Yes, Arduino is very cool. I'm working on a project that I'm calling "Spaceduino" to use an Arduino as the flight computer for a high-altitude balloon.
The idea is to have the Arduino interfaced to a Telit GPS and GSM unit plus a 433Mhz radio for telemetry downlink. I'll launch the balloon to take photos/videos of near space and use the telemetry to track the balloon for recovery.
So, balloon will be attached to the payload box via rope which has two things attached to it: a parachute and a radar reflector. Reflector is there so aircraft can see the contraption, the parachute will be open all the time, but doing nothing as the balloon ascends.
Once the balloon bursts (somewhere over 30km) the parachute will slow the fall.
Throughout the payload will broadcast its location derived from GPS over a 433Mhz link (probably RTTY) which I can track. Once at a low altitude (not sure how low yet) and within GSM range the payload will also SMS its location to another cell phone.
This should give me enough data to find the balloon. Plus there is software to do balloon trajectory prediction based on weather forecasts. That'll give me a general idea of route as well.
Also, on a clear day the entire thing will be visible from the ground using high-powered binoculars or a small telescope. If I have time I'd like to get automatic ground tracking in place by attaching a Yagi antenna to a powered telescope and using the downlinked balloon location to steer the telescope to train the antenna on it and to get video of the balloon's ascent. That last part may be beyond my budget unless someone's got a steerable telescope mount that has a computer interface (RS-232/423 or USB) that they'd like to give me :-)
I plan to use a small video camera that will be activated at altitude to record HD video of near space (and the descent). Currently thinking of a Slide HD which can record 4 hours of video and is small and light, but this part isn't nailed yet.
This is actually a really complex project: navigation, telemetry, all the balloon related aspects, interface with ATC, temperature extremes, weather, ...
This sounds really ambituous. I seriously hope your plan will succeed! Just be sure to document the whole adventure in a.series of blogposts. Project like that is bound to top HN's front age in an instant.
I want to downlink some pretty simple information: lat, long, altitude, temperature, so I think RTTY should be good enough. There are some nice modules by Radiometrix which work on 433Mhz (unlicensed band in the UK) (e.g. http://www.radiometrix.com/content/ntx2). The CU Spaceflight folks have been using these modules at 50 baud. I think I'm going to stand on their shoulders for this part.
And because it's so popular, there are a bunch of libraries and tutorials for it. It also abstracts or designs away a lot of the tedium in programing micro-controllers.
At first, I loved it because it's pretty cheap, and I could just plug it into my USB port and start coding, without having to worry about making a programmer or buying a USB/Serial interface.
Now, I love it because it's pretty cheap, and it has full GCC toolchain support, so I don't have to use some proprietary IDE.
This Daft-Punk-inspired helmet I made is controlled entirely by an Arduino Duemilanove (http://vimeo.com/10684953).
It's fairly complex as far as Arduino Projects go (it involves timer interrupts, a Serial Peripheral Interface with shift registers, and a homemade PCB), but hey, I managed to do it, and I didn't have a lot of prior experience with either C or hobby electronics.
I ordered a sanguino kit and learned to solder at the same time :)
If you're interested in learning the whole parts procurement process, you can also buy a PCB of pretty much any arduino-based board with no components, then pick out the components you want yourself and solder away.
For auxiliary hardware for the Arduino, use a breadboard, and get components from a salvaged electronic device - an old pre-2000s TV is best, but anything will do: a radio, stereo system. They should ideally be pretty old so that they contain enough discrete components. Newer stuff is all SoC - System on a Chip - just a big proprietary non-reusable chip and very few other components.
You can desolder and reuse pretty much any component: connectors, passive elements (resistors, capacitors), switches, active elements (diodes, transistors), LEDs, speakers, relays. You can even reuse bigger functional blocks like power supplies or power amplifiers. I used to do this all the time when I was a teenager.
Nothing like having an idea at midnight and scavenging the components to try it out at once, and not having to wait next day to buy them. Plus, it reduces e-waste, which is horrible to recycle, despite what we are told.
And stay safe, learn you electric safety rules!
PS: nice first project idea - light alarm that wakes you up in the morning, like the fancy, pretty expensive Phillips Wake-up Light Alarm devices. Later, make the light able to fade-in using maybe a thyristor. A nice enclosure is hard to make though - that's where Phillips has the advantage.
The BeagleBoard is stuck in a no-man's-land between low-end 8-bit controllers like Arduino and the many other AVR/PIC platforms, and a full-fledged embedded PC. If I want to monkey with Linux and take advantage of a full 1 GHz ARM core, I probably want to just use a commodity PC motherboard. Chances are, it won't cost much more, and it comes with the huge advantage that I already know how to develop for it.
It is also a total PITA to get 'booted' since it comes blank you need to mkfs a sd card. if you have linux on a computer already, this is pretty easy. but what if you dont have a linux machine ready to go? if you're me you spend 2 days trying to maneuver something, giving up and borrowing a laptop. I've had an easier time getting freebsd onto a Mac SE/30!
In the end what i really wanted to do is use the S-Video port on the BB and after much googling, determined that there was no driver, no easy way to do it and i was basically SOL especially without onboard ethernet/wifi.
The BB is a great embedded SBC, it completely slaughters the (hopefully now completely dead) PC104 but right now I'm tinkering with a Chumby: Its not nearly as powerful but for $110 you get a full retail package, LCD, composite video out, speaker, wifi, & power adapter. In 2 seconds i was ssh'ed in and porting SDL-based graphical apps with ease.
Do you think the lack of drivers and external support for things like the SVideo is a TI deficiency, or a linux community problem in general?
I'm trying to decide between i.MX51 and OMAP3/AM37 for an upcoming project and it seems like the level of support from TI or Freescale is going to be the only thing making a difference in the decision...the chips are just too similar in functionality.
im like the last person you should ask about embedded linux. ive done a smattering of projects but nothing detailed enough to advise :)
that said, i think the svideo thing is a 50/50 linuxwiki/TI problem. on one hand, the linux community loves to say "oh you know, just recompile the kernel" and on the other hand, TI should have made a completely ready-to-go bootimage one could download that demos -everything- on the board.
True, it would be nice if the SDK was complete on first silicon. But that never seems to be the case. And if there is code to demo a feature, it's probably a hack.
At $149, it is more expensive than many Mini-ITX boards with a 1.66MHz dual core Atom, which come with quite a bit more on board peripherals (SATA, 7.1 audio and the like). Once you add RAM the price difference narrows (or even reverses if you put more than 1GB, which you couldn't do withe the BeagleBoard), but you get quite a bit more power and flexibility at a comparable price.
The main advantage of a BeagleBoard over a Mini-ITX system is significantly lower power consumption. The entire BB can run on a single USB port (5V 500ma). Otherwise, cell phones would be running Intel and AMD processors.
The only thing I dislike about the beagle board is that the generation 1 had a faulty USB interface. Also the Serial Connection is a 9 pin. I'd rather they just let me use a standard ft cable.
=/ Arm units are definitely cool. Look at the maple leaf, and other Arm 3 based units. Very very cheap stuff. Very very powerful.
Arm is definitely seizing control of the embedded device market, and linux with it.
The USB interface still has issues. They've supposedly fixed them in rev. C4, but I have a C3. My main concern apart from the USB stuff is being able to buy enough units in quantity, with a reasonable turnaround time, if the BB (or any other hobby board) is used in a design.
Because it's cheap, and because it's awesomely simple to use. It's a hobbyist's dream come true.
If you're interested in some of its uses, just look back through Make magazine's site. Loads and loads and loads of projects appear there.