According to the picture and text on https://theretroweb.com/motherboards/s/chicony-ch-486-33-50p this contain a Nickel-Cadmium battery. These were generally recharged by the motherboards when they were online.
A NickelCadmium cell usually have 1.2 Volts, so this battery most likely have 3 cells.
The charging voltage is generally between 1.4-1.6 volts per cell.
So if you measure the voltages between the plates with a multimeter while the motherboard is running(careful to not short something), and it is between 4.2 to 4.5 then it tries to charge it.
But even if the voltage measure is not that value I would still assume to be a charging circuit to be present just to be safe.

MagefromAntares wrote on 2026-05-09, 23:31:

According to the picture and text on https://theretroweb.com/motherboards/s/chicony-ch-486-33-50p this contain a Nickel-Cadmium […]
Show full quote

According to the picture and text on https://theretroweb.com/motherboards/s/chicony-ch-486-33-50p this contain a Nickel-Cadmium battery. These were generally recharged by the motherboards when they were online.
A NickelCadmium cell usually have 1.2 Volts, so this battery most likely have 3 cells.
The charging voltage is generally between 1.4-1.6 volts per cell.
So if you measure the voltages between the plates with a multimeter while the motherboard is running(careful to not short something), and it is between 4.2 to 4.5 then it tries to charge it.
But even if the voltage measure is not that value I would still assume to be a charging circuit to be present just to be safe.

Perfect, and your mathematics seems right; I'll check with multi anyway, but at this stage you're absolutely right, just assume charging and use NiMH. Cheers!

I don't recommend directly replacing the onboard battery.
Most boards had poorly designed charging circuits that shortened the lifespan of already leak prone Ni-CD batteries. A direct Ni-MH replacement may still pose a leak hazard.
It appears the board has an external battery header (J6). I recommend taking advantage of that, instead of another onboard battery.

akimmet wrote on 2026-05-10, 01:01:

I don't recommend directly replacing the onboard battery.
Most boards had poorly designed charging circuits that shortened the lifespan of already leak prone Ni-CD batteries. A direct Ni-MH replacement may still pose a leak hazard.
It appears the board has an external battery header (J6). I recommend taking advantage of that, instead of another onboard battery.

Indeed, I'll be using J6 and a holder! The last owner removed the batt.

My thinking was just to use AA NiMH inside a holder, 3 x 1.2v to match the original voltage, that way if they leak at some point the goo is contained in the plastic box?

Or maybe an LIR2032 but I haven't decided yet.

I glanced at a board the other day and noticed it seemed to have 2 footprints in one, with a Nicad pair of holes and a lithium non-chargeable set of holes, so presumably the latter would not charge.

Dan386DX wrote on 2026-05-10, 01:07:

My thinking was just to use AA NiMH inside a holder, 3 x 1.2v to match the original voltage, that way if they leak at some point the goo is contained in the plastic box?

The "goo" creeps unfortunately. What appears to happen is that the hydroxide eats the nearest metal, then briefly is chemically "spent", but oh no, regenerates with atmospheric oxygen and moisture so it is able to attack the next nearest clean metal, repeat and repeat, until it has crawled all the way down a wire or halfway across the board.

BitWrangler wrote on 2026-05-10, 01:16:

I glanced at a board the other day and noticed it seemed to have 2 footprints in one, with a Nicad pair of holes and a lithium non-chargeable set of holes, so presumably the latter would not charge.

That isn't always the case on motherboards that had a dual footprint. Many have both footprints connected in parallel. Sometimes other component changes were required to disable charging.

Of course they charged the barrel battery. If not, it would have discharged quickly. NiCd and NiMH both have self-discharge of 20%+ per month, even.

Dan386DX wrote on 2026-05-10, 01:07:

Indeed, I'll be using J6 and a holder! The last owner removed the batt.

My thinking was just to use AA NiMH inside a holder, 3 x 1.2v to match the original voltage, that way if they leak at some point the goo is contained in the plastic box?

The external battery header does not charge the batteries unless it's an unusually deviant board. It was designed for an external primary battery like this: Computer Clock Battery rebuild

Unfortunately there really wasn't a standard for the CMOS battery header.
A few boards had a charge circuit for the header, but most didn't. Sometimes battery header charging could be selected by jumper.
Most boards tolerate a 3V lithium coin cell and a charge blocking diode. Some boards require at least 3.6V to maintain CMOS settings.

A while ago I asked the experts here about certain boards that have an EXT BATT header that also serves the purposes of an "internal vs. external battery jumper" and a "CMOS clear jumper"; you can read more about the diodes and such in that thread. Better understanding barrel battery charging circuit

akimmet wrote on 2026-05-10, 01:40:

Most boards tolerate a 3V lithium coin cell and a charge blocking diode. Some boards require at least 3.6V to maintain CMOS settings.

In particular, if you try to connect a 3V coin cell across the EXT BATT header that already has the diodes and doesn't charge across the EXT BATT header, you will probably find that the CMOS settings hold, but the date/time freeze in place when the system is powered off.

BitWrangler wrote on 2026-05-10, 01:20:

Dan386DX wrote on 2026-05-10, 01:07:

My thinking was just to use AA NiMH inside a holder, 3 x 1.2v to match the original voltage, that way if they leak at some point the goo is contained in the plastic box?

The "goo" creeps unfortunately. What appears to happen is that the hydroxide eats the nearest metal, then briefly is chemically "spent", but oh no, regenerates with atmospheric oxygen and moisture so it is able to attack the next nearest clean metal, repeat and repeat, until it has crawled all the way down a wire or halfway across the board.

This is the closest thing I've read to a horror script in here, but thanks for the heads up!

jakethompson1 wrote on 2026-05-10, 01:34:

Of course they charged the barrel battery. If not, it would have discharged quickly. NiCd and NiMH both have self-discharge of 20%+ per month, even.

Yeah, I should have spotted the capacity of the barrel in that TRW picture, just 60mAh + self drain.