Video (watch in 720p):
Complete instructions / project log
I'm thinking a more open area would have been better for a first flight.
Sun setting, Instructables contest deadline in a few hours
$15 frame, whatever, I have a spare frame if I have to put it together, but I'll probably invest in an aluminum frame later
As long as you had a plan ...
Did you break many props? Hard to tell from the video. Would have been really exciting hitting the power lines 8)
I'm doing a project that is almost the same as yours, diverging that I'm gonna use XBee as the transmitter.
Now I'm really scared, cause everyone I see using this frame ends in tragedy in the first flight.
Look, was the stabilization system working well? Just accelerating up is not enough to a smooth take off?
I broke 2 props in the final crash, I buy them 8 at a time so I have spares.
As you can see, most of the time the gyros kept the helicopter really REALLY level, it was mainly wind and inexperience that caused the most troubles.
The accidents you saw are mostly due to hitting the tree, if you look closely, even after that really wild takeoff, it recovered almost immediately.
Don't be afraid! My frame cost only $15 (I think it's economical because a single stick of aluminum from Home Depot costs around $10 anyways), I got two sets so I can rebuild within a day or two if I wanted to. Get a better frame if you want.
The standard XBee doesn't seem to have enough range, or at least enough for me to be comfortable using it as the main control method. You might need a higher powered one like the XBee Pro.
FrSKY has two-way telemetry modules that might be better suited if you do not need "digital" data
Hello everyone,
I'm new here, but thought I would just share this site I found since it is very relevant.
From the site: "The AeroQuad is an open-source hardware and software project dedicated to the construction of remote controlled four-rotor helicopters, also known as quadcopters or quadrocopters."
Frank,
I watched the video again and thats true, only the takeoffs were instables.
I also noticed that there was a XBee in your quad. Are you using XBee to control the quad either?
What controller are you using?
I made a controller myself and I'm gonna try in the quad. It's motion sensitive. I posted in this topic: http://arduino.cc/forum/index.php/topic,65685.0.html
And I'm aware of the XBee range and Ill get a XBee pro. (up to 1.6km) Thanks for the tip anyway 
It's always made me shake my head how many newcomers try and jump into a complex project like a Quadrotor as their first arduino project. I would never recommend such a project to a newcomer unless they can get hands-on help with someone local that has built and flown such a device. The safety issues alone should concern most people.
Lefty
retrolefty,
What makes you think this is frank's or my first arduino project?
or was it just a random comment?
I didn't get your comment...
Rivello:
retrolefty,What makes you think this is frank's or my first arduino project?
or was it just a random comment?I didn't get your comment...
It was meant to be a general comment on the topic, as I certainly don't have any knowledge of your specific experience and skill levels with such projects. I have seen at least a half dozen newcomers post here in the last 4-6 months ask about what they need to get started build and fly such a device using an Arduino.
Having been active in the past with R/C building & flying, I do know first hand how difficult, expensive, and yes even dangerous that 'flying' projects can be.
Do try and do everything possible to ensure safety for yourself and others nearby when out 'flying' your project.
Lefty
My radio is a special case. It's a old VEX robotics radio ($25 new), which uses 75 MHz, but I've upgraded it to 2.4 GHz using a Corona DIY conversion kit (another $30).
The XBee onboard my quadcopter is only there to make wireless configuration extremely easy, the GUI can display my sensor values and I can adjust many flight characteristic parameters wirelessly.
If you are looking for a good radio for control, I'd look towards the $60 Turnigy 9x, the main appeal is that its firmware is hackable
Motion sensitive controls will be cool but I don't know how well it'd work in practice
I wouldn't recommend this project to a newcomer either, you are right about the safety issues too. I've done other stuff like spudguns and go karts and plenty of electronics, not much scares me. A good starting point for a newcomer would probably be a upside-down pendulum type robot aka self balancing robot, which involves a lot of the same control theory and sensors.
retrolefty,
Thanks for the advice, I will for sure do all the tests I can before start flying. XD
Frank,
about my controller, I'm searching for a RC simulator to test on PC before flying the real Quad. Just if it looks acceptable I'll try for real.
My controller I'll do myself.. I'll add analog sticks and the "motion sensitive flying" will be just a mode of operation that I plan to be switchable while in flight, but the main control setup will be by the sticks.
Post the news when you have it!
Rivello
I know there are RC simulators out there, I'm not sure which ones have quadcopters. The behaviour of a quadcopter is kind of hard to model mathematically but not impossible (it depends on how accurate you want the model, and the flight controller is very variable).
Although I don't think you really need to "train" to fly a quadcopter, you can do hand-assisted takeoffs first, with one person who keeps a loose grip at an arm distance. This is sort of dangerous but if it flies stable, you should be fine.
The main problem I forsee with motion sensitive controls is how steady your hand is. You can low-pass filter (software smoothing) shaky hands but that introduces some lag into your controls. If your flight controller can keep your quadcopter reasonably stable then this should not be a problem, unless you make the latency over ~100 ms. (estimated, go play one of those "reaction time" button games, I find I average a result of about 300 ms, and I'm not that athletic)
but the main control setup will be by the sticks.
That is probably a better plan then expecting the hardware/software to function properly for the initial flights. However you may or may not realize how difficult it is to learn to fly manually, especially helicopters. Have you mastered any other R/C flying skills before? If not this is a almost impossible task for a new R/C pilot without local assistance. Joining a local R/C flying clubs or locating a local flying mentor and starting off with simpler fixed wing models will save you lots of money and time in the long run. Learning to fly R/C models is like learning to juggle, it looks a lot easier then it really is. And helicopter/quads are amoung the most difficult to learn to control manually. An model airplane want's to fly, but pilot error can crash it, a helicopter wants to crash but a skilled experianced pilot can keep it flying. 
Lefty
frank26080115:
Although I don't think you really need to "train" to fly a quadcopter, you can do hand-assisted takeoffs first, with one person who keeps a loose grip at an arm distance. This is sort of dangerous but if it flies stable, you should be fine.
Hand launching a prototype quadcopter with an inexperienced pilot is probably one of the most foolish things a person can do. The only time hand launching should even be considered is with a known stable quadcopter that already has hours of real flight time, flown by an experience pilot that has hours of flight time on THAT quadcopter.
I've seen first hand the kind of damage even a small plastic propeller can cause to human flesh. It ain't pretty. I wouldn't want to be within 20 yards of any sort of prototype quadcopter flying, regardless how experience the pilot or builder. People would be well off thinking of quadcopters as modern day Phantasm-like flying flesh shredders. Show them the respect they deserve. If you ever get bit by one, you WILL regret it.
what about making some sort of "training gear" for a new quadra like you use for a newbie RC heli pilot.
it will add mass far from the CG therefore stabilizing it a bit, and it will help prevent tipping when it contacts the ground.
once you have it flying stable with the gear, it should be fairly easy to retune it to fly without.
I have not had any experience with a quadra but am fairly accomplished RC pilot (plane and heli)