Poll, new accelerator for A500/A600.

[Doobrey]

hppacito wrote:

I was also thinking in replacing the original Kickstart with a 32 bit version (read, with 4x8 bit epros or 2x16 bit eproms), for even faster access. That should need to get *ROMEN from the old ROM socket, or autogenerate it.

I think taking *ROMEN from the socket is the only way,otherwise it won't boot. (No way of knowing if the rom overlay is enabled for mapping  0x00000000 to 0x00f80000 )

[Doobrey]

hppacito wrote:
 If the new eproms are fast enough (could also be flash), that can be a 3 cycle access, or a 4 cycle access. A shift register can be used to generate the delayed signal.

Just another brain fart..

 If you're gonna use faster eproms/flash, then you need to get the /OVR signal to stop Gayle automatically generating DTACK for the old ROM timing. AFAIK, on an A600 that means the old 'upside down PLCC socket' over Gayle.. which kinda blocks Jens' A600 clockport adapter  :-(
 Anyway, if you use the socket method it'll be easy to take /SPARE_CS and /INT_6 and add a clockport header on your board. (Or for the grand cost of another 50p, use a 2-4 decoder and have upto 4 clockports)

Also, ethernet should be pretty easy too. Use one of the clockports from above, (maybe replacing /SPARE_CS with /NET_CS for quicker read/writes) and use a CS8900 or RTL8019.  EDTP sell kits/modules for both for around $35 that only need INT inverted and the small matter of writing a SANA2 driver.

[Doobrey]

xeron wrote:

Whats to stop you using some of the reserved areas for fast RAM?

The  $e00000 is for the 1st half of a 1MB ROM (No idea why CBM split it that way) and I think Gayle generates a /ROMEN in that range.
 The $f00000 range is for 'diagnostic' flash rom.. Exec checks this space and boots from it instead of the normal Kickstart if a rom is found here, again Gayle generates the /FLASH chip select. This space is also scanned for residents on startup too, so it could prove to be a simple way to put in drivers for an IDE interface on the A500 version or parts of the 3.9 RomUpdate.

As for the $a00000 my brain's drawn a blank, but I got a sneaky feeling it's another unused rom space on the CD32 (all I can remember is that the CD32's exec scanned another area for residents that none of the other kickstarts did.).. but then again I also remember someone telling me it was another place that Zorro II cards can have there I/O space mapped to.

[Doobrey]

Try ExecDis.lha, it's a commented dissassembly of the 1.2 Exec, but the startup hasn't really changed much over the years.
 The only noticable startup differences on 3.x are an added check for a ROM in the PCMCIA slot that can take over instead of the Kickstart (although I've been told that part is broken and it fails to see a ROM), and the A600/A1200 doesn't use the CIA to turn off the ROM overlay as it's controllable by writing to Gayle.

[Doobrey]

hppacito wrote:

So far, no luck with usb host controllers. So no USB for the time being. I think that may be USB can be done completely offchip.

Wouldn't it be easier to just include an A1200 compatible clockport header. That way, users can just add a Subway card if they want USB. (Plus the board will be cheaper for those that don't want it).

 BTW, last time I checked, RS Components stocked 2mm IDC headers, also the Jelu Web Shop has right angled 44 pin headers for 1.21 euros.