will a flat cmos battery prevent boot up?

[Vagrant]

I have been putting together an old T-bird 750 for a friends kid.
When I turn on the powersupply, the cpu cooling fan turns and the leds on the board flash. This is normal. The Monitor also powers up, again normal.
But when I push the power button, nothing happens. I have checked the switch and also the reset switch with a multi meter, they are fine.
I have removed the drives, ram and video card, without any luck.

Could a flat cmos battery be preventing this system from even starting to post?

It has an MSI K7t mainboard
leadtek geforce2 mx video card
cd rom
20 gb seagate hard disk.

It worked fine till about a year ago when I retired it, all I removed from it was the cd burner.
It has been sitting in the rack with my two working machines, so is warm and dry.

Thanks for any help

 

[Bozo Galora]

that computer is old enough that it might

http://computer.howstuffworks.com/question319.htm

easy enough to take it out and check it with VOM

 

[Zepper]

Is it possible that you have the power switch connected to the wrong pins? AFAIK nothing should happen when you plug the PSU in or turn it on (no CPU fans etc.) - nothing until you press the power button. That leads me to believe that something is wrong with your PSU.
.bh.

 

[Vagrant]

Bozo Galore, thanks for the link, I will try a new battery, multi meter says the old one is putting out 3 volts, as labled, but that is a no load reading.

Zepper, MSI call it "instant power on" dont' know why they did it, but as soon as you conect power, the fans and diagnostic lights start up for about a second, then turn off. Nothing was changed from last time I ran it, apart from removing the burner.

 

[montag451]

when you remove devices, it is very easy to 'just disconnect this cable for a sec to get to this screw', then forget to put something back, or pull something out without realising it, or knock something

 

[hopejr]

I don't think a flat battery would stop it from starting. It will clear the bios settings which you'd have to set every time you start, but that's about all.

 

[JC at Ultra]

Originally posted by: hopejr
I don't think a flat battery would stop it from starting. It will clear the bios settings which you'd have to set every time you start, but that's about all.

I've never had a dead CMOS battery prevent a PC from posting. It's purpose it only to store CMOS settings when the PC is off.

BozoGalore's link is to someone that's talking about a PC so old, that the battery is soldered to the board. A socket A uses a standard CR2032 battery.

IF your battery is soldered to the motherboard (looks like a box and probably says "DALLAS" on it) then it's got a real-time clock built into it and the PC wouldn't POST if it was dead. But I doubt that's the case.

 

[VirtualLarry]

Originally posted by: Zepper
Is it possible that you have the power switch connected to the wrong pins? AFAIK nothing should happen when you plug the PSU in or turn it on (no CPU fans etc.) - nothing until you press the power button. That leads me to believe that something is wrong with your PSU.
.bh.

.. unless the CMOS battery is dead, and various things are in an unstable or random state, such as the RTC power-on alarm firing due to the circuits matching all of the bits from the RTC reading with the stored power-on-time value in the CMOS, etc. I do know that the RTC is involved someone in the ATX power-on signal in most boards, it's quite conceivable that funky CMOS settings/bad battery could cause issues.

However, usually it simply never powers on at all, never even tries to, on most of the boards that I've seen. More likely, something is shorted or the mobo is just plain bad if the system attempts to power-up when you plug in the AC power cord. (The CMOS data controls what happens in that case too, actually, so I guess a dead battery could cause that too, perhaps.)

 

[JC at Ultra]

Originally posted by: VirtualLarry

Originally posted by: Zepper
Is it possible that you have the power switch connected to the wrong pins? AFAIK nothing should happen when you plug the PSU in or turn it on (no CPU fans etc.) - nothing until you press the power button. That leads me to believe that something is wrong with your PSU.
.bh.

.. unless the CMOS battery is dead, and various things are in an unstable or random state, such as the RTC power-on alarm firing due to the circuits matching all of the bits from the RTC reading with the stored power-on-time value in the CMOS, etc. I do know that the RTC is involved someone in the ATX power-on signal in most boards, it's quite conceivable that funky CMOS settings/bad battery could cause issues.

However, usually it simply never powers on at all, never even tries to, on most of the boards that I've seen. More likely, something is shorted or the mobo is just plain bad if the system attempts to power-up when you plug in the AC power cord. (The CMOS data controls what happens in that case too, actually, so I guess a dead battery could cause that too, perhaps.)

Good point, actually.

I've had people RMA motherboards that acted the same way that simply had the clear CMOS jumper in the clear position, or had the CPU jumpers (remember those?) set incorrectly.

 

[VirtualLarry]

Originally posted by: Vagrant
Bozo Galore, thanks for the link, I will try a new battery, multi meter says the old one is putting out 3 volts, as labled, but that is a no load reading.

Zepper, MSI call it "instant power on" dont' know why they did it, but as soon as you conect power, the fans and diagnostic lights start up for about a second, then turn off. Nothing was changed from last time I ran it, apart from removing the burner.

That's interesting. I was just at the QDI (mobo) site today, reading their FAQ entries, and they had one that mentioned the system *automatically* powering on for a split-second after flashing the BIOS on a particular mobo. Their explaination was that the system was auto-starting, briefly, to "detect" the PSU.

Personally, I think that sounds a bit bizarre to me, I've never heard anything like that. But it's possible that the act of flashing the BIOS clears the CMOS, and on that (and perhaps others, like the MSI board in question?), triggers auto-power on due to the CMOS settings, and it powers on enough to adjust the settings/clear the checksum/etc. Something like that, anyways, that would prevent any further auto-power-ons. (Perhaps when the RTC gets power, when it counts the next second, it triggers a compare on the stored power-on-alarm values, and since the CMOS was reset, it triggers a power-on event, but then the BIOS POST recognizes that the CMOS checksum is invalid or otherwise cleared out, and then disables the RTC alarm setting, and possible sets the CMOS checksum, or something, etc. so it doesn't happen again, and then powers back down.)

I would still think that would cause the system to at least partially POST (bootblock early initialization) if that were true, so I don't think that's the correct explaination. It wouldn't be safe for the HDs either, to power-on for such a short period of time that they don't get to fully spin up and run internal firmware diagnostics, which often write to an internal error/event log stored on a non-host-accesable sector on the HD. Cutting the power to the drive (at least on IBM's Deskstars), while writing, could semi-permanently damage the low-level format of the drive and possibly the on-media firmware. (Never had that problem with WDs though.)

Edit: Oops, sorry, O-C disorder kicking in, following is actually diagnostics for a board that fails to start up, not starts up when power is plugged in to the supply, although it could explain why it suddenly powers off a split-second later. Personally, I suspect a buggered mobo though, period.

Edit: Also, most (half-decent) ATX PSUs, will auto-shutdown if they: 1) have a short, 2) have issues with the properly loading/operation of the switching regulator circuits, 3) overheat, 4) PSU fan failure, or other things.

I would also double-check that the ATX mobo power connector is properly connected, and that there isn't anything blocking one of the pins, and/or or that non of the pins are loose and got "pushed" back into the connector, resulting in no load on that voltage plane/regulator in the PSU. Likewise for drive molex power connectors, sometimes they fray, and one little wire strand pops out of the connector, and touches another one, an accident waiting to happen. Seems to happen most often when using cheapo aftermarket power-molex Y-splitters.

Oh yeah, one last "duh" thing, make sure that you have a fan with RPM/tach feedback plugged into the appropriate connector in the mobo, usually the CPU fan. Some boards were famous for not powering up due to not recieving a tach signal, to prevent CPU overheating/burn. A later BIOS for my old Abit BX6-r2 board was that way. If the CMOS is cleared, either manually or due to a a dead/flaky battery, it could reset that internal setting to default which was "ON" for that particular board/BIOS version.

 

[oldmacs]

I just had my mother-in-laws, seldom used, 8 yr old Compaq Pavillion fail all attempts to POS. Replaced cmos battery and all is well.

 

[VirtualLarry]

While it seems like I used to be more knowledgeable about CMOS batteries, I tried replacing the battery on a PC that would automatically power-on after being powered off, but it didn't help.

 

[GrumpyMan]

Just had to replace one on the ex's Optiplex 745. It would stop at boot with a prompt to hit F1 to go into Bios or hit F2 to keep on booting due to a low system battery. The date in the Bios was at 2006. Bought a 2032 battery at CVS and all was good to go after that and boots normally.

 

[mfenn]

Necro thread, locked.

mfenn
General Hardware Moderator