exxos's Atari, Amiga & retro forum

Okay, this nut has been cracked. The 68000 and the TF536 have different requirements to sit nicely on the bus and let the other one run the system.

BGACK on TF536 grounded, BG, BR, and E floating.

BG and BR on 68000 grounded, BGACK and E floating.


So, this will take two muxes to build, but that's not a problem. They're cheap and small. And I'll probably go ahead and ground some of the things that I'm currently leaving floating, because it seems better form, even if not strictly necessary.

Well done @derkom , this sounds like a mod that would be cool for any ST, let alone your Stacy.

Grounding, grounding, grounding.... Wonder why that always seems to be
a big issue with so many ST's (and the STacy especially)?

The Pak 68/3 documentation emphasizes it big time. I think I have 3-4
grounds coming off my Pak board to various places on the STacy's MB.

Darklord wrote: 04 Aug 2020 19:48 Grounding, grounding, grounding.... Wonder why that always seems to be
a big issue with so many ST's (and the STacy especially)?

The Pak 68/3 documentation emphasizes it big time. I think I have 3-4
grounds coming off my Pak board to various places on the STacy's MB.

I haven't actually run into any grounding issues that I'm aware of. The question here is whether the 68K and/or TF536 care if the disconnected pins are grounded or not. Makes sense to me that this matters with BR, since we want the disabled device to be certain it's not allowed to talk. In the hunting for a workable configuration today, it appears that the other pins can just be allowed to float, and indeed in the case of the c't PAK switcher, E and BG are allowed to float, and it's apparently fine.

I haven't laid out the schematics yet for doing this with two muxes, so I don't yet know how the pins will work (since the muxes are directional, only certain arrangements will work). I'm deliberately saving this task and PCB layout to do on my summer holiday, because that's better than going to the beach. :lol:

derkom wrote: 04 Aug 2020 20:59 and indeed in the case of the c't PAK switcher, E and BG are allowed to float, and it's apparently fine.

Considering that BG and E are both outputs from the CPU, grounding them would indeed not be a wise idea.

czietz wrote: 04 Aug 2020 21:47 Considering that BG and E are both outputs from the CPU, grounding them would indeed not be a wise idea.

True, and indeed now that I went back to check, the TF will boot fine with BG floating on the 68K. So it looks like the only things that actually need to be grounded are BGACK on the TF when booting the 68K or BR on the 68K when booting the TF. The rest can all float.

Very impressive! You're a wizard @derkom :)

Indeed, well done! :)

I just sent off the new switcher board design to Aisler today. Hopefully arriving in my mailbox around the time I get home from summer holiday. Will give me something to do during my quarantine. :lol:

Guess this is turning into my autumn project now. :lol:

Progress has been made! I got my new switcher boards and built one up:


Yes, my design calls for a SMT pullup resistor and I discovered I don't have anything appropriate. I will acquire appropriate components before my final build. :D


Still looking pretty science-projecty when installed:


The pair of wires coming out toward the front and stuck in foam are connected to the switch terminals on the board, thus switchable via jumper right now. And here's what we get!


:excited: :girldance: :thumbup: