There may be other non-obvious reasons why the bus isn't exposed.
One of them may be increased EMI. To keep the noise low, you'd need
a lot of capacitors and series resistors for input- and output lines.
In addition, bus drivers might be necessary if the singnal traces become
longer. Alltogether, the board would become more expensive and larger.
I'd check all possible alternatives first before thinking about an external bus.
There's a lot of hardware available with SPI or I2c/SMBus interfaces,
including flash memories, temperature and voltage sensors, ...
All those devices can be interfaced to the uCsimm quite easily.
Carl M Holmberg[SMTP:email@example.com] wrote:
> Welcome to the world of the electrical engineer. Imagine yourself
> between classes over at Ryerson, sketching out an embedded controller
> design for which you harbor some whacked out fantasy of selling to the
> general public. Primary considerations: 1) Useful, 2) Produceable, 3)
> CHEAP. If RT-Control hadn't managed to keep the price point down at the
> ~US$200 level, there would have been little interest.There are already
> plenty of $1K devices.
> I suspect that even now a built in IDE controller would be an unwanted
> cost for most customers, and at the time the design was frozen, 72 and
> 168 pin hardware wasn't rock bottom cheap enough for a device that
> wasn't intended as a completely opened up prototyping system such as
> Motorola sells.
> There are h/w designs and software available that can be molded to
> provide IDE for the uCSIMM, and Tom Walsh has a working 72-pin design on
> openhardware.net. However, if my reading of past email and his EZ328SIMM
> page is correct, Tom's intent is only to provide the design, not to
> contract with someone to crank it out. The cost analysis of making do
> buying off the shelf vs. building the expanded design is left as an
> Richard Laxton wrote:
> > I have two fairly fundamental questions:
> > 1. Why was no IDE interface designed into the uCSIMM?
> > 2. Why was the M68K bus not exposed for use by the
> > experimenter. When you look at it the IO is pretty pitiful considering
> > the plethora of IO available at the CPU. This makes retrofitting for
> > Q1. a lot more difficult.
> > [snip]
> > The new device could have the following integrated peripherals:
> > - IDE interface prehaps with DMA support
> > - Colour LCD support (built in to the MC68VZ323)
> > - Ethernet
> > - Full MC68K bus exposed for additional peripherals. This
> > would make it easy to develop new powerful peripherals such as high
> > speed mutichannel A/D samplers or whatever.
> Carl M. Holmberg Maui High Performance Computing Center
> 808.879.5077x279 550 Lipoa Parkway, Kihei, HI 96753
> firstname.lastname@example.org email
> email@example.com spam
> This message resent by the firstname.lastname@example.org list server http://www.uClinux.com/
This message resent by the email@example.com list server http://www.uClinux.com/
This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:38 EST