This is a very similar experience to what I had when my capacitors failed on my power supply board. I ended up having to send power supply and cpu out for repairs. Good luck.

If this helps any with your troubleshooting here is what Clive did to my boards

Data East -04 CPU board (SN 91987. Jurassic Park (version 6.00, non-OEM, non-sanctioned software. Boot failure. NVRAM installed. Electrolytic corrosion through leaking capacitors): All transistors, SCRs, diodes and zeners curve-traced and tested, -- new 2n3904 transistor (Q50 – switch logic). Removed battery holder glue from circuit board. All socketed ICs removed, checked and tested, -- replaced and returned dead NVRAM chip. Installed custom SMT 8K NVRAM chip and adaptor. ROM verified. Removed electrolytic corrosion. 47uf/25v/105c capacitor (C89). 470uf/25v/105c capacitor (C157). Boot failure: Signals in, out and around board analyzed. Busses exercised. Signals checked for timing and geometry, -- new 74LS244 IC and 20-pin chip socket (6D address bus driver). Signals in, out and around board re-analyzed, -- okay. Board cleaned up. Added 3AA remote battery pack system. Two-tier full functional testing

Data East 520-5047-02 Power Supply (SN 91987. Prior repairs. Bad capacitors, some physically leaking. F4 fuse clip fractured. Incorrect fuses etc.): All fuses removed, checked and tested, -- new F4 fuse clips due to fracturing (common). ½ amp slow blow (F7) and 7 amp slow blow (F2) fuses due to incorrect types/ratings. 7812 voltage regulator (VR1). MJE15030 transistor (TR6). Bridge rectifier tested and reflowed. 2x 22uf/50/105c (C14, C15), 100uf/50v/105c plus power trace reinforced (C2), 47uf/50v/105c (C3), 150uf/160v/105c (C8), 330uf/25v/105c (C7), 1000uf/25v/105c (C1) capacitors. Header pins straightened and reflowed. Board cleaned up. Two-tier full functional testing