tony359 wrote on 2025-11-06, 18:18:

Any ideas? What would be an expected current for a 2032 battery? I read something around 0.19mA.

I've recently removed some batteries from DS12887 equivalents. Those were BR1225 batteries.
Specified capacity is around 48mAh.
Seeing that those modules are specified as retaining the memory & clock for 10 years. I accounted for 8 hours/day of powered time.
which gives around 1µA.
A CR2032 seems to be around 235mAh, that should allow for about 50 years of data retention. though I have never yet seen any mainboard do that.
So let's say 10 to 20µA would be a good guess.

Feel free to correct me if I did my math wrong.

edit: don't forget diodes can also fail.

Thanks. What puzzles me is that I read 0.000mA on two other motherboard. I wonder whether I should attach the battery via current meter, turn on the board, save the settings and turn it off? Maybe there is no current because I have already removed the battery to wire the current meter and that disables the circuit altogether?

I believe a quick search suggests 0.19mA as what I should expect which matches your 0.20mA if I've done the math right 😀

tony359 wrote on 2025-11-06, 18:56:

Thanks. What puzzles me is that I read 0.000mA on two other motherboard. I wonder whether I should attach the battery via current meter, turn on the board, save the settings and turn it off? Maybe there is no current because I have already removed the battery to wire the current meter and that disables the circuit altogether?

I believe a quick search suggests 0.19mA as what I should expect which matches your 0.20mA if I've done the math right 😀

FYI, 200µA != 20µA

edit: oh and btw, the math suggests 1µA ie, 0.001mA
the 20µA is just a guess 😀

I knew I should have checked better 🙂

tony359 wrote on 2025-11-06, 18:18:

Hi all, […]
Show full quote

Hi all,

I'm working on a GA-6WMM7 which seems to deplete CMOS batteries in a very short time.

I've tested a new battery, reading 3.3V new, and the following day it was already at 2.8V.

I compared the currents being drawn by other boards, somehow I cannot even measure it with the board unpowered and my meter goes down to 0.001mA (yes mA, not A). The Gigabyte instead reads 0.7mA all the time. When I apply power (just 5VSB actually) the batery current goes to zero as expected.

I've swapped three capacitors I had on the VccRTC line which were between VccRTC and ground but no change. There was some minor corrosion in the area but so far I don't see anything. I found the attached as a reference circuit for the ICH on this board.

Any ideas? What would be an expected current for a 2032 battery? I read something around 0.19mA.

Ciao Tony,
So for the battery that drains, I would check as they have already suggested a diode, it could have a fault or fail, this in my opinion should be checked at least.
From the photo I see D2 D3 D5 and D27, I don't know which one goes to the battery 🪫, but I would check those, the problem with this type of motherboard, with i810E chipset, should be that with the battery low it doesn't start, but have you noticed if the PC clock works and remembers the date and time, or does it restart when you start the PC?
I hope 🤞 I was useful to you, good job.

Ciao

I think it follows intel's schematic I posted. One diode is for the PSU and one for the battery. But they don't go to ground so why would a bad diode cause the issue? D2 and D3 both read ok BTW but they're in circuit. I don't think D5 and D27 are part of the battery circuit but I can check for sure.

Indeed this board doesn't start with no or low battery but I was wondering the same thing: does the RTC work while the battery is good? I'll find out.

Thanks so far!

tony359 wrote on 2025-11-06, 22:22:

Ciao […]
Show full quote

Ciao

I think it follows intel's schematic I posted. One diode is for the PSU and one for the battery. But they don't go to ground so why would a bad diode cause the issue? D2 and D3 both read ok BTW but they're in circuit. I don't think D5 and D27 are part of the battery circuit but I can check for sure.

Indeed this board doesn't start with no or low battery but I was wondering the same thing: does the RTC work while the battery is good? I'll find out.

Thanks so far!

Consider for a moment, in the schematic, the diode that stops the 3v3 rail from being fed by the battery is leaky and any circuitry that is attached to the 3v3 line is now badly being fed by the battery.

Ah - ok, I didn't consider that.

I'll take a look, thanks! I could temporarily remove it to be sure.

It was the diode!!!!

I removed one leg of it to basically disconnect the voltage coming from the PSU - so it would only run on battery. It read 0.007mA and then went to 0.000 (I think the meter can do 0.007mA minimum). If I reconnect the disconnected leg even for a second, it would go back to 0.7mA!

I removed the diode and it read 0.3V as it should in one direction but 1.8V in the other! Wow!

I replaced it with a BAT85 I had, should be similar voltage drop, and it now reads 0.000mA!

One question: with no battery I read 0.3V at the battery terminal with the PSU turned on so I thought I'd replace the other diode as well and now I read 1V! But it looks like a "phantom" voltage, it slowly goes down and if I touch it while touching ground it goes down to 0.3V again. Shall I be concerned?

I've set the clock and will test again tomorrow but it's clearly improved a lot.

THANKS SO MUCH for your input!

Not really following the voltages you're measuring... I guess it's ok.

Oh wait, maybe. You've removed the battery and then measured the voltage on the battery terminal. right?
Diodes aren't 100% perfect blockers in one direction, and there's also some capacitive coupling, especially when switching power states.
Nothing to be concerned about.

Have fun & Enjoy !

Yes, I removed the battery and then turned on the PSU, I wanted to make sure there wasn't any voltage at the battery terminals.

tony359 wrote on 2025-11-07, 19:50:

Yes, I removed the battery and then turned on the PSU, I wanted to make sure there wasn't any voltage at the battery terminals.

Congratulations, great job 👍

I think we'll see a video on your YouTube channel, right?

tony359 wrote on 2025-11-07, 19:50:

Yes, I removed the battery and then turned on the PSU, I wanted to make sure there wasn't any voltage at the battery terminals.

I wouldn't have been surprised even if you measured about 3V3 on the battery terminal.

[/quote]
I think we'll see a video on your YouTube channel, right?
[/quote]

You will 🙂

weedeewee wrote on 2025-11-07, 20:10:

I wouldn't have been surprised even if you measured about 3V3 on the battery terminal.

Is that a similar phenomenon to the 68V AC I could measure on a wire which was disconnected from anything but running in a wall along with other live conductors in the same jacket? As soon as I wired that cable to something, the voltage would disappear.

tony359 wrote on 2025-11-07, 22:33:

weedeewee wrote on 2025-11-07, 20:10:

I wouldn't have been surprised even if you measured about 3V3 on the battery terminal.

Is that a similar phenomenon to the 68V AC I could measure on a wire which was disconnected from anything but running in a wall along with other live conductors in the same jacket? As soon as I wired that cable to something, the voltage would disappear.

No, that would be more like inductive or capacitive coupling like in a transformer or an antenna.

edited to correct the quotes.

Here you go 🙂

This thread mentioned in the description! Thanks again all for the help, it was a fun repair!

https://www.youtube.com/watch?v=mEf2yl6X458

tony359 wrote on 2025-11-21, 18:17:

Here you go 🙂

This thread mentioned in the description! Thanks again all for the help, it was a fun repair!

https://www.youtube.com/watch?v=mEf2yl6X458

Hi thanks and congratulations nice video. 😺

At those microamp levels, the meter voltage would probably increase if you breathe near it