Not-so-dead DX7 Mark I - The saga continues...

Hi everyone - Just starting a new thread as the older one hit a dead end; not sure if people are actively checking this forum.
 

 Thankfully, my DX7 with E! is no longer dead. I have things partially working - and I think the answer was that the E! needed to be reseated and the contacts cleaned. But now, something else is going on - as the battery replacement seems to have zero effect.
 

 When it's been sitting, when I power it up, I get the "RAM is Corrupt!" error message from E! -- and I've lost all the settings. Plus, E! seems to behave a bit erratically with different symptoms each time (no MIDI, missing internal banks 8-10, etc.). 
 

 If I keep it on and wait for the LED digit to change to "01" (maybe 4-5 seconds?), I can shut it off and turn it on again a second time - and then I won't get the same error message. But then, of course I've lost all the settings as I've said.
 

 So I'm thinking that the RAM and E! board may actually be fine, but somehow the first power-up doesn't have enough voltage to get started...and the second power-up then has enough voltage to get things to get rolling...but by then either the backup battery isn't applying enough voltage in there to keep everything stored in RAM - or it's losing the current elsewhere. BTW, the backup battery voltage appeared to be about 4.5V if I remember (as measured by E!), and the CR2032 battery was 3.5V. 
 

 So I am wondering if the E! board may be fine, but the drop in voltage is screwing up E! with strange results upon rebooting.
 

 @aerobaticant - Are you saying that you think the diode may be the reason why it's losing the voltage upon power-down?
 

 
 Thanks to anyone who has any suggestion as to what to test or replace next!