Re: Apple iigs ADB fails

[dempson@actrix.gen.nz (David Empson)]

Michael J. Mahon <mjmahon@aol.com> wrote:

> Michael Black wrote:
> > "Michael J. Mahon" (mjmahon@aol.com) writes:
> > 
> >>If that fails, you will probably need to replace the ADB controller
> >>chip--for which the likely source is another (spare?) IIgs board. ;-(
> > 
> > That's not one of the crossover ICs, that a relatively early Mac could
> > supply as a transplant donor?
> 
> I'd be surprised--the IIgs was the *first* implementation of ADB,
> and I'd bet that the microcontroller firmware changed quite a bit
> by the time ADB showed up in Macs.

It was the first _released_ implementation of ADB (September 1986,
compared to early 1987 for the Mac SE and Mac II), but ADB was developed
for the Mac and then acquired by the IIgs development team, so it is
really a Mac technology which happened to get released on the Apple IIgs
first.

Despite that, it is very likely that the IIgs ADB microcontroller is a
completely different beast to the Mac one. The IIgs ADB implementation
includes a fairly large amount of support in the microcontroller to map
the ADB keyboard onto a standard Apple II keyboard interface (C000/C010
with extensions), and to map the ADB mouse into a reasonably simple
hardware interface. In the ROM 00/01 IIgs, it also includes hardware
support for scanning the IIe keyboard (in a IIe-to-IIgs upgrade).

The Mac version of the ADB controller is much simpler and relies on
software running on the 68000 processor to handle some details like
modifier keys and translating keycodes to ASCII values.

They probably use the same microcontroller, or different variants of the
same family (which I believe is a Mitsubishi embedded micro based on the
6502).  The code being run in that microcontroller is completely
different between the Mac and Apple IIgs, but they probably share some
parts of the code which deal with the actual ADB communication.

-- 
David Empson
dempson@actrix.gen.nz