First suspect is JP25, pins 1-2 need to be covered for normal operation. Check that jumper, if it's missing it could lead to strange behaviour, and if it's on 2-3 (CMOS CLR), it's pretty clear that it won't keep settings.

Second would be BIOS. It's pretty obscure, but I know of at least one instance (GVC FR520, also with a SiS chipset) where an old BIOS revision led to the CMOS battery being discharged in very short time indeed. A later BIOS update fixed that. No idea about MTI boards (never seen one in the flesh myself), but something to look into if JP25 seems OK.

Thanks dionb ... J25 pins 1-2 are covered. I'll see what I can find out about this old BIOS this weekend.

So I Googled the BIOS of this old Mustang motherboard (ROM PCI/ISA BIOS (2A5IHR0C) Standard CMOS Setup Award Software, Inc.) ... specifically BIOS 2A5IHR0C ... and came up with one website that has a BIOS update (I think) for this board. But it's not an Award website. And it's not clear whether the download is a BIOS upgrade. So I'm too shy to try upgrading the BIOS.

Sooo ... after powering off/on this board for about a week ... I still get the same behavior ... even when trying a third new battery:

*When powered off for 24 hours ... the BIOS/Win98SE loses a day ... and loses the time.

I guess not a big deal because I can set the correct time in the BIOS when I first turn it on ... but I do use this old PC somewhat regularly for a couple old DOS games. I can play those games in DOSBox on an XP PC too ... but it's more fun on this older PC 😀

Anyway ... one question I have is ... is there anything else I can try besides upgrading the BIOS to see if I can keep this board in correct date/time?

PS: Just remembered ... I believe this board will also lose the HDD settings if off for like over a week. I haven't tested for that again ... but I know it lost the HDD setting recently when it was powered off for about a week. Again ... not a big deal because I can reset the HDD info. ... but kind of a pain. Just wondering.

OK ... just thought I'd update this thread in case someone else ever has this problem.

1) I checked all three batteries (with a multi-meter) that I've tried in this motherboard since they all came from the same package ... all are OK at 3 volts and work fine in other PC's.

2) I gently cleaned the two metal battery holder contacts with a pencil eraser ... and inspected the contacts for any bends. Both contacts seem OK and push up nicely against the negative side of the battery while the battery is in the slot (so seems like good battery contacts).

3) I left the PSU plugged into power when the PC is off to see if the CMOS still loses time (it does). I had read that if you leave the PSU plugged into the house power ... that it sometimes continues to provide power to the motherboard and battery ... allowing the CMOS clock to keep it's data (time, etc).

4) I checked one more time the behavior of the CMOS time when the PC is turned off. It pretty reliably loses a minute if you shut it off and turn it back on in a minute. And same for any other shut off time interval (off for 30 minutes, 4 hours, etc). Also loses date after 24 hours.

So I'm thinking I either need to upgrade the BIOS (I'm still trying to find one that I feel confidant is for this old motherboard) ... or ... I may need to replace the RTC crystal. I've never done that before ... but I see it's near the CMOS chip. I may try to find a crystal on another motherboard if it is of the correct values. I'm leaning towards replacing the RTC crystal at some point.

Anyway ... I forgot how much fun it was to try to bring these old PC's back to life 😀

Could there be any battery acid anywhere in the area of the clock/battery area? How about post some pics? BTW: current draw for a CMOS settings/clock circuit should be in the very low uA range, just a couple of uA at most!