[COLUG] OT: OneWire hardware hacking

[Mike Schoenborn]

At 12:29 PM 1/2/2008 -0500, you (jazzcat?) wrote:

>My timing was off before:
>
>The Pulse "low" time can range from 0.1-10 Microseconds
>The pulse "high" time seems to be able to range from 0.1 microseconds to 
>infinity
>The pulse transition time must be less than 100 nanoseconds

Those are *much* more reasonable timing constraints.   I wouldn't
mess about with 555's and TTL, but would instead throw $2 worth
of "smart glue" at the problem, in the form of a little 8 pin AVR CPU.

Something like an ATTINY2313 ($1.40 at DigiKey) would do: two
PWM channels, 1k FLASH, 64k RAM, internal oscillator (no need
for a crystal), a smidgen of EEPROM, 6 programmable I/O lines
(on an 8-pin DIP!), etc, etc.  Great little devices, lots of "community"
around it (avrfreaks.net), Linux build tools, etc, etc, etc.

http://www.atmel.com/dyn/products/product_card.asp?part_id=3229

(And if you can live without RAM (which you shouldn't need for
just banging out 16 pulses), check out the ATtiny15L.  I've got a
spare I could give you, or swap for an 1821 if you have an extra).

On the other hand, with the relaxed timing, you might just generate
those pulses within the Chumby itself.  I haven't studied its hardware
details, but the little daughter card inside might have a spare bit you
could twiddle, or a serial control line you could toggle.

Mike