Lenovo T43 (with XP) Wont Power On

I bought this Lenovo T43 laptop as a refurbished computer years ago.
It's been reliable, except about a year after I bought it, I had to
replace the system CMOS battery. I've also re-installed XP due to
malware issues.

I only use this computer about once a week, to go to a WIFI place. (I
only have dialup internet at home).

A few weeks ago, I powered up the T43 and got a CMOS error telling me to
set the clock. I set the clock time and date and it worked fine after
that. I assumed the CMOS battery was dying again, and ordered one
online. I've gotten the battery, but not yet changed it, since I cant
remember how to get to it. (I removed the keyboard, and can see the
battery next to the mouse, but that cover has to come off to get to the
battery and I dont recall which screws take that cover off).

Anyhow, I used the computer a few times since that "clock failure" with
no problems. (I usually keep the charger plugged in, which I suppose
keeps the clock set).

Yesterday I pushed the power button and nothing happened. I put a meter
on the "main" battery terminals, and am showing 12volts. I also tested
the charger and it's putting out voltage (around 14v). So, I know it's
getting power.

I removed the hard drive and CD drive. It still wont power up from
either the battery or the power supply. All I get, is a very brief flash
of the battery indicator light when I push the POWER button, and that
only occurs if the battery was removed from the computer.

I looked on some websites and found that I'm supposed to remove static,
but pushing the power button 10x, then holding it for 30sec. (with
battery and power cord removed). Then install the battery and it's
supposed to work. That has made no difference. It still wont power up,
and I've done this at least 20 times.

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

In message ,
writes:
[]
I removed the hard drive and CD drive. It still wont power up from
either the battery or the power supply. All I get, is a very brief flash
of the battery indicator light when I push the POWER button, and that
only occurs if the battery was removed from the computer.
[]
Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?
[]
I don't know that particular model, but have seen similar on others: I
suspect the power-in socket's solder joints (or one of them) have
failed, with the result that power is reaching neither the main computer
nor the battery (which has run down). A very high-impedance connection
remains, giving you the brief flashes, and making the 12 volts you
measured at the battery terminals (do you still get that if you put say
100 ohms across it?).

No _easy_ way to verify this; waggling the power connector _might_ show
some result. If you can charge the main battery in another machine, that
would probably prove it.

If this _is_ the problem, it's trivial to fix - you either replace the
power-in socket, or more likely just reflow the solder connections on
the existing one - once you've actually got to it, which involves, I'm
afraid, a significant amount of dismantlement of the laptop. Whether
that's worth it or not, I leave to you - for me, it would be (and was),
just for the frustration of such a simple fault laying low a computer,
but in terms of the value of the computer as against your time to do it
viewed coldly, it might not be.

Of course, that might _not_ be the fault.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

If you carry on hating, you're the one who's damaged.
- Sir Harold Atcherley, sent to the Burma/Siam railway in April 1943

On 8/15/2016 3:57 PM, wrote:
I bought this Lenovo T43 laptop as a refurbished computer years ago.
It's been reliable, except about a year after I bought it, I had to
replace the system CMOS battery. I've also re-installed XP due to
malware issues.

I only use this computer about once a week, to go to a WIFI place. (I
only have dialup internet at home).

A few weeks ago, I powered up the T43 and got a CMOS error telling me to
set the clock. I set the clock time and date and it worked fine after
that. I assumed the CMOS battery was dying again, and ordered one
online. I've gotten the battery, but not yet changed it, since I cant
remember how to get to it. (I removed the keyboard, and can see the
battery next to the mouse, but that cover has to come off to get to the
battery and I dont recall which screws take that cover off).

Anyhow, I used the computer a few times since that "clock failure" with
no problems. (I usually keep the charger plugged in, which I suppose
keeps the clock set).

Yesterday I pushed the power button and nothing happened. I put a meter
on the "main" battery terminals, and am showing 12volts. I also tested
the charger and it's putting out voltage (around 14v). So, I know it's
getting power.

I removed the hard drive and CD drive. It still wont power up from
either the battery or the power supply. All I get, is a very brief flash
of the battery indicator light when I push the POWER button, and that
only occurs if the battery was removed from the computer.

I looked on some websites and found that I'm supposed to remove static,
but pushing the power button 10x, then holding it for 30sec. (with
battery and power cord removed). Then install the battery and it's
supposed to work. That has made no difference. It still wont power up,
and I've done this at least 20 times.

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

Jonas:

The service manual can be found he

http://download.lenovo.com/ibmdl/pub...df/39t6193.pdf

There should be a description of getting the CMOS battery replaced. Might even
have a troubleshooting section.

HTH & GL

John

--
\\\||///
------------------o000----(o)(o)----000o----------------
----------------------------()--------------------------
'' Madness takes its toll - Please have exact change. ''

wrote:
I bought this Lenovo T43 laptop as a refurbished computer years ago.
It's been reliable, except about a year after I bought it, I had to
replace the system CMOS battery. I've also re-installed XP due to
malware issues.

I only use this computer about once a week, to go to a WIFI place. (I
only have dialup internet at home).

A few weeks ago, I powered up the T43 and got a CMOS error telling me to
set the clock. I set the clock time and date and it worked fine after
that. I assumed the CMOS battery was dying again, and ordered one
online. I've gotten the battery, but not yet changed it, since I cant
remember how to get to it. (I removed the keyboard, and can see the
battery next to the mouse, but that cover has to come off to get to the
battery and I dont recall which screws take that cover off).

Anyhow, I used the computer a few times since that "clock failure" with
no problems. (I usually keep the charger plugged in, which I suppose
keeps the clock set).

Yesterday I pushed the power button and nothing happened. I put a meter
on the "main" battery terminals, and am showing 12volts. I also tested
the charger and it's putting out voltage (around 14v). So, I know it's
getting power.

I removed the hard drive and CD drive. It still wont power up from
either the battery or the power supply. All I get, is a very brief flash
of the battery indicator light when I push the POWER button, and that
only occurs if the battery was removed from the computer.

I looked on some websites and found that I'm supposed to remove static,
but pushing the power button 10x, then holding it for 30sec. (with
battery and power cord removed). Then install the battery and it's
supposed to work. That has made no difference. It still wont power up,
and I've done this at least 20 times.

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

You *must* replace the CMOS battery.

After that, everything will be fine.

It'll probably take a little searching to find
a decent keyboard removal procedure. Sample here.

https://support.lenovo.com/ca/en/documents/migr-46515

*******

Some SuperI/O chips, if the CMOS battery drops to zero
volts, the SuperI/O will prevent the computer from starting.
Even though the main battery is full. That's why the
CMOS coin cell needs to be replaced.

Paul

wrote:
I bought this Lenovo T43 laptop as a refurbished computer years ago.
It's been reliable, except about a year after I bought it, I had to
replace the system CMOS battery. I've also re-installed XP due to
malware issues.

I only use this computer about once a week, to go to a WIFI place. (I
only have dialup internet at home).

A few weeks ago, I powered up the T43 and got a CMOS error telling me to
set the clock. I set the clock time and date and it worked fine after
that. I assumed the CMOS battery was dying again, and ordered one
online. I've gotten the battery, but not yet changed it, since I cant
remember how to get to it. (I removed the keyboard, and can see the
battery next to the mouse, but that cover has to come off to get to the
battery and I dont recall which screws take that cover off).

Anyhow, I used the computer a few times since that "clock failure" with
no problems. (I usually keep the charger plugged in, which I suppose
keeps the clock set).

Yesterday I pushed the power button and nothing happened. I put a meter
on the "main" battery terminals, and am showing 12volts. I also tested
the charger and it's putting out voltage (around 14v). So, I know it's
getting power.

I removed the hard drive and CD drive. It still wont power up from
either the battery or the power supply. All I get, is a very brief flash
of the battery indicator light when I push the POWER button, and that
only occurs if the battery was removed from the computer.

I looked on some websites and found that I'm supposed to remove static,
but pushing the power button 10x, then holding it for 30sec. (with
battery and power cord removed). Then install the battery and it's
supposed to work. That has made no difference. It still wont power up,
and I've done this at least 20 times.

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

Very easy to change the battery. Remove the kb and about 14 screws from
underneath. Use a thin knife to pop the top plastic bezel off of the retainers.

snip
Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

Very easy to change the battery. Remove the kb and about 14 screws from
underneath. Use a thin knife to pop the top plastic bezel off of the
retainers.

I know you don't want to watch video, this youtube video claims to apply
to the T41-T43 and shows him removing the kbd and palm rest section of
the top, exposing the battery. You remove several bottom screws to do
that, and there's one screw under a small plastic cover on the bottom
that had to be removed, so the plastic cover has to be pried out, it's
near the bottom left corner of the case when the computer us lying on
the top lid (upside down) with the hinge away from you. The small
plastic cover is maybe 1/2 the size of your smallest finger fingerprint
and about that shape.
https://www.youtube.com/watch?v=8ZI-8Q2j-zQ

On 8/16/2016 2:16 AM, Mike S wrote:
snip
Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

Very easy to change the battery. Remove the kb and about 14 screws from
underneath. Use a thin knife to pop the top plastic bezel off of the
retainers.

I know you don't want to watch video, this youtube video claims to apply
to the T41-T43 and shows him removing the kbd and palm rest section of
the top, exposing the battery. You remove several bottom screws to do
that, and there's one screw under a small plastic cover on the bottom
that had to be removed, so the plastic cover has to be pried out, it's
near the bottom left corner of the case when the computer us lying on
the top lid (upside down) with the hinge away from you. The small
plastic cover is maybe 1/2 the size of your smallest finger fingerprint
and about that shape.
https://www.youtube.com/watch?v=8ZI-8Q2j-zQ

This repair is for a T41 but it looks similar to what was described in
the above video. You might end up taking a few more screws out than
needed but you should be able to do the repair after seeing this.
https://www.ifixit.com/Guide/IBM+Thi...placement/2916

On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

You *must* replace the CMOS battery.

After that, everything will be fine.

It'll probably take a little searching to find
a decent keyboard removal procedure. Sample here.

https://support.lenovo.com/ca/en/documents/migr-46515

*******

Some SuperI/O chips, if the CMOS battery drops to zero
volts, the SuperI/O will prevent the computer from starting.
Even though the main battery is full. That's why the
CMOS coin cell needs to be replaced.

Paul

I guess the laptops are different in this sense from desktop computers.
When the coin cell dies in a desktop it still powers up, but just goes
to the bios settings, which need to be setup on every boot.
I hope this coin cell solves this. It was a nice computer, and is my
only way to get online with WIFI, since I cant get high speed at home.

On Tue, 16 Aug 2016 02:16:17 -0700, Mike S wrote:

snip
Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.

Very easy to change the battery. Remove the kb and about 14 screws from
underneath. Use a thin knife to pop the top plastic bezel off of the
retainers.

I know you don't want to watch video, this youtube video claims to apply
to the T41-T43 and shows him removing the kbd and palm rest section of
the top, exposing the battery. You remove several bottom screws to do
that, and there's one screw under a small plastic cover on the bottom
that had to be removed, so the plastic cover has to be pried out, it's
near the bottom left corner of the case when the computer us lying on
the top lid (upside down) with the hinge away from you. The small
plastic cover is maybe 1/2 the size of your smallest finger fingerprint
and about that shape.
https://www.youtube.com/watch?v=8ZI-8Q2j-zQ



I'd like to watch the video, but it's impossible on dialup. I dont have
another laptop to use for public WIFI.

I did get the cover off (thanks to all the help from you guys). There
were actually 3 of those fingernail shaped plastic covers that were
glued on over the screws. They serve no purpose so they will stay off
now. There were about 14 screws total. One was stubborn, and when the
palm rest cover finally came off, that one got launched and hit the
wall, never to be seen again. I doubt the hardware store sells such tiny
screws, but I'm sure I can leave one out without any problems.

With any luck, I'll get that CMOS battery in the mail today and be able
to get this computer working again.

It sure would be nice if they made it easier to change that coin
cell.... Like a simple little door held on by one screw.....

wrote:
On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

Maybe this computer is "toast", but I cant understand why it would just
quit working while it was not being used, but was plugged in to the
charger. (I have always left it plugged in most of the time when not in
use).

Can anyone suggest anything else to try or do?

Also, does anyone have a photo showing how to remove that top panel, to
access the CMOS battery.
Note: I can not watch online videos on dialup. but should be able to see
any photo.
You *must* replace the CMOS battery.

After that, everything will be fine.

It'll probably take a little searching to find
a decent keyboard removal procedure. Sample here.

https://support.lenovo.com/ca/en/documents/migr-46515

*******

Some SuperI/O chips, if the CMOS battery drops to zero
volts, the SuperI/O will prevent the computer from starting.
Even though the main battery is full. That's why the
CMOS coin cell needs to be replaced.

Paul

I guess the laptops are different in this sense from desktop computers.
When the coin cell dies in a desktop it still powers up, but just goes
to the bios settings, which need to be setup on every boot.
I hope this coin cell solves this. It was a nice computer, and is my
only way to get online with WIFI, since I cant get high speed at home.


I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.

In some cases, it is absolutely worth while taking
care of your CR2032, because the battery is on the
solder side of the mainboard, and almost impossible
to reach. One clever poster here, used a hole cut saw and
made a 1" hole in the bottom of an old laptop, just
to have quick access :-) Which is certainly better than
removing 30 screws and having to take the thing entirely
apart.

You would think they would arrange one of those access
doors, to expose the coin cell, but that isn't always
the case. In some cases, it's a complete teardown.
It's like the car design, where you have to pull
the engine up out of the engine bay, just to
change the fourth spark plug. And as a result,
platinum spark plugs are used. Because the labor cost
is so high to change the plugs, it's worth it to
take care of them and stretch the service life.

Paul

On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.

Coming into this thread a bit late...

Is this the common scenario for a laptop, that if you remove the laptop
battery, the CR2032 will run the CMOS clock even if the laptop is plugged
into the mains? Or is this more the exception, than the rule?

Just would like to know, as I thought you could use a laptop without a
battery if left plugged into the mains (like a small desktop), if so
desired.

Bill in Co wrote:
On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.

Coming into this thread a bit late...

Is this the common scenario for a laptop, that if you remove the laptop
battery, the CR2032 will run the CMOS clock even if the laptop is plugged
into the mains? Or is this more the exception, than the rule?

Just would like to know, as I thought you could use a laptop without a
battery if left plugged into the mains (like a small desktop), if so
desired.



The general principle is one of "ORing logic".

If multiple power sources are available, they connect with
forward conducting diodes to the load. If all the conventional
power sources are removed, and only the CR2032 is present,
the CR2032 powers the CMOS RAM and RTC clock. The higher power
sources would run at a slightly higher voltage, causing the
diode on the CR2032 to be reverse biased. The diode is then
not conducting. This is important, because the CR2032 is
not rated for charging. You are only allowed a 1uA current
flow back into the CR2032, which defines the diode type
used for the ORing logic. I have three computers here,
that I've verified a BAT45 dual diode is used (it's a three
lead device that looks like an SMT transistor). You need
a magnifying glass to read the logo.

The other kind of coin cell, is an LR2032. It is a coin
cell that supports charging, and is a rechargeable battery.
It is not interchangeable with the CR2032. The LR2032
charges from whatever main power source is present, but
the capacity of that coin cell is a lot less than a CR2032,
and it cannot run the clock for as long. So if you find an
LR2032, then the analysis ("what's broke") might be quite
different. As one reason for an LR2032 to break, would be
a charging circuit failure (the thing that charges the LR2032).
Logically, the circuit design is the same (ORing of power),
but the implementation would be different. I've not seen
a schematic for an LR2032 based design.

Paul

Paul wrote:
Bill in Co wrote:
On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.

Coming into this thread a bit late...

Is this the common scenario for a laptop, that if you remove the laptop
battery, the CR2032 will run the CMOS clock even if the laptop is plugged
into the mains? Or is this more the exception, than the rule?

Just would like to know, as I thought you could use a laptop without a
battery if left plugged into the mains (like a small desktop), if so
desired.



The general principle is one of "ORing logic".

If multiple power sources are available, they connect with
forward conducting diodes to the load. If all the conventional
power sources are removed, and only the CR2032 is present,
the CR2032 powers the CMOS RAM and RTC clock. The higher power
sources would run at a slightly higher voltage, causing the
diode on the CR2032 to be reverse biased. The diode is then
not conducting. This is important, because the CR2032 is
not rated for charging. You are only allowed a 1uA current
flow back into the CR2032, which defines the diode type
used for the ORing logic. I have three computers here,
that I've verified a BAT45 dual diode is used (it's a three
lead device that looks like an SMT transistor). You need
a magnifying glass to read the logo.

The other kind of coin cell, is an LR2032. It is a coin
cell that supports charging, and is a rechargeable battery.
It is not interchangeable with the CR2032. The LR2032
charges from whatever main power source is present, but
the capacity of that coin cell is a lot less than a CR2032,
and it cannot run the clock for as long. So if you find an
LR2032, then the analysis ("what's broke") might be quite
different. As one reason for an LR2032 to break, would be
a charging circuit failure (the thing that charges the LR2032).
Logically, the circuit design is the same (ORing of power),
but the implementation would be different. I've not seen
a schematic for an LR2032 based design.

Paul

Interesting! Let's see if I understood this fully...

So I guess this means that (as a generality), if you intend to use a laptop
as a mini-desktop, without ever using or replacing the laptop battery, you'd
better keep it always plugged into the mains, or like you said, the CMOS RAM
and RTC will drain the CMOS battery (typically a CR2032, I presume).

But even that is assuming that when the laptop is plugged into the mains and
NOT turned on it is running the CMOS from the mains, and not using the
CR2032. (is that correct)

However, I think you said that was only a 10 ua draw, anyways, which could
last perhaps 3 years? Still, something to be aware of. :-)

Bill in Co wrote:
Paul wrote:
Bill in Co wrote:
On Mon, 15 Aug 2016 22:57:17 -0400, Paul wrote:

I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.
Coming into this thread a bit late...

Is this the common scenario for a laptop, that if you remove the laptop
battery, the CR2032 will run the CMOS clock even if the laptop is plugged
into the mains? Or is this more the exception, than the rule?

Just would like to know, as I thought you could use a laptop without a
battery if left plugged into the mains (like a small desktop), if so
desired.


The general principle is one of "ORing logic".

If multiple power sources are available, they connect with
forward conducting diodes to the load. If all the conventional
power sources are removed, and only the CR2032 is present,
the CR2032 powers the CMOS RAM and RTC clock. The higher power
sources would run at a slightly higher voltage, causing the
diode on the CR2032 to be reverse biased. The diode is then
not conducting. This is important, because the CR2032 is
not rated for charging. You are only allowed a 1uA current
flow back into the CR2032, which defines the diode type
used for the ORing logic. I have three computers here,
that I've verified a BAT45 dual diode is used (it's a three
lead device that looks like an SMT transistor). You need
a magnifying glass to read the logo.

The other kind of coin cell, is an LR2032. It is a coin
cell that supports charging, and is a rechargeable battery.
It is not interchangeable with the CR2032. The LR2032
charges from whatever main power source is present, but
the capacity of that coin cell is a lot less than a CR2032,
and it cannot run the clock for as long. So if you find an
LR2032, then the analysis ("what's broke") might be quite
different. As one reason for an LR2032 to break, would be
a charging circuit failure (the thing that charges the LR2032).
Logically, the circuit design is the same (ORing of power),
but the implementation would be different. I've not seen
a schematic for an LR2032 based design.

Paul

Interesting! Let's see if I understood this fully...

So I guess this means that (as a generality), if you intend to use a laptop
as a mini-desktop, without ever using or replacing the laptop battery, you'd
better keep it always plugged into the mains, or like you said, the CMOS RAM
and RTC will drain the CMOS battery (typically a CR2032, I presume).

But even that is assuming that when the laptop is plugged into the mains and
NOT turned on it is running the CMOS from the mains, and not using the
CR2032. (is that correct)

However, I think you said that was only a 10 ua draw, anyways, which could
last perhaps 3 years? Still, something to be aware of. :-)

Analyze all the "main" power sources. Something
from there can be used to run the CMOS RAM and
RTC clock.

It's when all the identifiable primary sources
are not present, then the diode from the CR2032
is in a conducting state. And the CR2032 runs
the CMOS RAM and RTC clock.

Your digital watch needs around 2uA to run a
32768 Hertz oscillator. The 32768 Hertz oscillator
in the computer, is on the Southbridge. The Southbridge
houses the 256 byte CMOS RAM and the RTC. The RTC uses
a ripple counter to divide the 32768 Hertz signal down
to 1 Hz (for time-keeping purposes).

The area in the Southbridge, is located in the CMOS "well".
All the control signals leading into that circuit
section, use transmission gates. That prevents
power leakage from the well, into the rest of the
Southbridge. Without leakage protection, the load on
the CR2032 would be more than 10uA.

The CR2032 lasts a bit less than 3 years, if asked
to provide 10uA on a continuous basis. If one of the
other "main" power sources is available, the CR2032 could
easily last more than 10 years.

And computers have used more ungainly power sources
for the same purpose. My Mac G4 uses a much larger
pen cell format battery, which would probably
cost $20 to replace. Since the computer still starts,
I'm not in a hurry to open it up and check (I still
use the computer to run my SCSI scanner). My main
motivation for replacing pesky batteries like that,
is if they stop a computer dead. None of my PCs here,
have that issue with the CR2032 - none of the SuperI/O
seem to have the issue with VBAT = 0 volts.

Paul

On Tue, 16 Aug 2016 07:24:54 -0400, Paul wrote:

I've never seen an explanation for the "why" of it.

It just seems that certain models of SuperI/O chips
have a VBAT input, and the terminal does not seem
to be connected to the hardware monitor. So it's connected
for some other reason. And I don't think I was able to
find a line of text in the datasheet, that said "computer
won't start if VBAT=0 volts". But those are the symptoms.

The symptoms go "one day the clock loses setting or the
other BIOS settings are corrupted". Then "four weeks
later, the computer will not start - a check of the
CMOS coin cell CR2032 reveals the terminal voltage
is then very close to zero volts". So the voltage
drops from 2.4V to 0V on a CR2032, under a 10uA load,
and it takes 4 weeks.

Removing the main battery pack, causes the CR2032
to run the CMOS clock. Which accelerates the demise
of the CR2032. It has a slightly less than 3 year
rating, if you keep pulling the main battery pack
out of the laptop after using it.

In some cases, it is absolutely worth while taking
care of your CR2032, because the battery is on the
solder side of the mainboard, and almost impossible
to reach. One clever poster here, used a hole cut saw and
made a 1" hole in the bottom of an old laptop, just
to have quick access :-) Which is certainly better than
removing 30 screws and having to take the thing entirely
apart.

You would think they would arrange one of those access
doors, to expose the coin cell, but that isn't always
the case. In some cases, it's a complete teardown.
It's like the car design, where you have to pull
the engine up out of the engine bay, just to
change the fourth spark plug. And as a result,
platinum spark plugs are used. Because the labor cost
is so high to change the plugs, it's worth it to
take care of them and stretch the service life.

Paul

Everything you said makes sense, however in my case, I think the
computer is toast. I replaced the CMOS battery with a new one. Nothing
changed. I did the hold the power button for 30 seconds after pushing it
10 times, and I even tried to power up without the RAM. NOTHING....

I still get that fraction of a second flash on the battery light, but
only after the main battery was removed. With or without the AC adaptor
plugged in, and it dont indicate any charging of the battery lights or
anything else.

I dont see anything that appears burned, cracked, or broken, but the
parts are so damn small and my eyes not the best anymore, so what is
there to see. I've done all I can do with limited test equipment and
cant find anything else to try. I guess this computer is junk.

Maybe I can use someone elses computer and find another T43 on Ebay, and
then all I have to do is put in my present harddrive and will be back in
business. This is the reason I prefer desktop machines. I can always fix
them, but they are too big to haul around to places that have WIFI.
Laptops are not made to be user repaired. I guess they are just
disposible once they fail.

I cant afford a new laptop, and I dont want any OS newer than XP anyhow,
so I may as well just find another identical computer. These T43's are
about 10 years old, so I'm sure I can get one pretty cheap.