Could this be the battery? Appreciate any insights

[Hybyd2]

This is a P3 733 MHz with an P34VX. The computer is on 24x7 - only the
monitor gets turned off when not in use, the disks are on and the system
does not go into standby or hibernate. One of the reason it's on 24x7 is
that if turned off, turning it on is a problem. And, when one succeeds in
turning it goes into the BIOS with the CPU speed set to Manual and a warning
that the CPU has been changed so its requesting the speed be set. I set it
at 733 and then all is well. Also, I've noticed the clock will need to be
reset once it goes into Windows as the time goes back to 1999 or some such
year. Could run for days / weeks until for whatever reason it gets turned
off. Because of the flaky behavior I've connected it to a UPS so your
normal brownout has no effect.

So what happens when it gets turned off? Typically, pressing the power
button has no response at all. It's totally dead. I then typically flip it
on it's side and open the case. There have been times when with the case
open pressing the power button has brought it back to life. It happened
again earlier this evening. With the computer open on it's side I actually
had to unplug it twice and replug it before it sprung to life. Some flaky
capacitor on the mother-board? Bad battery? If it's a bad battery,
wouldn't the only problem be that it wouldn't remember any bios settings?
Why would it refuse to start until I jumped through some still to be
determined mysterious set of steps.

Any insights into what the problem is would be greatly appreciated.

Thanks,

Henry

 

[Michael W. Ryder]

This is a P3 733 MHz with an P34VX. The computer is on 24x7 - only the
monitor gets turned off when not in use, the disks are on and the system
does not go into standby or hibernate. One of the reason it's on 24x7 is
that if turned off, turning it on is a problem. And, when one succeeds in
turning it goes into the BIOS with the CPU speed set to Manual and a warning
that the CPU has been changed so its requesting the speed be set. I set it
at 733 and then all is well. Also, I've noticed the clock will need to be
reset once it goes into Windows as the time goes back to 1999 or some such
year. Could run for days / weeks until for whatever reason it gets turned
off. Because of the flaky behavior I've connected it to a UPS so your
normal brownout has no effect.

So what happens when it gets turned off? Typically, pressing the power
button has no response at all. It's totally dead. I then typically flip it
on it's side and open the case. There have been times when with the case
open pressing the power button has brought it back to life. It happened
again earlier this evening. With the computer open on it's side I actually
had to unplug it twice and replug it before it sprung to life. Some flaky
capacitor on the mother-board? Bad battery? If it's a bad battery,
wouldn't the only problem be that it wouldn't remember any bios settings?
Why would it refuse to start until I jumped through some still to be
determined mysterious set of steps.

Any insights into what the problem is would be greatly appreciated.

Thanks,

Henry

Have you checked the power supply? I had similar problems getting a
computer to restart until I replaced the generic power supply with a PC
Power and Cooling one. No problems after that.

 

[JAD]

lol I have a PIII slot 1 850 in a P3V4x 'wacky' board, but a great
workhorse. same sort of situation except off on weekends. with a
windows9x Os i would occasionally get the cpu setup after a weird
shutdown (hangs etc) and the clock was always a little off. I think it
was a winbond thing. your problem seems to be cmos battery as the
problems seem to be chronic,every boot?

 

[Matt]

Any insights into what the problem is would be greatly appreciated.

A standard battery costs about $2.

 

[David Maynard]

This is a P3 733 MHz with an P34VX. The computer is on 24x7 - only the
monitor gets turned off when not in use, the disks are on and the system
does not go into standby or hibernate. One of the reason it's on 24x7 is
that if turned off, turning it on is a problem. And, when one succeeds in
turning it goes into the BIOS with the CPU speed set to Manual and a warning
that the CPU has been changed so its requesting the speed be set. I set it
at 733 and then all is well. Also, I've noticed the clock will need to be
reset once it goes into Windows as the time goes back to 1999 or some such
year. Could run for days / weeks until for whatever reason it gets turned
off. Because of the flaky behavior I've connected it to a UPS so your
normal brownout has no effect.

So what happens when it gets turned off? Typically, pressing the power
button has no response at all. It's totally dead. I then typically flip it
on it's side and open the case. There have been times when with the case
open pressing the power button has brought it back to life. It happened
again earlier this evening. With the computer open on it's side I actually
had to unplug it twice and replug it before it sprung to life. Some flaky
capacitor on the mother-board? Bad battery? If it's a bad battery,
wouldn't the only problem be that it wouldn't remember any bios settings?
Why would it refuse to start until I jumped through some still to be
determined mysterious set of steps.

Any insights into what the problem is would be greatly appreciated.

Thanks,

Henry

Dead CMOS battery. Replace it.

THEN see if there are 'other problems'.

 

[Paul]

This is a P3 733 MHz with an P34VX. The computer is on 24x7 - only the
monitor gets turned off when not in use, the disks are on and the system
does not go into standby or hibernate. One of the reason it's on 24x7 is
that if turned off, turning it on is a problem. And, when one succeeds in
turning it goes into the BIOS with the CPU speed set to Manual and a warning
that the CPU has been changed so its requesting the speed be set. I set it
at 733 and then all is well. Also, I've noticed the clock will need to be
reset once it goes into Windows as the time goes back to 1999 or some such
year. Could run for days / weeks until for whatever reason it gets turned
off. Because of the flaky behavior I've connected it to a UPS so your
normal brownout has no effect.

So what happens when it gets turned off? Typically, pressing the power
button has no response at all. It's totally dead. I then typically flip it
on it's side and open the case. There have been times when with the case
open pressing the power button has brought it back to life. It happened
again earlier this evening. With the computer open on it's side I actually
had to unplug it twice and replug it before it sprung to life. Some flaky
capacitor on the mother-board? Bad battery? If it's a bad battery,
wouldn't the only problem be that it wouldn't remember any bios settings?
Why would it refuse to start until I jumped through some still to be
determined mysterious set of steps.

Any insights into what the problem is would be greatly appreciated.

Thanks,

Henry

You could possible have one or two problems.

Try another power supply, as your current one could be having
a problem with +5VSB.

As for the time being off, have you ever cleared the CMOS on
this board, without switching off/disconnecting the PSU from
the wall ? Sometimes the ORing diode gets damaged, when the
CLRTC jumper shorts the +3V supply, and then, even if the
battery is good (reads about 3V), the clock could have trouble.
The ORing diode is a dual diode in a tiny three legged package
that looks like a transistor. Normally, this device is next
to the CMOS battery. One poster fixed his, by replacing the
three legged device, with a couple of 1N4148 type diodes. The
part typically has "K45" printed on it, in letters that can
only be read with a magnifying glass.

As far as diagnosing a bad dual diode, using a multimeter
to check the forward and reverse characteristics of the
dual diode would be one way to check it. Checking the battery
itself is pretty easy with a multimeter - it should read
3V when new, and the CMOS might become "forgetful" at around
2.4V (allowing 0.4V drop across the diode, and a minimum of
2.0V to run the Southbridge). The battery is only used, when
+5VSB is not available on the motherboard - the green
LED on an Asus motherboard is usually an indicator of the presence
of +5VSB, and when the green LED is lit, the CMOS clock and CMOS
RAM are running off +5VSB and no current flows out of the CMOS
battery.

The +5VSB is also part of the control circuit in the PSU. If
a power supply won't start when you press the power button, it
could be the +5VSB has "collapsed", and there is nothing to get
the rest of the PSU running. Another reason can be something
shorting to the bottom of the motherboard, such as an extra
standoff that doesn't line up with a plated (grounded) hole on
the motherboard. If a brass standoff touches a copper track on
the underside of the board, it can break just about anything it
touches.

HTH,
Paul

 

[Dave Hines]

Two problems:
Power Supply is marginal - replace it with a current generation PSU
(Antec NeoPower 480 or better $125 or PC Power & Cooling 400W + units)
Beware of PSU's that are cheap & under spec.
CMOS Battery is under voltage - that's why the date defaults back to 1999.

 

[Hybyd2]

I appreciate all the responses from folks who took the time to do so. I do
realize based on this and other responses that the battery by itself would
not cause the startup problems. That there could be problems with the PSU.
Some basic followup questions (I'm a newbie):

1) About checking the battery: if a multi-meter reads 3.0V is this a
definitive sign that the battery is good? Or, can it read 3V when the
battery is out of the board and still not perform under load? Since the
system is running at this point and acts as my web-server, I'm reluctant to
experiment with it until I have a good plan of action. But I do have a
spare battery which registers 3.1 using a multi-meter.
2) Paul, in response to your question below about clearing the CMOS. If
that involves manipulating jumpers on the board, no I've never done this in
the year and a half that I've owned the computer (I bought it used).
3) If the battery is dead and the system unplugged from the wall power
socket is it a given that this (P34VX) system will lose it's clock setting?
I do realize based on what Paul's said before that there are other ways for
the clock to lose it's setting but I'm curious if the answer to the question
in the previous sentence is guaranteed to be yes.
4) If the battery is dead and the system unplugged from the wall power
socket is it a given that this (P34VX) system will lose it's knowledge of
what CPU is plugged in - as I mentioned when it does startup it goes into
the Advanced bios setup screen with the CPU speed at Manual and in the right
column saying that I've changed the CPU and so should adjust the speed.
5) If the +5VSB has collapsed is there a way to jump start the PSU / board?
6) What's really puzzling (to me) is that there is some weird sequence which
causes it to start. ???

Thanks,

Henry

 

[Al Smith]

This is a P3 733 MHz with an P34VX. The computer is on 24x7 - only the

monitor gets turned off when not in use, the disks are on and the system
does not go into standby or hibernate. One of the reason it's on 24x7 is
that if turned off, turning it on is a problem. And, when one succeeds in
turning it goes into the BIOS with the CPU speed set to Manual and a warning
that the CPU has been changed so its requesting the speed be set. I set it
at 733 and then all is well. Also, I've noticed the clock will need to be
reset once it goes into Windows as the time goes back to 1999 or some such
year. Could run for days / weeks until for whatever reason it gets turned
off. Because of the flaky behavior I've connected it to a UPS so your
normal brownout has no effect.

So what happens when it gets turned off? Typically, pressing the power
button has no response at all. It's totally dead. I then typically flip it
on it's side and open the case. There have been times when with the case
open pressing the power button has brought it back to life. It happened
again earlier this evening. With the computer open on it's side I actually
had to unplug it twice and replug it before it sprung to life. Some flaky
capacitor on the mother-board? Bad battery? If it's a bad battery,
wouldn't the only problem be that it wouldn't remember any bios settings?
Why would it refuse to start until I jumped through some still to be
determined mysterious set of steps.

Any insights into what the problem is would be greatly appreciated.

Thanks,

The only insight I have is that it's time to upgrade. Why are you
hanging onto a P3 that won't boot, when for around $100 you could
buy a P3 that *will* boot? Or better still, spend $500 and get a
more modern computer with some power.

 

[Zotin Khuma]

Dead CMOS battery. Replace it.

THEN see if there are 'other problems'.

I've seen a few motherboards that run fine with a good CMOS battery or
with none, but have great difficulty starting with a weak one. The
difference between a good battery and none is it won't retain settings
with none after a power-down.

 

[Arnie Berger]

The battery in this circuit is never under load. It provides a very
small holding current to the CMOS memory and real-time clock (RTC)
cicuitry. I doubt that it is at fault if it is reading 3.0V. I think
that ASUS uses a CR2032 battery that is supposed to put out 3.0V.

I think the problem is likely the power supply being right on the edge
of functioning. Most power supplies have overload protection circuits
that will shut it down if it has to supply too much current. One place
where the most current is drawn is at start-up, because the disk drive
motors draw much more current at spin-up then when they are at normal
rotational speed. So, I could see that as you try to turn it on, the
power-on circuit (yes, there is a start-up circuit to prevent big
current inrushes. It is called a soft-start circuit)detects an
overload and shuts it down before anything ever really gets going.
I think a beefy power supply should fix it, or at least eliminate the
power supply as a problem.

arnie

 

[GEO Me]

The only insight I have is that it's time to upgrade. Why are you
hanging onto a P3 that won't boot, when for around $100 you could
buy a P3 that *will* boot? Or better still, spend $500 and get a
more modern computer with some power.

Following your logic the reply to any question regarding hardware
problems would be: Buy a different one.
Did you work as a mechanic for a car dealership in your previous
life?

Geo

 

[Guest]

if a multi-meter reads 3.0V can it read 3V when the battery is
out of the board and still not perform under load?

Unlikely, unless the circuit powered by the battery has a short in it.
But measuring the battery in-circuit is just as easy; connect the red
meter probe to the top of the battery and the black probe to a ground
point (any bare metal on the computer case or power supply).

5) If the +5VSB has collapsed is there a way to jump start
the PSU / board?

Not the motherboard, and probably not the power supply. Have you
measured the +5Vsb at the purple wire (pin 9, next to yellow +12V wire
at the end)?

6) What's really puzzling (to me) is that there is some weird sequence which
causes it to start. ???

You wrote that it starts when it's on the side and a panel is removed.
That points to a problem with a mechanical connection, possibly even
a marginal short between the motherboard and case. Inspect each
mounting point carefully, even on the bottom side, to verify that no
traces come close enough to the mounting studs or screws to cause a
short). Also see if the motherboard or case is warped enough to allow
a short at a corner, and place stick-on rubber or vinyl bumper feet
where needed to prevent contact.

Electrolytic capacitors may have gone bad in the power supply or
motherboard, 2 sure signs of this being bulging on top or brown stains
on top or bottom (but faulty capacitors can also look normal). There
was a huge scandal over faulty electrolyte put into many Taiwan brand
capacitors made a few years ago. If this problem is in the power
supply, of course a new one will cure it, but you certainly don't need
anything expensive (Newegg.com has Fortron-Source -- top quality, very
low prices) or rated for more than 300W.

 

[David Maynard]

The battery in this circuit is never under load. It provides a very
small holding current

You mean a 'big' load because holding current *is* a 'load'.

to the CMOS memory and real-time clock (RTC)
cicuitry. I doubt that it is at fault if it is reading 3.0V. I think
that ASUS uses a CR2032 battery that is supposed to put out 3.0V.

I think the problem is likely the power supply being right on the edge
of functioning. Most power supplies have overload protection circuits
that will shut it down if it has to supply too much current. One place
where the most current is drawn is at start-up, because the disk drive
motors draw much more current at spin-up then when they are at normal
rotational speed. So, I could see that as you try to turn it on, the
power-on circuit (yes, there is a start-up circuit to prevent big
current inrushes. It is called a soft-start circuit)detects an
overload and shuts it down before anything ever really gets going.
I think a beefy power supply should fix it, or at least eliminate the
power supply as a problem.

A weak power supply wouldn't cause the settings to be lost.