Could you get us a photo of the board without anything plugged in ?

I would suggest verifying if the battery voltage gets to the chipset but since it's a bga device... that's not an option 😒

Could be corroded/dirty battery contacts on the board though...

Had similar things happen to me.
It was:
Low battery.
Corrupted BIOS chip.
And a BIOS chip pin that was broken.

Sorry i havent had time to pull the rig apart , heres a quick pciture of the board

Corroded contacts have been mentioned. Another possibility is something going to ground that isn't supposed to. I once had a system that would clear the CMOS every time I moved it, and it turned out that there was a motherboard standoff with no corresponding screw hole - move the case and it would ground the CMOS circuit. If the contact had been a little bit tighter it would have been unable to retain the CMOS settings just as yours is doing.

Make sure the JBAT jumper is in the proper position.

http://www.win3x.org/uh19/motherboard/show/3670
Comparing the manual and your board it appears it is (on pins 1-2) but you never know.

Yeah so I've been poking around a bit and can confirm it's not the bios , I swapped the bios with the duplicate board and had the same issue.
I'm sure it's not the JBAT either since it wont boot if its not set properly.

Two possibilities might be either corroded contacts , I've considered de - soldering the battery holder because it looks like there might be some green stuff on the pins.
Or the Diodes might be faulty since the other duplicate board gives a reading of 0.60V and this one does not , but I'm not sure whether it should be reading any volts or not.

There's definitely corrosion on that pin and the via that the trace disappears into.

AppleSauce wrote on 2021-04-30, 14:16:

Two possibilities might be either corroded contacts , I've considered de - soldering the battery holder because it looks like there might be some green stuff on the pins.
Or the Diodes might be faulty since the other duplicate board gives a reading of 0.60V and this one does not , but I'm not sure whether it should be reading any volts or not.

Assuming the diodes are there to stop current flowing to the battery when the PSU is on then, when the PSU is off, they should have a voltage across them, 0.6V sounds about right. If there's no voltage then either they've failed short, or they're not connected to the battery (broken trace?), or the battery is flat. From the photo, I'd guess broken trace. Try, with the PUS off and battery out, measuring the resistance from battery holder terminals to the diode terminal on the replacement board and the problem one. You might find a connection that's 0 ohms on the replacement and open circuit on the problem one.

So I did some more testing and found a main trace under the battery that had no continuity on the working board with the battery problem ,while the dead board did.
After a temporary bodge wire up with alligator clips I connected in a battery and found the diode was now finally getting voltage but 0.51v instead of 0.60v , that might be because of the wire length or the alligator clip setup.

So I'm going with corrosion being the culprit and not the diodes.
I think I'll desolder the battery holder to see if there's any damage on the top of the board and after that ill solder in a doner battery holder.
If it still wont work I'll solder in a shorter bodge wire to link the damaged traces.

Okay so I managed to take off the battery holder and luckily there wasn't much going on only some corrosion on the via leading to the other corroded bit , I cleaned that up.
So I soldered the holder back and then soldered a bodge wire on.
It seems to be keeping the date now so hopefully it's all sorted , ill do some more stress testing though to make sure nothing else is wrong with the rest of the board.

Just a small update ,
motherboard seems to be working pretty well.
I've installed all the expansion cards and other bits and bobs and tested some games.
Very happy that the solution was a simple one in the end.