No it should be about same voltage as the original Varta and 3.7-3.8v is standard. The diode keeps the board volts from trying to recharge the coin cell.

Horun wrote on 2022-07-10, 21:37:

No it should be about same voltage as the original Varta and 3.7-3.8v is standard. The diode keeps the board volts from trying to recharge the coin cell.

3.7V 3.0V is the nominal voltage for most lithium coincells, and while yes, regular NiCd/NiMh batteries have a nominal voltage of 1.2V (so 3.6V for a 3-cell barrel battery) that's not the entire story, you need to consider charge voltage and that barrel batteries on motherboards are usually kept charged at all times and how the charging circuits work with current limiting, that means they can be at 4.5/4.6V or even just 5V, sometimes maybe even higher.

though I'm surprised why there is any voltage present at all at the battery holder, shouldn't there be a diode to block that?

Ok I thought you meant at the original Battery pins, not at the replacement battery.
The schematic calls for a diode in series with a 220 ohm resistor off the Coin V+ to the original battery + pin. My guess is that the diode in backwards...

Someone made a external battery using 3xAAA in series for 4.5v I looked up the charging profile for the varta which I believe is 3s nicad or nimh and up to 4.5v is acceptable for that chemistry. So 4v may be charging voltage.

These boards may work with 3.3v also. This adapter might be designed to prevent charging.

imi wrote on 2022-07-10, 22:59:

3.7V is the nominal voltage for most lithium coincells, and while yes, regular NiCd/NiMh batteries have a nominal voltage of 1.2V (so 3.6V for a 3-cell barrel battery) that's not the entire story, you need to consider charge voltage and that barrel batteries on motherboards are usually kept charged at all times and how the charging circuits work with current limiting, that means they can be at 4.5/4.6V or even just 5V, sometimes maybe even higher.

though I'm surprised why there is any voltage present at all at the battery holder, shouldn't there be a diode to block that?

Regular LiSO2 cells (like the legendary CA2032) have a nominal (idle) voltage of 3.0V.
The only Lithium cell that has 3.7V is the LiSOCl2 (Lithium Thionyl Chloride) cell that´s very uncommon in consumer electronics.

Horun wrote on 2022-07-10, 23:12:

Ok I thought you meant at the original Battery pins, not at the replacement battery.
The schematic calls for a diode in series with a 220 ohm resistor off the Coin V+ to the original battery + pin. My guess is that the diode in backwards...

Ye my thought also but it is added after the bord markings, but will try to reverse it.

is your diode shorted by any chance?

don't just reverse it, it seems to be installed correctly.

majestyk wrote on 2022-07-11, 05:36:

Regular LiSO2 cells (like the legendary CA2032) have a nominal (idle) voltage of 3.0V.
The only Lithium cell that has 3.7V is the LiSOCl2 (Lithium Thionyl Chloride) cell that´s very uncommon in consumer electronics.

you're right, sorry I was thinking of rechargable lithiums ^^ ...which only makes it worse though 😁

I have now changed diodes from ”1N4148-TAP” to ”1N4148” i guess ”tap” has something to do with it?

Now it just jumps random between mv and v which I understand is normal.

What about LIR2032 lithium-ion batteries? They are 3.6V and rechargeable.
You can use them with regular CR2032 battery holder and being rechargeable suggest you don't need the charge blocking diode. Also these are like 2€/$
.

that's not how that works, you'd still need a dedicated charging circuit for them and not just connect them to one that is built for charging 3S NiCd batteries

imi wrote on 2022-07-18, 10:34:

that's not how that works, you'd still need a dedicated charging circuit for them and not just connect them to one that is built for charging 3S NiCd batteries

What is the charging profile for those rechargeable coin cells? If it’s not exact, it may be close enough.

I believe nicad tops out at max 1.5 VPC (x3)

it's right there in the link that RandomStranger posted: Constant Voltage 4.20V, Constant Current 17mA

So it sounds like it will work.

It mentions a cutoff voltage, so it sounds like it even has a basic integrated bms.

Which is pretty typical of many consumer cells.

what? no, those are the specs your curcuit has to adhere to, which on any board that wasn't designed for it obviously does not, if you discharge the cell below 3.0V you risk damaging it or reducing it's lifespan.
same obviously goes for overcharging it, there is no BMS in that cell.

OP said they are getting 3.8 to 4 V constant when energized that should maintain the battery level and that is totally within safe range of that cell.

A Mechanically confined Lithium ion cell like that will probably survive a over discharge event as long as the BMS can be reset if it has one that is
I have actually seen some remarkably small single BMS boards, it is not impossible that that cell has one.In any case it doesn’t seem that the application is capable of over volting that cell which is the major concern Should a thermal event occur a cell of that size does not contain much thermal energy, enough to damage the motherboard possibly, probably enough to give you a burn if one of the pieces touches you but I highly doubt that it would come anywhere close to exiting a closed and sealed case and spreading fire But again the voltages seem within the range of the cell So such an event does not seem very likely

As long as OP is sure about the voltages that they are seeing it does seem like it should work

Concerns about current can be addressed with a inline resistor

Might be looking into a project like this for a 486 board. what’s everyone’s opinion on this? Any updated user testimonials? What’s the best configuration?

Sphere478 wrote on 2024-01-09, 22:56:

Might be looking into a project like this for a 486 board. what’s everyone’s opinion on this? Any updated user testimonials? What’s the best configuration?

I wish I have good answers.

Sometimes I use the flat holders, which are nice if you have the space on your board, but sometimes they overlap the edge of the board, block slots, and it's hard to get the diode in there.

The vertical holders are cheap and they take up less space, making it easier to fit the diode in there, but after you put the diode in, things can be fragile. I tried hot glue, but the tall coin slot has enough leverage to break free pretty easily.

So I don't know what to say.

Tony359 recently did some work on one of my boards and was able to identify a diode I beleieve near by that when removed stopped 5v from reaching the battery terminals. And the rtc functioned properly with a 3.3v battery. So simply removed the diode and installed a coin cell holder.

Yeah, whenever possible disabling the charging circuit is a much better option if you're trying to use a CR2032. Not only a coin cell provides a lower voltage (3V) than the original barrel (3.6V) but also the extra diode you add will lower that voltage even further, which can result in the battery not being able to power the RTC or the memory that stores the CMOS settings.