RAM expansion schematics

[McArti0]

Bryce.

NO. TotO simply confused the supply BUS when drawing the schem.
I showed above.

[McArti0]

so there's no misunderstanding.

[GUNHED]

Do you want to shadowed rambank 1?

The Inicron RAM expansion allows to either use the second 64 KB of the CPC or the buffered S-RAM of the RAM-Box. A switch can select it.

[McArti0]

I don't understand at all why the CPC 464 doesn't allways write to the internal RAM when the RAM extension is connected. After all, RAMDIS blocks read only. GA does not know if additional ram is connected. GA should always write to rambank0. Does anyone know how this happens?

[IanS]

I don't understand at all why the CPC 464 doesn't allways write to the internal RAM when the RAM extension is connected. After all, RAMDIS blocks read only. GA does not know if additional ram is connected. GA should always write to rambank0. Does anyone know how this happens?

In the special case '**' marked in column '011' internal memory at 0x4000-0x7FFF is remapped to 0xC000-0xFFFF. This is handled internally in the CPC6128 and plus machines. The 464 and 664 don't perform this remapping. Further, the 464 and 664 do not write protect internal RAM when external memory is accessed. On the original DK-Tronics expansions these two issues were handled by backdriving the A15 and MREQ* signals on the bus, overdriving the values from the Z80 CPU in an electrical conflict.

from https://github.com/revaldinho/cpc_ram_expansion/wiki/'Old-School'-CPC6128-512KB-and-1MB-RAM-Expansion-Cards

[McArti0]

High MREQ electrical forcing agains Z80? Did someone want to kill the Z80? 
It is dangerous to use RAM extensions for 464 and 664. 

[eto]

High MREQ electrical forcing agains Z80? Did someone want to kill the Z80? 
It is dangerous to use RAM extensions for 464 and 664. 

obviously it isn't. This trick is used for 38 years and no Z80 were killed yet. 

[McArti0]

OMG. I didn't know that was possible! 
Ok. Mystery solved. Thanks!