it will not be Y2K compliant so have you tried a date like November 24 1999 ?

No, it doesn't even save the BIOS settings.

You need a diode and have most likely installed the battery not correctly. Show a picture of what you did.

derSammler wrote:

You need a diode and have most likely installed the battery not correctly. Show a picture of what you did.

Should I fill D2/D3 and R22? 😕

Have you tried a jumper on the middle pins of the Ext. Battery connector ? Many of my older 286, 386 and 486 boards that have onboard battery need 2-3 jumped to use the onboard battery, it is removed for connecting external battery. You should not have to add any components to the board IF it originally had an onboard battery.

Your board did not have a rechargeable battery. The battery's positive goes to an empty spot for a diode. You can use 1N4148 for the diodes and you also need to populate R22 but I don't know the value. That being said, why do you insist on a barrel battery? That's destruction waiting to happen...

Wow thanks quicknick, yeah just noticed it is a Ni-MH which is never used in computers AFAIK. Those diodes and resistors would only help to recharge so the battery must not be charged at all or else it would still hold cmos (if a Lithium or Alkaline was used then no diodes/resistor needed). Trickle charge rate for NiMh is typically 0.05-0.1 of it's max ouput so for a 60mAh it would be 3-6 mA. Assuming ~3.6-3.8v after diodes then a 1k resistor would work. All the images but one of that board (Google) show no on board battery so most were used with external. just mumbling on...

Horun wrote:

Wow thanks quicknick, yeah just noticed it is a Ni-MH which is never used in computers AFAIK. Those diodes and resistors would only help to recharge so the battery must not be charged at all or else it would still hold cmos (if a Lithium or Alkaline was used then no diodes/resistor needed). Trickle charge rate for NiMh is typically 0.05-0.1 of it's max ouput so for a 60mAh it would be 3-6 mA. Assuming ~3.6-3.8v after diodes then a 1k resistor would work. All the images but one of that board (Google) show no on board battery so most were used with external. just mumbling on...

Are you sure about the 10k resistor value?
It seems those components aren't just for battery charging but for connecting the battery to the board.

One thing we know for sure: the board had a battery, whether external or onboard. CMOS typically draw 4 to 10uA to keep settings so even if a 1k resistor is used it would supply 3.6mA max to the cmos which is more than enough to keep it alive and not damage it.

D2, D3 and R22 are gonna be necessary. Diodes should have lowest possible voltage drop and resistor is something between 150...270ohms, setting the charge current for the battery.

Tiido that would create a 20mA charge rate which is too high for a 60mA NiMH according to all I have read about. Here is a good article:
According to this: https://www.electronics-notes.com/articles/el … de-charging.php

That's the resistor values seen on all the existing boards I have had access to, most have 150ohms, some have 220 and one had 270ohms. They did have NiCd batteries originally though.

@RiP - it's a long time ago, but meanwhile I got the same problem. Did you added the 2 Diodes and the resistor, and if, what resistor value have you taken (e.g. 270 ohms... but I would start with the highest to be considered value always) ?

Peter z80.eu wrote on 2022-12-06, 14:16:

@RiP - it's a long time ago, but meanwhile I got the same problem. Did you added the 2 Diodes and the resistor, and if, what resistor value have you taken (e.g. 270 ohms... but I would start with the highest to be considered value always) ?

Yes, I used 1k resistor and it worked.