RE: Article PC-card

jbevren_at_starbase.globalpc.net
Date: 2001-02-26 07:14:21

On Sun, 25 Feb 101 g.baltissen@hccnet.nl wrote:

> Hallo Nick,
> 
> > >Yes. But also tristates the address- and databus. And that is something
> > that
> > >does not happen with the PC-bus. And I'm very sure not even wanted.
> 
> > That's right, you don't want it... then 8237DMA controller is on the PC 
> > mainboard and though the CPU is tri-stated, it controls read/writes 
> > from/to IO/MEM.... if the concept is to have a PC card plug into the C64, 
> > then someone has to be awake to make the transfer else a DMA controller 
> > will be needed between the two.
> 
> I'm sorry, either you do not understand it, or I don't understand what you 
> wrote. So for sure I repeat everything:
> - the IOCKRDY line is to halt the CPU and nothing more. The CPU stops 
> processing but address- and databus reamin active. My comment was that the 

IOCKRDY is not used for dma.  Its for stallin ghte cpu for slow devices.

> 6502 has this feature as well. But the trouble was that this input could 
> not be used as it was not connected to the expansionbus.
> Marko commented to use the DMA-input which resulted into my comment that 
> this input could _NOT_ be used as using this input would not only result in 
> stopping the CPU but also tristating the address-and databus. The IOCKRDY 
> input was meant to be used by slow memory or I/O. You surely don't help the 
> poor slow device by tristating the databus during a read....
> 
> As I understand, you say that we need a DMA controller before we can use a 
> PC card? Why? The MGP card for example does not use DMA. And As I already 

If your target card doesnt use dma, you dont need a dmac.

> said, it run fine with my C64 from 1989 on. DMA is only needed with 
> floppies, harddisks and networkcards, at least for the PC. I already proved 
> that the C64 doesn't need one for a IDE-disk (OK, speed is not "je-van-het" 
> but...). I don't see any reason why it should not work for a 360KB floppy, 
> for the 1.44 I have my doubts.

The FDC does use dma, but it doesnt need to.  Again, I recommend using dma
whenever possible. :)

UHS (my scsi project, see http://uhs.c64.org ) will have to deal with the
PC-DMA->c64 issue.  i've decided to either make a very simple DMAC or use an
8237a and allow others to use the remaining 3 dma channels via a
shortbus-style cabling setup.

Side note: I'm working on matchign ISA signals with the c64's bus (as well
as what I blieve is likely on the supercpu's ram socket), so that one can
use an ISA bus with thet c64 and/or scpu.  This involves having just a PIC
and DMAC set, and no other chipset involved.  There would be no x86 cpu on
the bus, which means it would require a passive backplane.  

COnveniently enough, the local junk store has 7slot backplanes for $5 each
;)

comments?

-
This message was sent through the cbm-hackers mailing list.
To unsubscribe: echo unsubscribe | mail cbm-hackers-request@dot.tml.hut.fi.

Archive generated by hypermail 2.1.1.