I see only 2 diodes off the battery area. I would use 3 x AA alkaline in EXT battery pack (for 4.8v+ when new, minus about 1.2v for the 2 diodes gives about 3.6) so is about right.
Have seen boards that require 6v external and they have more diodes than 2 off the Ext battery. Now your board may require a bit more but that is where I would start.
If when you save setting to CMOS or the time does not hold then I would consider going to 4 AA. Better to not over volt the Ext battery than over volt it on an old board.
Use a good Digi Volt meter and make sure pin 1 is + and pin 4 is ground...

That's good to hear. I was thinking exactly what you wrote and was thinking this weekend trying it. Then from what I have seen just removing the pin 2&3 pin jumper that disables the charging so I shouldn't need a diode in there for the battery pack. I am thinking I can do a quick test without the battery and jumper and just test to see if pin one or 4, which ever one I confirm as positive, should not have voltage with the jumper removed and board powered on.
Guess I should find me a battery holder so I can make this happen.

If you want one in a hurry, you can try a dollar store, look for small LED light strings with a battery box.

Well finally got the stuff together to get rid of this ticking time bomb nicad. Checked out the 4 pin connector pin 1 had 0.6 volts pin 2 was internal battery positive at 3.8 volts and pin 3 was charging power had 4.6 volts when computer on. (Took jumper off pins 2 and 3 for these tests) pin 1 battery neg. So after checking out pin 1 it went to the diodes / resistors so figured it surely was for external positive. The 0.6v would just be leakage either way no actually charging coming back through so I hooked up 3xaaa battery pack to pin 1 and 4 kept the jumper off pins 2 and 3, and tested. So far so good CMOS kept settings after 20min unplugged. Calling it a win for now.

curggles@gmail.com wrote on 2023-04-04, 00:52:

Well finally got the stuff together to get rid of this ticking time bomb nicad. Checked out the 4 pin connector pin 1 had 0.6 volts pin 2 was internal battery positive at 3.8 volts and pin 3 was charging power had 4.6 volts when computer on. (Took jumper off pins 2 and 3 for these tests) pin 1 battery neg. So after checking out pin 1 it went to the diodes / resistors so figured it surely was for external positive. The 0.6v would just be leakage either way no actually charging coming back through so I hooked up 3xaaa battery pack to pin 1 and 4 kept the jumper off pins 2 and 3, and tested. So far so good CMOS kept settings after 20min unplugged. Calling it a win for now.

Good job !