On Mon, Aug 28, 2006 at 11:03:13PM +0100, Peter TB Brett wrote:
> On Monday 28 August 2006 22:07, you wrote:
>
> > There can be different external tools for different purposes. Deciding
> > to implement one particular type isn't a decision not to do something else.
> > There are many useful things a tool with an RS-232 to SPI microprocessor in
> > it could do.
<snip>
>
> Well, the point was to do away with the DIP switches entirely so as to fit
> the
> device into a *tiny* box.
That might be your point. Mine was to do the job with nothing but
DIP switches. :-)
In the U.S., some DIP switches are really cheap. I saw some in the
Digi-Key catalog priced under $1.20 for a 12-gang switch. That's a dime a
switch, or around 7 pence the last time I looked up the exchange rate.
And that goes back to what I said earlier: there can be different
tools for different purposes. If we make the on-chip hooks as simple as
possible and as general as possible, we leave open the maximum opportunity
for others to exercise their creativity. That way, somebody else can design
a tool to do things we didn't think of, or don't have time and resources to
take on.
From the SPI protocol viewpoint, we could think about tools that
look like serial EEPROMs, and tools that looks like bare shift registers.
Incidentally, re-reading one of today's earlier messages, someone
asked whether a microcontroller could send SPI at 32 MHz. It doesn't matter
whether it can or not. SPI doesn't have discrete rates the way RS-232 does,
only a maximum rate. The master sends a clock pulse with every data bit,
and it can go as slow as it wants to -- even if it's one bit on Monday, four
bits on Wednesday, and three bits on Thursday. The slave has to keep up.
But most slave devices emulate shift registers at the bit level, and can
handle an irregular incoming clock at any speed from zero to 32 MHz.
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
