Excessive battery draw is a common symptom of a failing chipset IC, when the RTC/CMOS is integrated in the chipset. Although it's possible a cap/diode in the vbatt circuit has excessive leakage. But in my experience it's almost always the chipset.

You need to measure the current draw on the battery. It should be <1 uA. (Check datasheet for actual figure)

You can take the clear cmos jumper off and place your voltmeter between the two pins used for normal operation. Or can put one lead on the battery top, and one lead on the side spring contact, and push the contact away with your probe, which will redirect the current through your meter.

Do this with the board powered down.

You need a meter that can measure micro amps. nano amps would be even better. Although, if it's going dead that quick, you might see low mA.

The bios e/eprom isn't battery backed. It's the RTC/CMOS integrated in the SIS chipset on your board.

kingcake wrote on 2024-03-02, 01:32:

Excessive battery draw is a common symptom of a failing chipset IC, when the RTC/CMOS is integrated in the chipset. Although it' […]
Show full quote

Excessive battery draw is a common symptom of a failing chipset IC, when the RTC/CMOS is integrated in the chipset. Although it's possible a cap/diode in the vbatt circuit has excessive leakage. But in my experience it's almost always the chipset.

You need to measure the current draw on the battery. It should be <1 uA. (Check datasheet for actual figure)

You can take the clear cmos jumper off and place your voltmeter between the two pins used for normal operation. Or can put one lead on the battery top, and one lead on the side spring contact, and push the contact away with your probe, which will redirect the current through your meter.

Do this with the board powered down.

You need a meter that can measure micro amps. nano amps would be even better. Although, if it's going dead that quick, you might see low mA.

The bios e/eprom isn't battery backed. It's the RTC/CMOS integrated in the SIS chipset on your board.

Damn, these are not good news... How did you solve that problem or did u scrap the board? Or is there chance to replace it, although I think it will not be easy to source it. There are two SIS chips - 85C496 and 85c497...
Other thing I forgot to mention, I have currently 1.71 bios version installed and it somehow messes up tha date... If I set year 2024, everytime I return back to BIOS settings, it says 1996 (but it keeps date and time correctly). In dos or if I create new folder, year it reads is 2096..... Maybe some relation to the first problem? Currently no checksum/battery error message as I have replaced it only yesterday.

sounds like a Y2K bug. most 486 bios have em...