Hallo Bill, > The 6545 chip (Cathode Ray Tube Controller) in this picture > is hot to the touch, probably is bad. I am currently searching > for a replacement. I know some ICs can become very warm but IMHO the 6545 is not one of them. I don't have a spare either but this could do the trick: Two years ago I have been experimenting with the Motorola 6845 wondering why that one doesn't work inside a 8032. And to be honest, I still haven't any idea. But during these experiments I took a Hitachi 6845 out of the box and suprise.... that one worked fine! And so did the UMC version. All three worked fine on a PC MGP card but all Motorola 6845's refused to work on a 8032. Having a reliable source of 6545 replacements, I quit the research. Remark: I never tested it on a 8296. > Note also that the processor is a 6502 A, not a "B" like other > 8296's. Would be slower? No. The A means it is the 2 MHz version, the B means 3 MHz version. This means they can run reliable at these frequencies. The 8032 is only a 1 MHz system so the 6502A should run very fine. Remark: I wonder if the A, and especially the B, is made for these frequencies or are they justed tested. A 80486-33 running on 25 MHz is much cooler then the original 80486-25. One idea of reducing the heat output in, for example a 1541, could be replacing the parts with their A or B versions. > 2) could the display be causing the problem? You failed to mention the bee-de-bee-de-bee beep at start-up. So I presume that it didn't occur. And that isn't a good sign either. > 3) Any diagnostic ideas? All the brooken 8296's I met so far had all the same major problem: bad PLA's. Your luck: replacements exist, IIRC Jens Sch”nfeld sells them. There even exists an EPROM replacement, no URL at hand. A not so hi-tech solution: start swapping parts with another 8296. In some cases a 8032 will do as well (6522, 6502, 6545, 6520). The B700 won't do as it is 2 MHz IIRC. For the next one you need an osciloscope: replace the upper ROM with an EPROM filled with NOP's and JMP $F000 at $FFF7. This should produce square waves at the address lines A0..11. Taking this idea a step further: use a EPROM filled with a diagnostics program, one that outputs the result at the userport. Problem: I don't have such a ROM and if others cannot help you, you must program your own one. A hi-tech solution: I myself repaired quite some 6502 machines using my debugger,. I simply compared what the 6502 was doing against the source code. That enabled me to detect bad RAM, bad 6522's etc. A logic-analyser will do as well of course but I don't think you have that laying around somewhere (me neither). Good luck!