»

[freqmax]

I saw the Braun A1000 motherboard replacement GBA1000 again. And it's a new motherboard that fits the old case but uses old chips from other motherboards. So it made me think about the FPGA implementations of Amiga (A500-Minimig, A1200-Replay).

Would a motherboard with the old chips socketed and running in clock-by-clock in sync with an FPGA implementation and wired unlike GBA1000 so it can compare bus signals in realtime be the most effective setup to reverse engineer the old chipsets ..? wired in such way that the FPGA chip(s) can be reconfigured to feed simulated input into the old custom chips in order to test special cases.

The other way to do it would be to replace the chips in an existing motherboard with a FPGA, chip by chip. But that probably is less flexible. The idea here is to straiten out even the last bits of incompatibility before the supply of custom chips dry out (just look at 6581).

[Iggy]

Sounds like what Jens did with the CloneA.

[NorthWay]

Any ETA on those die shots to get the logic decompiled?

[freqmax]

The difference is that in this setup the FPGA pokes around with the custom chip instead of just trying to replace it. It might require fewer FPGA chips and a setup that allows one to feed test stimuli into the custom chips.