I ended up building the prototype soil moisture sensor boards I mentioned in a previous post, and... it works! Somewhat.

This is what it ended up looking like:

The sensor section of the board I coated with black Plasti-Dip to protect it from moisture, and the results turned out rather well. In later versions I'll probably be potting all of the circuity in either Plasti-Dip or epoxy.

It turns out my method for trying to get two zones to work properly on the limited pin count of the ATTiny13a was not effective—instead of getting independent values for each depth, it acted like a single large sensor. In light of this observation, I've decided to jettison the cleverness and just focus on a single-zone capacitive moisture sensor design—at least for now. I also removed the ground plane as well, as it doesn't seem to be helping the readings. Here it is the new single-zone version:

I just wanted to share a few pictures of project I'm working on. This is the next logical step from the appliance module I mentioned a while back: a smart light switch. The design is modular, which allows for all sorts of configurations, but the first prototype will be be a simple on/off switch with the ability to be remotely controlled.

There are four boards: a main board, a relay board, a power board(which also handles remote communications), and the touch board. Eventually I'll be making a dimming version of the relay board, and I may possibly implement a power-over-ethernet version of the power board.

The user interface is a Decora-sized 1-dimensional capacitive touch plate with a frosted glass surface, with slight indentations on the upper and lower portions. Pressing once on the top will close the circuit, pressing once on the bottom will open the circuit. If the switch is equipped with a dimmer module instead of a simple relay module, touching the top and dragging your finger down will dim, and the reverse will brighten. A piezo-electric element on the main board will provide some tactile feedback as you touch the glass, making a satisfying click when you press it. Right-angle Status LEDs shine light into the top and bottom of the glass, which is diffused by the frosted surface.

Cool, huh?

I've been sitting on these designs for a while now, so I figure it is about time I pull the trigger and order a few boards to populate.

In my previous post, I elaborated on how I reverse engineered the low level protocol the GE Color Effects G-35 strings used. It seems to have been quite popular, and several people have taken that information to hack their own G-35 strings to do cool and amazing things. But what about me? What am I doing with my strings?

Well, it's a bit late in the year to start putting together some sort of advanced choreographed light show, so I'm staying somewhat simple for now. No music, just a few different algorithmic patterns. And here's the kicker: I want to be able to control them with the web browser my phone too. Sounds like a job for… the ybox2.

Behold, the networked Christmas light controller in an Altoids tin:

In short: it works great. See for yourself:

I went to Costco the other day with my wife to pick up some Christmas lights. While there, I found the most awesome christmas lights ever: the GE Color Effects G-35.

This string contains fifty "bulbs". Each bulb contains a red, green, and blue LED—allowing for any color to be produced. The biggest drawback is how you control them: there are 14 pre-set "programs" that you can select, and they are all pretty lame. However, with a little reverse engineering and an Arduino (or any other microcontroller) they can be made into the coolest Christmas lights on the block. If you buy several strings, you could easily make a large 2D LED array. The possibilities are endless. At $59.99USD for a string (A tad over $1 per node), these things are a steal for what you can do with them.

I wanted to share my efforts so that other people can give this a try. At the end of the post I have a link to the code I'm now using to drive my Christmas lights. Feel free to use it for your own hacking endeavors.

So far I've reverse engineered both the radio protocol and the protocol used on the LED data bus. Of the two, the LED data bus protocol is by far the most interesting. So that's where I'll start…

UPDATE: See follow-up post here.

This is an experimental capacitive soil moisture sensor I am working on. Just thought I would share.

These images are actually a little old, but I don't feel like making new ones. The new revision uses a slightly different mechanism for measuring capacitance that should be more accurate.

I'm using capacitance to measure the moisture instead of resistance because it allows me to void having any metal directly touch the soil. Resistive moisture sensors are not practical for semi-permanent use due to electrode corrosion. This moisture sensor should work indefinitely with no corrosion.

Assuming it will work at all, that is. We shall see.

I plan to have several of these daisy-chained around my yard. At the end of the chain will be a solar-powered wireless micro-controller unit which will allow me to read any of the sensors from the wireless home automation network I'm building. The goal is to use the moisture information to help the irrigation system know when it should water less than it normally would.

My hobby is complicated.

Nothing quite celebrates one's geekdom like showing off pictures of an OLPC XO-1 using an RZUSBSTICK 6LoWPAN network adapter. I couldn't resist sharing.

Don't give me that look. I needed to test my CDC-ECM additions on a Linux box, and it was the closest one handy. It's all in the name of science. SCIENCE.

You can see it actively (and successfully) pinging ff02::1 on the XO-1's screen.

The Contiki distribution comes with some very useful firmware that turns the RZUSBSTICK into an ethernet interface. This firmware is informally called "Jackdaw".

Unfortunately, out of the box it only supports RNDIS and CDC-EEM (Ethernet Emulation Model) USB protocols. There is no RNDIS driver on the Mac (and likely never will be). There is no usable CDC-EEM driver either.

There is a standard protocol for USB ethernet devices, and that is CDC-ECM (Ethernet Control Model). In order for the Jackdaw firmware to ever 'just work' on a Mac, someone is going to need to add support for CDC-ECM to the firmware.

I've been intending to do this for a few months now, but only recently managed to scrape up the time to put all of the pieces together. After a good 8-10 hours of work this weekend, I am now pinging my RAVEN board directly from my Mac using a RZUSBSTICK running my updated Jackdaw firmware.

I'll be sending my changes upstream so that they can hopefully be included in the next Contiki release. If you are curious or just impatient, you can grab the changes from my private subversion repository here.

Update: I've moved the changes to their own branch here. I'm also working on merging the changes into contiki-2.x, (which you can find here), but that is still a work in progress.

Additionally, you can find builds here that are ready to upload to a RZUSBSTICK.

I just wanted to share some pictures of another aspect of the same home automation project.

What I'm doing here is connecting a TI CC2520 802.15.4 transceiver to a ybox2, using a break-out board that I designed a few months ago. I'm in the process of developing the CC2520 driver for the Propeller now, and hope to have it working in a week or two. After that, I'll be developing a 6LoWPAN stack for the propeller chip. Fun times.

So what are the possibilities?

• 802.15.4 tunnel, for connecting two Zigbee networks in different locations.
• 6LoWPAN to ethernet bridge.
• 6LoWPAN border router.
• TV-based 802.15.4 network monitor

I'll be exploring all of these options in the weeks to come.