exxos's Atari, Amiga & retro forum

@PhilC Mine's currently in the bottom right one...I was worried about it blocking the floppy cable (my H4 is using the Gotek in an external drive case until I properly sort out the FDD power cable...another job on the To-Do list... :lol: :roll: )

I'll try the new EmuTOS .prg file from @agranlund in a few minutes & see what time I have left for experimenting with the location ...

I think that there's no way to have the top shield on with the TF536 anyway, looking at the clearances. And the RAM expansion (which I'll probably swap with an Exxos one and disable the on-board RAM) has some height to it too.
How did you get the CPU connection over to that side of the board @JezC ?

If ram works fine on the ribbon cable, would be possible to use it also to relocate TF536?

Cyprian wrote: 21 Sep 2020 22:40 If ram works fine on the ribbon cable, would be possible to use it also to relocate TF536?

You'd need to have two 32-way cables at least. The biggest problem would be with the clock I imagine.

There's a lot more going on with the CPU relative to the RAM.

stephen_usher wrote: 21 Sep 2020 22:46

Cyprian wrote: 21 Sep 2020 22:40 If ram works fine on the ribbon cable, would be possible to use it also to relocate TF536?

You'd need to have two 32-way cables at least. The biggest problem would be with the clock I imagine.

There's a lot more going on with the CPU relative to the RAM.

Not to mention how close to the edge the operation of many of the old machines is nowadays, even in stock configuration... :hide:

Without the top shield, I wonder if the TF536 could sit over the top of the shifter can. It might need a socket tower to get it up there though.. but then again the relocator/decode/ROM board could probably provide the height required.

PhilC wrote: 21 Sep 2020 20:48 That and I liked @derkom's idea to keep both CPUs inside the case to switch between them.

Since the design seems to work well at this point, I'll stick the KiCad files in my thread. The physical PCB is obviously Stacy-specific, but the electrical layout should transfer fine to any similar projects.

Edit: Here it is.

Well, having removed the Marpet RAM upgrade I can see better. There would be space in front of the power supply for the TF536 and to have a relocator board based upon @derkom 's with the 68000 and also maybe the extra decode etc. circuitry too.

It would require the socket for the TF536 to overlap the pins into the current CPU location.
If the ROM is onboard, then it doesn't matter that it overlaps the ROM sockets.

What do people think?

I've just realised that I've got the CPU representing the TF536 the wrong way around. I'm guessing the routing's going to be hell.

Out of interest, how does your decoder work, @exxos ?

Does it initially just trap addresses with A4 to A23 all zeros and change the address to 0x0e000000 and then set the correct signal (BGACK?) to the CPU when the ROM address range is accessed and enabling the ROM CE?

I’m trying to think of the simplest way of doing the decode.

I just decode the whole address bus and address zero for ROM access. Think other decoders still use the GLUE to decode address zero, and not sure if they decode more than just the ROM area or not.

Have you considered @cmorley USB ROM board, or @DoG decoder.. would just seem easier than reinventing the thing again.