Yoghoo wrote on 2025-03-23, 11:18:

...motherboard also looks pristine without any visible damage or defects.

It´s still PC-Chips 😉

Did you connect the external battery to the "EXT BAT" 4-pin header?
The battery feeds the clock oscillator "4069UBP" and the CMOS in "CHIP6". Near the osc. there´s cheapo electrolytic "C2" You can remove it to check if it was leaking.

You could do with leaving them in another day and checking voltage again.

Reason I say that, is because if you measured them out of circuit initially at 1.6V, it could either have been a truly new fresh battery, or because modern meters pull so little current a part used or stale battery that would have been near 1.2V under load, but was showing 1.6V of "surface" charge that would have dropped the instant it was connected. But if you read 1.6V in circuit with them operating, it was probably a new new battery.

However, we have a conundrum where the discharge curve is a bit S shaped, and just a quick short when installing could have dropped them near 1.2ish, since the decline down to 1.3ish is quite sharp at the beginning of discharge, then a low slope from 1.3ish to 1.1ish depending on battery, then it kinda falls off a cliff in the last 10%. Anyhooooooo 1.2V could be anywhere from 30% to 70% used depending on exact battery characteristics and deviations in lead and contact resistance at the meter. Curves for comparison https://www.powerstream.com/AA-tests.htm

So could use another data point or two to be sure how steep that middle "straight line" portion is. Not that you might not be right that there is a large draw, but not sure we can infer it for sure yet. Should be fairly easy to do an inline current measurement on this though.

majestyk wrote on 2025-03-23, 14:43:

It´s still PC-Chips ;) […]
Show full quote

Yoghoo wrote on 2025-03-23, 11:18:

...motherboard also looks pristine without any visible damage or defects.

It´s still PC-Chips 😉

Did you connect the external battery to the "EXT BAT" 4-pin header?
The battery feeds the clock oscillator "4069UBP" and the CMOS in "CHIP6". Near the osc. there´s cheapo electrolytic "C2" You can remove it to check if it was leaking.

Yes, using the EXT BAT 4-pin header. I checked C2 but there is no leakage whatsoever. As it's lifted a bit from the board I could check the underside as well as the top.

Last night I used some new batteries and after 12 hours it stopped updating the clock. Batteries were drained from 1.6V to 1.3V. So 0.3V in just 12 hours.

BitWrangler wrote on 2025-03-23, 15:10:

You could do with leaving them in another day and checking voltage again. […]
Show full quote

You could do with leaving them in another day and checking voltage again.

Reason I say that, is because if you measured them out of circuit initially at 1.6V, it could either have been a truly new fresh battery, or because modern meters pull so little current a part used or stale battery that would have been near 1.2V under load, but was showing 1.6V of "surface" charge that would have dropped the instant it was connected. But if you read 1.6V in circuit with them operating, it was probably a new new battery.

However, we have a conundrum where the discharge curve is a bit S shaped, and just a quick short when installing could have dropped them near 1.2ish, since the decline down to 1.3ish is quite sharp at the beginning of discharge, then a low slope from 1.3ish to 1.1ish depending on battery, then it kinda falls off a cliff in the last 10%. Anyhooooooo 1.2V could be anywhere from 30% to 70% used depending on exact battery characteristics and deviations in lead and contact resistance at the meter. Curves for comparison https://www.powerstream.com/AA-tests.htm

So could use another data point or two to be sure how steep that middle "straight line" portion is. Not that you might not be right that there is a large draw, but not sure we can infer it for sure yet. Should be fairly easy to do an inline current measurement on this though.

All batteries used were indeed brand new. First used some Amazon Basics AA batteries. Last night I used Grundig AA batteries. Both sets were drained the same way. Ordered some lithium batteries in the mean time. Will be here tomorrow.

Will leave the Grundig batteries in another day and check if it goes down further.

I meant electrical leakage, when a capacitor´s DC isolation / resistance is too low.

majestyk wrote on 2025-03-23, 16:11:

I meant electrical leakage, when a capacitor´s DC isolation / resistance is too low.

That makes more sense. 😀 But the RTC is updating correctly when the motherboard is turned on. Does an electrical leakage draw more power from the batteries then?

Was busy with a lot of other things the last couple of months but finally found some time to replace all capacitors, replace all diodes and also replace the RTC crystal.

But it still drains the 3 AA batteries in a couple of days. Also tried 4 batteries and 3 lithium batteries. The 3 lithium batteries lasted 2 weeks or so.

Finally I replaced it with a new 3.6V barrel battery but that also seems to be draining very fast.

So I am out of ideas what else it could be. If someone else has an idea please let me know.

See if it's the board itself or something connected to it. This can't be happening in your case of course but I've had USB speakers doing the same for example. Maybe you also have something out of the ordinary like a USB->serial converter for the mouse?

I'd try the board alone without anything connected to it. Wait some time, check battery, add PSU, cards & drives maybe. Wait some time, connect external devices... a process to help you find the culprit.

konc wrote on 2025-06-18, 08:38:

See if it's the board itself or something connected to it. This can't be happening in your case of course but I've had USB speakers doing the same for example. Maybe you also have something out of the ordinary like a USB->serial converter for the mouse?

I'd try the board alone without anything connected to it. Wait some time, check battery, add PSU, cards & drives maybe. Wait some time, connect external devices... a process to help you find the culprit.

Thanks for the suggestion. I tried it fully configured with all cards and other peripherals and with an empty board (only VGA card and keyboard). Also tried 2 different VGA cards to be sure but unfortunately in all cases the batteries are depleted at the same pace.

Took some time but finally solved the issue. Replaced all electrical capacitors, diodes and crystal before but that didn't solve it. In the end there was a ceramic capacitor which was faulty. After replacing it (and 2 others which were a bit out of spec) the battery doesn't drain anymore.

I had tested the ceramic capacitors before in circuit and none of them were shorted and gave ok values. But after de-soldering (as a last resort) it showed very strange results with a component tester. Every time I tested it it gave a different value back. Never saw that before so learned something new. 😀

So finally can enjoy this motherboard. After replacing the standard BIOS with MR BIOS it's a very nice board with a lot of down clocking options (Ctrl-Alt-Plus/Minus as well as the turbo button). Also don't need XT-IDE anymore for big HDD's with this BIOS.

Very common for ceramic capacitors to develop internal cracks that allow leakage. Often they go dead short, but, like in your case, it was modest leakage. It's doesn't take much to drain 4 alkaline AAs if the drain is 'round the clock. Remember, adding batteries in series doesn't increase Amp Hours. It does, however, increase Voltage and Watt Hours. So in this case of a steady current drain, 4 series AAs drain just as fast as 1 AA.