Yes, TOS 1.x, which underlies Moteworks. Although I just noticed
that the last release of T1 has a hardware impl for I2C in mica128.
Here is what I used, including my small patch to I2CM.nc:
http://www.etantdonnes.com/Motes/T1_I2C.zip
But I imagine that it will be fairly pointless to try to use it
as all the bit access methodology was changed in T2, so there'd
be a bunch of porting to do.
I also put the spec that I used to debug my problem here,
where the section of interest is "3.7 Clock synchronization":
http://www.etantdonnes.com/Motes/NXPI2Cspec_UM10204_3.pdf
My symptom was that I missed the first bit of the second byte
of data from the slow device and thus lost bit sync and failed
to complete that byte (if memory serves correctly...). I found
this by looking at the data line with a scope and counting the
bits...
If you really think your problem is the clock stretching issue
then it might be better to see how the ATmega hardware thinks
it should be dealing with it.
MS
Sebastian Dölker wrote:
> Unfortunately I have never worked with moteworks
> and T1 (T1 = tinyos 1.X ?)
>
> I work with tinyos-2.1.1 and the standard i2c component of atmle128L (iris)
> which uses the under laying twi (two wire interface)
> of the atmel128L chip but it seems that with this
> special co2-sensor (with long reaction times and the “clock
> synchronization mechanism”) this doesn’t work well.
>
> I guess I have to do bitbanging..
>
> Is it possible to use the sources that you mentioned as a starting point?
>
> Where can I find them?
>
>
>
_______________________________________________
Tinyos-help mailing list
Tinyos-help@millennium.berkeley.edu
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
