* David@ixibbs.iximd.com <David@ixibbs.iximd.com> on Mon, Apr 30, 2001:

> First, IDE Doesn't require (the way I've been told) an adapter in the
> way that SCSI does.  That IDE is linked to the bus directly.  I think
> this is wrong myself though.
> 
> If you look at this schematic:
> http://members.nbci.com/_bitmaster_/dc/dc-ide.pdf
> You will see that the only thing he has hooked up to the DC-Port is an
> IDE port, which means that IDE is in there if you want to try an atapi
> based device.  It also has some 5v and 3v lines coming out of the
> DC-Port which means you could power a ZIP drive (Why I say this is
> because my SCSI zip drive is External with a 5 volt power adapter. AND I
> also seen IDE ATAPI BASED ZIP drives that go in computers in a 3 1/2 in.
> bay which means ANYONE who wants to try would most likely succeed.
> 
> The problem that I see is that Harddrives need a bios of somesort to
> actually run (unless I'm wrong).  Atapi drives like CD-ROM's and ZIP
> drives don't need a BIOS on the controller.
> 
> Can someone tell me how you have the GD-ROM driver working?  I mean is
> there a way to find out if it is SCSI or IDE? (It has to be one or the
> other doesn't it)
> 
> At this site he also has a Pin-Out of a SoCalled CD-ROM port:
> members.nbci.com/_bitmaster_/dc-index.htm
> 
> Would this be the actual port that the GD-ROM is hooked to?
> 

Ok, since some of us like to dance around in circles, let me try to
summarize this thread the best that I can:

- Going through the port-dreamcast list archives (which I'd advise anyone
  to do before asking when and whether the DC modem will be supported),
  I've seen Marcus Comstedt say that he's unsure whether or not the GD-ROM
  is ATAPI-based, as the command numbers don't correspond to standard
  ATAPI.  Obviously, Marcus C. is the best one to speak on this, and given
  his track record on DC reverse engineering, 9 times out of 10 he's
  correct (hehe, I only subtract a point cause I often find my self
  duplicating some of his efforts - I wish I could have at those notes! ;-).

- In my own RE sessions with WinCE-based code, I've seen references (in the
  form of strings) that suggest the GD-ROM is ATAPI-based.  I have yet to
  confirm this == grain of salt.

- bITmASTER's IDE interface has a ton of glue before connecting with the G2
  bus via the DC's expansion adapter.  This glue probably does the
  necessary address decoding so that you can access registers on the IDE
  device.  No where on his site does bITmASTER give specifics so you know
  what?  Before making vague assumptions about aspects hardware of which you
  know nothing about, why don't you just ask bITmASTER?  Fair warning: I've
  e-mailed him before, and it may take him a few weeks to respond, but who
  better to ask than the man who did it?

Truth is, no one in the hobbyist circle doesn't know (yet) if the DC has an
internal ATAPI interface - so idle speculation is just that.

Facts:

- The DC expansion port does *not* sport a IDE/ATAPI interface in any way,
  shape, or form - just ask Marcus C. and Dan Potter, some of whom have
  probed it long enough.

- No one except Sega and Iomega know how the ill-fated ZIP drive was
  connected, but for an idea of how Sega did interfaces for devices that
  *do* exist, e.g. the BBA and Modem, both have glue chips that map the
  device's PIO/whatever into the SH-4 address space, e.g. there is no true
  PCI bus within the DC, it only exists on the BBA.

I've ranted like this to say: there are many others who've spent many long
hours RE'ing the hell out of the DC and related software, not to say that
my black box still isn't a mystery, but speculation for the hell of it
doesn't help at all and makes one question the "clue level" of those who
would do so.  If you want to know something and don't like the answer you
get, go figure it out for yourself.

Hell, I make wrong guesses on DC hardware all the time, but I correct
myself by probing the hardware or asking others who've already done it -
and the resident guru is Marcus C.

Cool? ;)

M. R.