[Could you all please start new threads not by replying to other threads and changing the subject line? Thanks.] > the address lines on VIC-II as used in the C64 are a mix of > multiplexed (A0-A5/A8-A13) and static (A7-A11). That's easy enough > to explain from the way the memory is set up. > > But then there is A6 which is, according to the datasheet, > multiplexed with a static '1'. From the schematics that doesn't > make any sense since it's routed through the 74LS258 just like A7 > and as soon as A6 switches to a static '1' the 74LS258 switches too > and the 2 address bits supplied by the CIA are valid. > > Is there a known reason why it was designed that way? Was it maybe > used in the CBM-II series in some way? The original VIC-II (the 6566) has A0..A13; it is used with SRAM. The 6567/6569 sacrifices two address pins for #RAS and #CAS, so it has twelve address pins left. It can be configured (at manufacture time, by changing the metal layer) to work with DRAMs with either 7 or 8 address pins. With 7 address pins to the DRAM (16kB max., just like 6566) you get A0..A6 muxed with A7..A13 as the DRAM address pins, and A7..A11 as non-muxed pins to attach to the colour RAM. The 8 address pin "mode" only changes which lines are muxed, so you end up with A0..A6 muxed with A8..A14 to DRAM, and (as before) A7..A11 to the colour RAM. But A14 does not exist, a fixed value is used instead. Any board with 8-address DRAM chips that wants the VIC-II to address more than 16kB will have to supply the extra data lines externally *anyway* (on a mux with #CAS or such); the only gotcha is if you have 8-address DRAM and do not want the VIC-II to address more than 16kB, hooking it up the simple way will make A15 = A7, likely not what you want, so you'll need to do the mux circuitry anyway (could be some fixed bank, of course). Segher Message was sent through the cbm-hackers mailing listReceived on 2012-02-20 10:00:04
Archive generated by hypermail 2.2.0.