Re: 68008 co-processor board complete working design available ...

[Jim Gilbert <brokerjim@gmail.com>]

On Nov 10, 9:48 am, "sfahey" <sfa...@a2central.com.remove-6n6-this>
wrote:
>   To: Jim Gilbert
>   Re: Re: 68008 co-processor board complete working design available ...
>   By: Jim Gilbert to comp.sys.apple2 on Tue Nov 09 2010 03:43 pm
>
>  > This is just the beginning of the technical information available.
>
> What is the board's application/function?
>
> Sean Faheywww.a2central.com

The board possessed substantially more memory and a much faster 32-bit
processor than the hosting Apple II; it also had a full-function
serial port with full modem control.  I had no particular application/
functional target other than to build a substantially powerful Apple
II co-processor using a bunch of parts that were sampled to me for
free by my Motorola, Intel, and Mostek sales reps.

It required programming in 68000 assembly language with knowledge of
its deeply buffered interfaces to the host processor.  I used two
trivial applications to convince myself it was working properly: (1) a
RAM Disk providing fast storage for the host Apple II (that would
evaporate when the power did, unlike a real disk drive), and (2) a
print spooler using the RAM to allow the host to run far ahead while
thinking it was printing large jobs.

The primary purpose was for a recovering systems programmer to
convince himself he could design, build, and debug a substantial board-
level computer.  I learned about many aspects of logic design, layout,
wirewrapping, use of discrete components, etc.  All the Apple and
68000 assembly language and the ROM monitor were pretty easy since I
had spent 15 years as a professional systems programmer.

I implemented this board after having completely understood my
Microsoft Softcard by deriving the schematic from the traces, studying
the datasheets, disassembling the software, etc.  My board was based
on a Motorola Application Note by Geoff Brown demonstrating what could
be done with a 68008 and a 68901 multi-function peripheral. The main
feature my board shares with Geoff's design was software refresh of
the dynamic RAM propelled by the timers and interrupt controller
integrated into the 68901.

If I were doing the board today I would have probably integrated most
of the TTL into a CPLD or FPGA and have included an Ethernet
controller.

This whole thing began when my Motorola sales rep gave me a 68008 in a
plastic tube after I had asked a curiousity question about it.  Then
came some design notes. And I asked him about a 68901 which appeared.
And my Intel guy gave me a pair of 64Kx4 FP-DRAMs, so I asked him for
another pair.  And I think I got the FIFOs from spares for one our our
printer controllers (I was manager of Operating Systems and Networking
for FileNet at the time).