Just a bit from my dabblings with the MaxStream XBees:

Gerhard Fiedler wrote:
> - I'm not sure whether one-way transmission will be good enough. I want to
> get close to 0 false positives at the receiver, and false negatives should
> ideally be less than 1 in 1M. Is this realistic with one-way transmission?
> Or would I need transceivers and a protocol? (In this I'm not considering
> concurrent presses of switches connected to the two transmitters.)

The XBee protocol will give you a return status on point-to-point 
transmissions.

> - In case running my own FCC certification doesn't make sense, the only
> small pre-certified module I've found so far is a MaxStream ZigBee module
> <http://www.maxstream.net/products/xbee/xbee-oem-rf-module-zigbee.php>. But
> it's bigger than I'd like it, it uses more current than I'd like it to use,
> and it's more expensive than simpler solutions. Are there any smaller
> and/or less power-hungry and/or cheaper pre-FCC certified RF modules out
> there?

I wish XBees cost less and used less power... but I wasn't able to find 
anything I liked better a year ago, and I really like these now that 
I've been using them.

> - The other question about this MaxStream ZigBee module I haven't yet
> answered is related to the protocol. How long does a typical ZigBee
> information exchange take, between power-downs of one device? My doubt is
> about resyncing with the mesh. Can a station wake up and start transmitting
> right away? What's the typical time for a complete protocol exchange of
> very little data (1 byte is enough in my case -- if the protocol takes care
> of error checking etc)?

I don't have hard data on this, but there's stuff in the API 
specifically for wakeup, poll, transmit, and sleep.  They're very into 
this kind of application, and very conscious that it requires a low duty 
cycle to get good battery life.

In fact, there's a new thing in the API that I haven't looked into much 
yet, that appears to me to mean you can just configure an I/O on each of 
two XBees to be "in sync," and declare the update period you want, and 
they take care of the rest - logic 1 on the near pin flows to logic 1 on 
the remote pin, 0 to 0.  Seems ideal for what you're doing.

> - Last not least is the antenna question. I'd like to come by without any
> antenna external to the (very small) enclosure. The MaxStream module comes
> in a version with what they call a "chip antenna". Its performance seems to
> be just fine for my needs. Is this more than just a skillfully placed
> track?

My RF understanding is weak (which is why the all-in-one solution 
appeals to me).  But there is a tiny extra board pasted on top of the 
module on the "chip antenna" versions.  And yes, I think the chip 
antenna would be quite adequate for a few meters.

Also on size - keep in mind that there's a Freescale micro inside the 
XBee that you can run your own code on, and several of the I/Os are 
brought out for general use.  Could save a bit of space by not using an 
external controller.  I haven't tried that yet, either.
-- 
Timothy J. Weber
http://timothyweber.org
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist