house..

Yup. Two on different floors of the house, and one in my work office.



--
Use the BIG mirror service in the UK: http://www.mirrorservice.org
My posts (including this one) are my copyright and if @diy_forums on
Twitter wish to tweet them they can pay me £30 a post
*lightning surge protection* - a w_tom conductor

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:


You may also want to replace mice and keyboards periodically as they do
suffer death - although mice are generally better now they aren't
trackball.

Only ever changed one keyboard. They are all Model Ms.



--
Use the BIG mirror service in the UK: http://www.mirrorservice.org
My posts (including this one) are my copyright and if @diy_forums on
Twitter wish to tweet them they can pay me £30 a post
*lightning surge protection* - a w_tom conductor

On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off. It
had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing". A
quick check revealed that the CMOS settings had been lost. The RTC chip
battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip. Unsoldered
old one, put a socket in, put new chip in and ... nothing. Machine
wouldn't boot. Turned out the chip came "off" and needed an I/O poke to
activate it. Managed to staighten pins on old chip. Put it in. Booted.
Pulled it out while machine was live. Put new chip in, and a couple of
debug instructions and I managed to activate the chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would have
made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

--
Graham.

%Profound_observation%

On 02/10/13 01:10, Graham. wrote:
On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off. It
had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing". A
quick check revealed that the CMOS settings had been lost. The RTC chip
battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip. Unsoldered
old one, put a socket in, put new chip in and ... nothing. Machine
wouldn't boot. Turned out the chip came "off" and needed an I/O poke to
activate it. Managed to staighten pins on old chip. Put it in. Booted.
Pulled it out while machine was live. Put new chip in, and a couple of
debug instructions and I managed to activate the chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would have
made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

I remember those, on an EISA 486 Acer.

Found this innovative solution when I just googled to remind myself
about them:
http://www.classic-computers.org.nz/...ttery-chip.htm

On 01/10/2013 12:29, The Natural Philosopher wrote:
I just thought I'd start a little thread for people to discuss what te
subject line might mean.


It isn't quite regular maintenance but the PSU capacitors on some boards
do fail by going overpressure and take on rakish angles after a while.
The failure mode typically showing up as ram faults and mysterious BSOD
failures with no theme in common.

It is possible to swap these capacitors for similar high ripple types if
you have a steady hand and a fairly beefy soldering iron after a
motherboard has become totally unreliable. Running a Linux memory tester
from a bootable CD is a good way to test for such faults.

(*) You should try removing and reseating the memory first to eliminate
a single block having gone bad. Ditto for the video card although that
usually shows more obvious symptoms like no picture and beeps at bootup.

Feel free to add any other things you consider are useful under the
general heading.

I've left out software upgrades, because these days most people do those
on a regular basis as they become avaliable, esp,.. if they are using
Linux.

I've left out backups as well, because those have been extensively
discussed before.

Suffice to say that irreplaceable data should be held in at least three
places if possible on at least two machines one of which if you REALLY
want to survive a house fire etc, ought not to be in the same house..

And preferably on some different media types since it is possible your
precious written data disk is only readable on a particular drive.

--
Regards,
Martin Brown

On Wed, 02 Oct 2013 01:10:41 +0100, Graham. wrote:

On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off. It
had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing". A
quick check revealed that the CMOS settings had been lost. The RTC chip
battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip. Unsoldered
old one, put a socket in, put new chip in and ... nothing. Machine
wouldn't boot. Turned out the chip came "off" and needed an I/O poke to
activate it. Managed to staighten pins on old chip. Put it in. Booted.
Pulled it out while machine was live. Put new chip in, and a couple of
debug instructions and I managed to activate the chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would have
made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

DS1287 mostly.



--
Use the BIG mirror service in the UK: http://www.mirrorservice.org
My posts (including this one) are my copyright and if @diy_forums on
Twitter wish to tweet them they can pay me £30 a post
*lightning surge protection* - a w_tom conductor

On 02/10/2013 09:58, Chris Bartram wrote:
On 02/10/13 01:10, Graham. wrote:
On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off. It
had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing". A
quick check revealed that the CMOS settings had been lost. The RTC chip
battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip. Unsoldered
old one, put a socket in, put new chip in and ... nothing. Machine
wouldn't boot. Turned out the chip came "off" and needed an I/O poke to
activate it. Managed to staighten pins on old chip. Put it in. Booted.
Pulled it out while machine was live. Put new chip in, and a couple of
debug instructions and I managed to activate the chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would have
made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

I remember those, on an EISA 486 Acer.

Sure it was not a VESA rather than EISA? (i.e. a fast local bus
extension for EISA systems that predated the introduction of PCI)



--
Cheers,

John.

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

On 02/10/2013 17:22, Jethro_uk wrote:
On Wed, 02 Oct 2013 16:05:26 +0100, John Rumm wrote:

On 02/10/2013 09:58, Chris Bartram wrote:
On 02/10/13 01:10, Graham. wrote:
On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off.
It had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing".
A quick check revealed that the CMOS settings had been lost. The RTC
chip battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip.
Unsoldered old one, put a socket in, put new chip in and ... nothing.
Machine wouldn't boot. Turned out the chip came "off" and needed an
I/O poke to activate it. Managed to staighten pins on old chip. Put
it in. Booted. Pulled it out while machine was live. Put new chip in,
and a couple of debug instructions and I managed to activate the
chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would
have made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

I remember those, on an EISA 486 Acer.

Sure it was not a VESA rather than EISA? (i.e. a fast local bus
extension for EISA systems that predated the introduction of PCI)

IIRC VESA was a *video* standard. EISA was an extension to the ISA spec

VESA is/was a video standard - but also a bus extension standard. It
added a third level connector on the board at the other end from the
E/ISA slot(s)

(it used dual level connectors in the board) which allowed an EISA slot
to carry either a normal ISA card, or a fully compliant EISA card, which
needed to be configured in the BIOS. It was quite elegant, and backwardly
compatible too.

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.
Hence it was becoming a serious bottleneck for anything needing fast CPU
/ DMA access - especially video cards. Hence why VESA got involved with
their own local bus to enable fast CPU to video card communications that
were not limited by the ISA bus speeds. (IBM were countering with MCA at
the time, and we know how well that went down!)

It just seems a little odd however going to the effort of describing a
486 machine as being "EISA" - since by that time everything had been
EISA for a long time (i.e. 8 bit only slots had pretty much vanished by
then) - and many fast machines of the day were sporting some form of
additional "local bus". (the VESA bus was also quite closely based on
the native 486 bus standard)

Some pictures of the VESAbus:

http://en.wikipedia.org/wiki/VESA_Local_Bus



--
Cheers,

John.

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

On Wed, 02 Oct 2013 16:05:26 +0100, John Rumm wrote:

On 02/10/2013 09:58, Chris Bartram wrote:
On 02/10/13 01:10, Graham. wrote:
On Tue, 01 Oct 2013 11:45:54 GMT, Jethro_uk
wrote:

On Tue, 01 Oct 2013 12:29:49 +0100, The Natural Philosopher wrote:

Finally, some computers with onboard batteries benfit from a new one
after 5-10 years,

Many years ago, I got a "Memorex Telex" PC from a surplus sell off.
It had an EISA bus, and the RTC chip from TI had an inbuilt battery.

After a few months, went to boot it, and the hard disk was "missing".
A quick check revealed that the CMOS settings had been lost. The RTC
chip battery had died.

After much phoning and faxing, I found a supplier in Brum (I lived in
London) who wholesaled the chip. A train journey and bottle of whisky
later (they couldn't do retail sales) and I got a new chip.
Unsoldered old one, put a socket in, put new chip in and ... nothing.
Machine wouldn't boot. Turned out the chip came "off" and needed an
I/O poke to activate it. Managed to staighten pins on old chip. Put
it in. Booted. Pulled it out while machine was live. Put new chip in,
and a couple of debug instructions and I managed to activate the
chip.

Turned out the reason the machines were "surplus" is that there were
shedloads that had "died". If the internet had been around, I would
have made a fortune repairing them.

Haven't seen many chips with batterys built in recently


They were called "Dallas chips"

I remember those, on an EISA 486 Acer.

Sure it was not a VESA rather than EISA? (i.e. a fast local bus
extension for EISA systems that predated the introduction of PCI)

Some IBM PS/2s had them too. The 55SX for a start.

http://www.tavi.co.uk/ps2pages/battery.html
--
Use the BIG mirror service in the UK: http://www.mirrorservice.org
My posts (including this one) are my copyright and if @diy_forums on
Twitter wish to tweet them they can pay me £30 a post
*lightning surge protection* - a w_tom conductor

On 02/10/2013 17:53, John Rumm wrote:

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.
Hence it was becoming a serious bottleneck for anything needing fast CPU
/ DMA access - especially video cards. Hence why VESA got involved with
their own local bus to enable fast CPU to video card communications that
were not limited by the ISA bus speeds. (IBM were countering with MCA at
the time, and we know how well that went down!)

It just seems a little odd however going to the effort of describing a
486 machine as being "EISA" - since by that time everything had been
EISA for a long time (i.e. 8 bit only slots had pretty much vanished by
then) - and many fast machines of the day were sporting some form of
additional "local bus". (the VESA bus was also quite closely based on
the native 486 bus standard)

You sure about all that? Because IIRC EISA only ever appeared on a few
servers, had a config setting per interrupt to go level triggered, and
never took off on desktops because PCI came along.

That Acer machine probably had some of my code in it - we (ICL) did
deals with Acer, and someof the bugfixes were fed back into their products.

Andy

On Wed, 02 Oct 2013 17:53:24 +0100, John Rumm wrote:

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.

Sorry, John, but that wasn't EISA. It was just known as 16 bit slots for
a while until someone invented the term ISA to describe both the 8 and 16
bit slots - '8 bit ISA' and '16 bit ISA' if you like. The 16 bit slots
appeared firts in the PC/AT around 1984.

EISA was a (mainly server only) much later (1988) architecture, produced
to compete with MCA, which was IBM proprietary. MCA had level triggered
interrupts (hence IRQ sharing) as well as a wider bus, some clever tricks
to use the address bus to widen the data bus, semi automatic 'plug and
play', etc. EISA had most of that, for the rest of the industry. It
didn't last long partly due to the restricted market, so not many cards
were made.

It just seems a little odd however going to the effort of describing a
486 machine as being "EISA" - since by that time everything had been
EISA for a long time (i.e. 8 bit only slots had pretty much vanished by
then)

No - see above. EISA machines were fairly rare, so the effort *was* worth
it!


--
Use the BIG mirror service in the UK: http://www.mirrorservice.org
My posts (including this one) are my copyright and if @diy_forums on
Twitter wish to tweet them they can pay me £30 a post
*lightning surge protection* - a w_tom conductor

On 05/10/13 21:31, Vir Campestris wrote:
On 02/10/2013 17:53, John Rumm wrote:

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.
Hence it was becoming a serious bottleneck for anything needing fast CPU
/ DMA access - especially video cards. Hence why VESA got involved with
their own local bus to enable fast CPU to video card communications that
were not limited by the ISA bus speeds. (IBM were countering with MCA at
the time, and we know how well that went down!)

It just seems a little odd however going to the effort of describing a
486 machine as being "EISA" - since by that time everything had been
EISA for a long time (i.e. 8 bit only slots had pretty much vanished by
then) - and many fast machines of the day were sporting some form of
additional "local bus". (the VESA bus was also quite closely based on
the native 486 bus standard)

You sure about all that? Because IIRC EISA only ever appeared on a few
servers, had a config setting per interrupt to go level triggered, and
never took off on desktops because PCI came along.

Not sure your memory IS correct..there was ISA and then a enhanced ISA -
fairly sure we sold Ethernet cards with that for a year or two - and
then as you say PCI came along.


Mmm. the truth is somewhere between our memories it seems. It was a bit
more common than you say, but less common than I remembered..

http://en.wikipedia.org/wiki/Extende...d_Architecture


That Acer machine probably had some of my code in it - we (ICL) did
deals with Acer, and someof the bugfixes were fed back into their products.

Andy


--
Ineptocracy

(in-ep-toc-ra-cy) €“ a system of government where the least capable to
lead are elected by the least capable of producing, and where the
members of society least likely to sustain themselves or succeed, are
rewarded with goods and services paid for by the confiscated wealth of a
diminishing number of producers.

On 05/10/2013 21:31, Vir Campestris wrote:
On 02/10/2013 17:53, John Rumm wrote:

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.
Hence it was becoming a serious bottleneck for anything needing fast CPU
/ DMA access - especially video cards. Hence why VESA got involved with
their own local bus to enable fast CPU to video card communications that
were not limited by the ISA bus speeds. (IBM were countering with MCA at
the time, and we know how well that went down!)

It just seems a little odd however going to the effort of describing a
486 machine as being "EISA" - since by that time everything had been
EISA for a long time (i.e. 8 bit only slots had pretty much vanished by
then) - and many fast machines of the day were sporting some form of
additional "local bus". (the VESA bus was also quite closely based on
the native 486 bus standard)

You sure about all that? Because IIRC EISA only ever appeared on a few
servers, had a config setting per interrupt to go level triggered, and
never took off on desktops because PCI came along.

Perhaps we are talking about different things, but I have always
considered ISA the original 6/8MHz 8 bit bus as per the IBM PC and
PC-XT, and the Extended (E)ISA as the 16 bit version with the extra slot
introduced with the PC-AT...

Ah ok, done some hunting about, EISA does indeed get used more than one
way. My apologies... Seems the one you are describing is a later thing -
contemporary with (or just before) VESA bus, and more commonly found in
servers.

In fact thinking about it, I must have used a system based on it at some
point since I recall using a configuration utility to configure
expansion cards on it.

That Acer machine probably had some of my code in it - we (ICL) did
deals with Acer, and someof the bugfixes were fed back into their products.



--
Cheers,

John.

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

On 05/10/2013 21:49, Bob Eager wrote:
On Wed, 02 Oct 2013 17:53:24 +0100, John Rumm wrote:

Yup EISA came in with the 286 PC-AT - basically adding the extra
connector beside the original ISA slot. It was still notionally limited
to about 8MHz bus speed (although some BIOSes allowed you to push it a
bit) with the same crippled edge triggered interrupt mechanism of ISA.

Sorry, John, but that wasn't EISA. It was just known as 16 bit slots for
a while until someone invented the term ISA to describe both the 8 and 16
bit slots - '8 bit ISA' and '16 bit ISA' if you like. The 16 bit slots
appeared firts in the PC/AT around 1984.

Yup, fairy snuff - I think I am remembering EISA being used (perhaps
incorrectly, or unofficially) from the mid 80's. So for the purposes of
this discussion, ignore me ;-)



--
Cheers,

John.

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

On 06/10/2013 06:13, John Rumm wrote:

Yup, fairy snuff - I think I am remembering EISA being used (perhaps
incorrectly, or unofficially) from the mid 80's. So for the purposes of
this discussion, ignore me ;-)

Major brownie points there for actually admitting to a mistake. We're
all here to learn.

No, scratch that last. Some of us are here to learn, others just for the
arguments

Andy

On Sat, 05 Oct 2013 21:56:40 +0100, The Natural Philosopher
wrote:

Mmm. the truth is somewhere between our memories it seems. It was a bit
more common than you say, but less common than I remembered..

http://en.wikipedia.org/wiki/Extende...d_Architecture

I was a bit ****ed off at having spent some serious wedge (for me at
tht time) on an EISA sound card, which was superceded very shortly
afterwards by the PCI version and by the makers dropping virtuall all
support for the older ones within another couple of years.
I still have it, it still sounds great, but using it means I have to
keep an ancient motherboard up and running. If I'd known about PCI
taking over the world at the time, I'd have waited a bit, and my PCI
version would likely still be running easily enough (voltage
variations allowed for, of course).