exxos's Atari, Amiga & retro forum

@kodak80 Nice compact card that. Be good to see what results you get with it

Raven board #1 is now being cased. White one for #1 and #2 is already in a back case.

Also updated, all chips to the latest release in preparation for testing video card.

I updated Raven #1 with latest firmwares to allow me to start testing some Cirrus video cards I purchased but it currently freezes after a couple of minutes use, regardless of which video card I am using, I switched back to an ET4000AX and it still freezing.

This raven was running an LC CPU and I purchased a full 68060 CPU which seemed to be when the freezing started. I switched back to the LC CPU thinking it was the new CPU which was faulty but after sending the CPU back to the seller, found that the system is still freezing.

I have not had time to diagnose it further to figure out if it is the newer firmware or something hardware wise has gone faulty. Could the new CPU have damaged something?

At least Raven #2 is working. Just need to find a permanent place in my room to set it up now it is cased.

PhilC wrote: 13 Oct 2025 22:48 @JezC I thought @kodak80 had made an ISA cart port board? It would just be a case of adding support for it on the Raven.

@JezC @PhilC

Moving this from the MC68060 post.

Yes, the ISA cartridge PCB I designed for the MicroATX ISA slot so may be able to work with the Raven. In the MicroATX the board needed ROM3 and ROM4 from the GLUE chip. Without a GLUE in the Raven we would need to be manually generate these which was mentioned here: https://atari-forum.com/posting.php?mod ... 0f69fdc20d by Arne

Here is what was advised:
ROM3 & ROM4 can be generated from /AS, R/W and the upper 8 bit of the address bus
I do not fully understand how this could be done.

We would need some additional hardware (e.g. GAL, CPLD, or discrete logic), with:
- Inverters
- AND gates to combine the signals
- Possibly a programmable logic device (like GAL16V8 or ATF22V10) for compactness

This is not at all thought through properly, just thinking out loud :)
But we can steer the logical cartridge address range to an ISA-MEM access with the PMMU.
Would "just" need to avoid that the (isa-space) range isn't colliding with what a graphics card may think belongs to it.
Possibly by A0-A19 being lower than $A0000, and then also some quite high address bit(s) being set to put it out of the way of potential linear framebuffers.

Something like Ax signals TBD, perhaps at ISA space 8MB or something like that.
(cpu) 0xFA0000 -> (isa) 0x800000
(cpu) 0xFB0000 -> (isa) 0x810000

The card would need some logic, similar to a normal ISA card, to decide if the access is for it. And then also to generate some of the signals for the cartridge port.

Cool, sounds like a possibility then, @JezC There you go, build yourself a Raven and you've already got yourself a little project ;)

PhilC wrote: 14 Oct 2025 14:03 Cool, sounds like a possibility then, @JezC There you go, build yourself a Raven and you've already got yourself a little project ;)

Thanks @PhilC ... I think...?

I could maybe start planning to do this in early 2027 (and then actually start on it before the end of 2028).... :roll: :?

Edit : if we start planning for CL 5 early, then I for one would be interested to see the ST ATX you built!

JezC wrote: 14 Oct 2025 14:30 Edit : if we start planning for CL 5 early,

:dizzy:

JezC wrote: 14 Oct 2025 14:30 then I for one would be interested to see the ST ATX you built!

Someone was doing a H5 ATX port but seemed to have not got completed now, can't remember who was doing it now.

exxos wrote: 14 Oct 2025 14:40

JezC wrote: 14 Oct 2025 14:30 Edit : if we start planning for CL 5 early,

:dizzy:

JezC wrote: 14 Oct 2025 14:30 then I for one would be interested to see the ST ATX you built!

Someone was doing a H5 ATX port but seemed to have not got completed now, can't remember who was doing it now.

I'm now expecting @PhilC to suggest that as another project for me... :lol: :roll:

Nah @JezC I'm not recommended anything involving jumpers for you :lol: