BIOS battery backup issue...

Thanks TV.... I'll try the jumper and the new battery. Although, just some random thoughts...

If the battery cannot retain two BIOS settings, others seem to be fine including time/date, it does not seem to be battery issues; especially, when the battery just had been replaced. While measuring voltage does not indicate strength of the battery, the old, unmarked battery shows 3.15V; pretty much the same as the new Duracell one.

The motherboard does have jumpers as you described and will try that with and without the battery. Based on the manual for this board, with the battery in and changing the jumper, the BIOS is not cleared at times. I fail to see how the BIOS would be cleared when the battery is not connected, but I'll take their word without questioning.

Shopping for repair is out of question; I've built this machine by myself and pretty comfortable with ripping it apart and putting it together. It's a pain working on it since replacing the BIOS battery requires removing the video card and the jumper is hidden behind the cables, but I am OK.

Replacing the motherboard is also out of question. It's hard to find the same board, Newegg and other "trusted sources" does not sell it anymore. Less known sources sell it for around 300 bucks, which is a ripoff. The current i5-760 CPU (first generation) is getting old anyway and rather just build a new machine from scratch.

A bit of (excessive?) precision:

The BIOS is stored in an EEPROM chip. It's completely nonvolatile; you can unplug the chip from the PC, and the BIOS is retained indefinitely.

The BIOS settings are retained in CMOS. That's volatile, but it takes very little power to maintain. The CMOS cell usually lasts for years. I usually replace a motherboard before its CMOS batter dies.

I'm not sure what might be causing your problem. I believe that most PCs won't boot if the ClearRTC (CMOS clearing) jumper is in the wrong position. I hope that you don't have a defective board.

Layback Bear said:

Just a though. One of my motherboards has the jumper in the instruction that came with the motherboard wrong. Went on line and got the correct jumper position. Next try to use a non Marvell SATA port if you have one.

This PC is almost two years old and running just fine until lately. Once the BIOS, or more correctly the CMOS is reprogrammed, the system works just fine. The Marvell SATA port that has its own BIOS and works just fine during the boot process and within Windows.

While I didn't have time to clear the CMOS via the jumper, unintentionally that had been done already. Couple of days ego the CMOS battery had been changed, that did a "ClearRTC". That was the only time when the time/date had been lost; other times it is the storage type and the boot order that are changed.

On the surface, it does seem that the PC has a bad motherboard. The "easy part had been done already like, change CMOS battery, ClearRTC, flush the EEPROM, or BIOS, without remediating the issue.

During my "training" session about the BIOS, or rather the EEPROM and CMOS, it became evident that the EEPROM should not be flushed within Windows. Nor should it be flushed utilizing Asus Update within Windows. The recommended method is "Asus EZ Flush2" within the CMOS configuration, or the "bupdater.exe" DOS utility.

The EEPROM had been flushed via "Asus EZ Flush2" and upon reboot, it came up with this message:

New CPU installed, please enter setup to configure your system

Pressed "F1", checked the configuration, saved it, and rebooted. The first reboot resulted in a "reset" where system in itself killed the power for 5-10 seconds. Interestingly, there had been no changes made to any other settings within the CMOS; the date/time, storage type, boot order, etc remained the same.

The system had been shutdown, power killed to the box for 5-10 minutes, turned the power strip switch on, and booted the PC. The system passed the BIOS screen and booted Windows, without any interaction with the BIOS.

On the surface, it does seem like that the issue had been resolved; however, the real test is still remaining. The power strip switch will be shut off overnight; if it boots just fine in the morning, then this issue is resolved.

Well, the issue has not been resolved and I doubt that it will be anytime soon...

Shutting off the power strip at night resulted in BIOS error and asking for running the setup. At this time, none of the BIOS settings were changed; the storage type, boot order, etc., were correct. After saving the BIOS without making any changes, the machine booted just fine.

I've rebooted the box, got into the BIOS again, change the storage type from IDE to ACHI, saved it, and booted to Windows. The OS started up just fine and didn't even asked for reboot. The system is running in AHCI mode, and the performance improvement is quite noticeable both during bootup time and general use.

I am curious as to why shutting off the power to the box has this issue? Everything points to the battery backup for the CMOS, or BIOS, but that doesn't make much sense.

Both the original, about two years old, and the new batteries are good. Shutting down the PC shuts of the power supply and there should be no power to the motherboard going through the power supply. Shutting off the power strip shouldn't make any difference, as far as the power supply and the BIOS battery backup is concerned. It's a weird issue...