Amiga 3000T need help troubleshooting

[Matt_H]

Just throwing out a few guesses here...

This looks an awful lot like what my 4000T was doing when my accelerator wasn't seated correctly. Double/triple/quadruple check that it's fully seated (and make sure all the standoffs/posts are in place to support it--and, though unlikely, that you're not experiencing this issue). Push down on the connector with a pencil eraser to ensure a tight fit.

It looks like it's crashing very early in the boot process, but can you get into the Early Startup menu (hold both mouse buttons)? What about the Cyberstorm menu (Esc key, I think)?

Can you confirm that your machine has an onboard 030? (Worth checking because some 3000Ts don't.) If so, try again without the accelerator and double check the motherboard jumper settings for the onboard CPU.

Have you tried disconnecting the hard drive? OS4 is extremely finicky and I wonder if something it installed in the RDB (e.g., a filesystem) is causing the machine to crash. I suspect it's not this because it looks like it's crashing before it even tries to access the drive, but you never know...

[Matt_H]

This won't fix the problem, but reconnect the floppy drive(s) as you continue testing. That may seem counterintuitive, but a barebones Amiga configuration expects a floppy drive to be connected. Not having a floppy drive produces its own set of odd symptoms, so keeping the floppies connected will prevent those symptoms from getting mixed up with the ones you're already experiencing.

It's a longshot, but consider removing the motherboard to see if there's any debris on the underside that may be causing a short, or if a solder point on the underside has punctured the insulation shield and is shorting against the case.

Capacitor failure is less likely on the 500/2000/3000(T) compared to the 600/1200/4000, but it is a possibility. If you're removing the motherboard, it's worth checking them with a multimeter.

[Matt_H]

This is the strangest hardware problem I've ever seen. I have absolutely no idea what the problem could be.
I've never heard of a Paula failure before, but it's an easy chip to replace so it might be worth trying on the remote chance that it fixes things.