Hi Dieter,

BV> Jeezzz Dieter, you just write your own i2C bus stuff and embed it in
BV> whatever single-chip micro you are using...
DM> You can go one further, and dispense with the I^2C bus - it's
DM> probably an overkill protocol wise.
BV> Sure, as far as the protocol itself is concerned, it probably is
BV> overkill. I quite like the idea of a self clocking serial link over
BV> a two wire interface though. i2C is great for that. You can
BV> implement your own commands or protocol over their basic physical
BV> layer which is what I would probably do unless I wanted to talk to
BV> another outside i2C chip.
DM> Have you got nothing better to do on your holidays than make
DM> more work for us ??

Nope.

DM> Thought not.... :-((

You got that right :) Actually, I wanted to do about 20hrs chargable
stuff whilst away, but so far all the laptop has been used for is
writing to you fuckers. Every time I try to start doing some work my
brain says, "Nah, fuck it. You're on holidays, remember" :)

DM> Anyway, isn't the I^2C bus a 2 line data transfer bus, with
DM> another two needed for power ??
BV> Nope. It can be reduced to a 2-wire interface which acts as a self
BV> clocking synchronous serial link. For this it is ideal. I haven't
BV> seen the EXACT specs myself, but I dont' think you need to run power
BV> down another pair.
DM> Lemme guess, was than an article in CC-Ink at some time ???

Dunno. I think that Elektor has done more on i2C than any other
magazine which would make sense considering that Philips devised it
and Elektor cover quite a lot of Philips stuff being European and all.

DM> Do I have to go through them yet again ???

I wouldn't bother. I really can't recall seeing much on i2C in CC-Ink
and even the bit's I've seen in Elektor haven't gone into much detail.
I didn't get the issue where they did the in-depth write-up and as
such what I know about it has been learnt the hard way.

DM> I've had a look through some of the CC-Ink photocopies, and
DM> there is a 2 line network called S-ARTnet, that includes power
DM> and bidirectional data in a single pair.

I didn't realise that S-ARTnet had power and data up a single pair.
I have seen it mentioned a few times in CC-Ink but never really
studied it.

BV> Sure, but why add all the protocol overhead of a token passing
BV> network if you don't need it.
DM> True, and the chips look horribly specialised and single
DM> sourced. Read expensive and difficult to obtain.

I would imagine so. Another problem with the i2C chips is that each
chip has it's own address pre-defined by Philips and as such you
can't really move it's addressing around much. THis would make it a
tad difficult for what you guys are thinking of. Much easier to just
steal the concept and roll your own from there.

DM> Anyway, Bob's back at the drawing board - his hummer didn't work
DM> too shit hot :-))

I'll bet it was the smell that gave it away then :)

Cheers, Brenton
@EOT:

---