Zigduino Mote

[Pierce Nichols]

I talked to a few potential users at MakerFaire who would like to see
a RTC on the Zigduino Mote. It will add a couple of bucks to the cost
of the full-up version and nothing to the cost of the stripped
version. Is this something you need? Does it help with applications
you are considering? Other thoughts?

-p

--
Pierce Nichols
Principal Engineer
Logos Electromechanical, LLC

[Frank26080115]

Hi Pierce,

When they say they want to see a RTC, do they mean a simple 32.768 KHz
crystal? Or do they actually want a dedicated RTC chip?

The ATmega128RFA1 definitely have a async timer, which should
eliminate the need for a dedicated RTC chip 90% of the time. If you
want to add a watch crystal to the async timer, I don't think it's
that hard to just solder the crystal plus two capacitors even if you
don't leave a dedicated footprint for them. This is assuming I already
have access to the async timer input pins.

Of course, if you did not break out the async timer input pins in the
first place, then add the crystal directly, or else you've just wasted
those pins.

-Frank

[Martin Haeberli]

Pierce,

In several applications, the need exists for low-rate, relatively time
accurate (let's say 1 second or better) data logging, then with
asynchronous backhaul of the measurements to one or more collection
points. I'd either want low clock drift even w/large temperature
changes, or, less preferred, a way of keeping the clocks in sync. So
a $2 RTC (what specifications do you have in mind?) would be worth it
to not hassle the alternatives.

Best,

Martin
----

[Pierce Nichols]

The RTC I am looking at is the S-35390A-I8T1G, which is barely bigger
than the crystal and talks I2C. In quantity, it's ~$0.60 in quantity
plus ~$0.40 for the crystal. It has two interrupts as well, and it
draws only a quarter of a microamp when sleeping. Thoughts?

-p

[Martin Haeberli]

Looks cool!

On Thu, Sep 22, 2011 at 9:11 PM, Pierce Nichols

[Pierce Nichols]

All the pins are broken out, regardless. The problem I see with
built-in asynchronous timer is that the longest possible interrupt
interval appears to be only 8 seconds (32.768 kHz/1024, counting
0-255)... and it's a little while before you can go back to sleep.
This strikes me as problematic for the kind of sensor network Martin
and at least one of the people I was talking to at MF are interested
in, as it is likely to substantially reduce battery life.

-p