Senior Project Quadcopter Build Log

[Snarls]

Hi FliteTest community! I am a senior in high school with the opportunity to partake in what is known as a "Senior Project". Basically, instead of taking a class, I am given the time to pursue any project I want. Back in October I was suddenly inspired to build a quadcopter. I had little to no flight experience, not to mention any clue of how to build a multirotor. Like many of you, I stumbled upon FliteTest and was immediately hooked. Over the winter (very cold in the northeast), I spent hours researching and planning for my project. I even taught myself to fly in below freezing conditions with two feet of snow. I spent the time making drawings and renders in Sketchup, as well as plan out the stages of my project.

My school divides the year into four terms, and I am very excited to say that now is the term that my project officially begins. I have from April 4th to approximately May 25th to complete my project goals and I will be using this thread to catalog my build. After the project is over I hope to turn this thread into an interesting article.

Here are the basic goals for the project:
-Construct a basic quadcopter and learn to fly
-Contruct and calibrate a brushless camera gimbal
-Film aerial video of various locations with the gimbal
-Develop autonomous (waypoint) flight
-Test out other quadcopter abilities (flips, lifts, etc)

And with that begins my build log...

I started where most people begin, which is with the frame. I went to Lowes and purchases some 1/2" square dowels, a 3/8" square dowel, some hardware, and spray paint. I used templates and measurements I made in my plans to cut out each piece of the frame.

The landing gear is based on the Hobbyking X666 frame. I designed the frame to have the smallest main plate area as possible, but still be able to fold together. I settled on a 4.5" square plate in the middle and 12" arms. I managed to break my coping saw in the process of cutting the landing gear, but eventually ended up with all the parts. I put everything together for a rough look and then proceeded to paint (The landing gears are clamped because they get attached with zip-ties along with the motors).

For the paint job, I decided to paint two arms black and two white for orientation help. I had enough wood for two extra arms which I paint red to signify if an arm was replaced. Two coats of paint with sanding in between and it the wood looks like aluminum or plastic if you stand far enough.

This concluded the frame for now as I am still waiting on parts from Hobbyking (its killing me), but you can begin to see its resemblance to this rendering I made:

With time to spare I decided to begin work on the camera gimbal. Again, I have not electronics so I have built all the parts, but could not do any final assembly. The gimbal is made out of wood and is designed to be adjustable along 3 points to allow for balancing. The parts are are based on this model I made:


It has been one week so far into this project and I am very pleased with my progress so far. The only frustration is the lack of electronics. I will update this thread with updates as often as possible and I will try to answer any comments or questions. Thanks and enjoy!

 

[FlyingMonkey]

Check out this light and simple mount...

http://www.rcgroups.com/forums/showthread.php?t=827488

 

[Mustang7302]

Back in high school (I am not -that- old) I spent my spare time tinkering on paintball markers. It was fun to learn basic principles, tinker, and have something tangible out of my efforts. But that wasn't for a school project by any means.

You're off to a fantastic start. Keep it up and don't be shy to ask for help.

 

[Snarls]

Thanks for the responses guys! FlyingMonkey, that mount you posted has definitely given me some ideas, and thanks for the response Mustang. I wish I had more to update you guys on, but unfortunately, my project has come to a standstill for the past two weeks due to problems with my orders at Hobbyking. One of my orders is stuck with the infamous Malaysia Pos shipping (read about it here: http://www.rcgroups.com/forums/showthread.php?t=2050859&page=13), and the other order has not shipped from the US warehouse after almost 7 weeks. I have contacted customer support, but hope to get more luck with the new NY based CS group.

I hope to update you guys as soon as possible!
By the way, my name is Sam.

 

[FlyingMonkey]

NEVER use the Malaysia POS. It's a real POS… I've learned my lesson.

Yours is probably lost with mine. It's probably from the same shipment.

I've put in a Paypal notice, that's about to escalate to a claim if CS doesn't refund, or do a credit soon.

Not as a threat, but just to make sure I get my money back.

 

[Snarls]

Good news, I finally received my shipment from the US warehouse at the beginning of the week and can continue the build process. Even better, I received a slip from the post office requesting I pick up the missing Malaysia Pos shipment which somehow found its way to me.



Here you can see some of the parts that I received. The two biggest items in here are the Multiwii Pro FC and Q Brain ESC. I have never set up a flight controller before, so Multiwii is definitely a learning curve. Despite this, I think I was able to connect the gps and upload the firmware to the board. I plan on running Multiwii 2.3navi b7 which will allow me to use waypoints. If anyone has any good tips for this version of the firmware please share.



I am not sure if I need to change things in the firmware code such as min/max throttle, as well as if I need to calibrate my Q Brain. Any help along these lines would be appreciated as I continue to build this week for a hopeful test flight this weekend!

 

[Snarls]

Hey guys, my missing international package showed up last week, but you wouldn't believe what its been through. For reference here is a picture of what the box looked like:


Despite this shipping catastrophe all the parts inside appear to be completely fine (still have to test some electronics). With all the parts at hand I quickly assembled the frame with electronics on board over this past weekend. Everything fit into place just like I designed. The only issue is that the Q brain barely fits into the center of the frame. Some shifting is necessary to collapse the frame down and the wires could not be a millimeter shorter. On the top of the main plate I installed the Multiwii Pro board as well as the gps, receiver, and telemetry module.

I spent most of my time this past weekend not building, but programming the Multiwii. It took a lot of time researching online to determine what exactly to do with the firmware, but eventually I was able to get the gps, telemetry, and basic flight abilities working. As of now I have only tried a few flights that have each lasted only seconds (safely landing thankfully). I can tell the quad definitely needs some tuning, and I need more confidence getting this thing off the ground. I will try to get some pictures, flight video, and possibly some gimbal use very soon. Thanks guys!

 

[jhitesma]

Amazing what international shipping can do to a box. I had one show up that was apparently rectangular when it left the place I ordered from...but was almost perfectly spherical when I got it! Oddly enough the only times I've had things arrive broken the packages looked fine...every beat up box I've gotten has had no damage to the stuff inside <fingers crossed>

 

[Snarls]

Hey guys, I finally have some pictures of the quadcopter to show you as well as some updates on the gimbal. I have spent most of my flying time trying to tune the PID values, and I think I have finally found the correct values. The quad has only crashed once so far luckily, but I am realizing more that I need more practice flying before I can attach the camera gimbal or do anything. I have only flown in my backyard so I am not use to anything besides general hovering and small movements. I fly in angle mode for auto leveling and have not tried any other modes. Can anyone suggest what order I should try out the Multiwii modes (Altitude hold, RTH, etc).

The gimbal is looking great. I did my best to balance along the roll and pitch axis beforehand. I am using the hobbyking 2-axis gimbal controller running BruGi and it works fairly well. The roll axis runs clean with a cold motor, but the pitch axis requires more power to the motor in order to not be choppy, so the motor gets fairly warm. This is probably because I am using a weaker motor. The biggest problem I have is that the gimbal freaks out when tilted too fast or to too extreme of an angle. Is this normal for BruGi? If not what can I change in the programming?

Anyway, here are the pictures:

 

[jhitesma]

Can't help with your gimbal questions...but as for MW modes.

I'd say get out to a bigger area (unless you've got an acre or two or a backyard) and give horizon a try. It's a rate mode like acro but still self-levels when you let off the sticks so it's easier to learn than full acro.

As for the other modes. Don't forget to do a mag calibration before trying any of the headlock/headfree/GPS modes. You'll want to get baro (alt hold) tuned in first since pos hold and RTH both depend on it. I'm still working on getting mine tuned in but flew with the defaults for some time...it worked but didn't hold altitude very well. Following the guide: http://www.multiwii.com/wiki/index.php?title=Altitude_PID it's getting better but I found going in .2 increments on D was way too little for me and I was actually able to max D out and never got the yo-yo effect...I did keep getting a more stable alt hold though. I didn't finish getting P and I dialed in before I swapped to new motors and I've yet to get my main PID's dialed in for the new motors thanks to my flyaway and high winds the only days I've had free to fly after getting the copter back.

 

[Snarls]

Hi everyone, sorry for not posting in a while. I was very rushed towards the end of the project period and had no time to update this blog. The project is officially over for my school, but in no way is it done for myself. I am taking a short break now but here is what happened the last couple weeks:

After obtaining a reasonable position hold with 1045 props (hovers in 2m radius), I decided to switch to 1147 props. This switch threw off all the tuning I had done resulting in an unstable quadcopter. Even after re-tuning the position hold failed to work. Instead, the copter would hover for a second, suddenly turn orientation and launch 30ft up in the air. I have no idea why this would happen so if anyone has any insight please comment.

Anyway I decided to use 1147 props because I needed more thrust to lift the camera gimbal. I attached the gimbal to the bottom of the quadcopter along with a set of retractable legs. You can see the whole set up below.



Here is where disaster struck. I brought the newly made quadcopter with gimbal to my school in hopes of obtaining my first stabilized aerial footage. I was excited and did not realize the extent of the winds that day. After takeoff the legs retracted and the camera was running smoothly. Suddenly the wind picked up and my quadcopter was pushed back behind me. Either a rising column of air or some programming error then pulled the copter up 50ft higher. Despite my best efforts to land the quad, the whole set up crashed headfirst into the cement at a reasonable speed. The wreckage was disheartening, parts were strung about, in places parts should not be, and all of this happening the day before the project due date.

Here's the light at the end of the tunnel. Upon further inspection at home, the basic quadcopter was up and flying again in 10 minutes. The gimbal had broken along easily repaired glue joints and the legs required a new central plate. All together I had the whole system repaired for my big presentation the next day. I attribute all this to aspects Flite Test uses in their multirotors, such as break away motor mounts and easily replaced wooden parts.

After this ordeal my presentation went to plan and my high school was truly impressed by the work I had done. At his point I have not achieved much autonomous flight or aerial footage, but I am putting the quadcopter aside to address some issues I have. First, the gimbal is obviously too large, maybe not in mass but size, for the quadcopter to carry in a stable manner. I plan to build a smaller, lighter gimbal that will attach to the front of the quacopter, eliminating the need for retractable legs and drastically reducing weight. Second, I need to work on flying with the 1147 props.

All in all after working on this project I have realized that flying around with a simple, light-weight quad is more fun than the stressful task of filming with lots of extra parts. In the future I plan on making a smaller quad to fool around with, similar to Alex and his mini h-quad. As for autonomous flight, I think I will put my multiwii pro on a versa wing instead. Waypoint navigation on multiwii is still under development and I feel much more comfortable using it on a cheap, easily repaired plane that I have much more experience flying.

This is not the end of this project, it's just the end of stage 1, so stay tuned for stage 2!

 

[Balu]

Nice save

Do you have any details on your retractable legs? They look quite nice

 

[Snarls]

The retracts are really simple. All I used was two servoless retracts from hobbyking, the kind used for landing gear on airplanes. The legs themselves are constructed out of the same 1/2" square dowels as the motor arms. The main plate is cut out of a 1/4" thick piece of poplar and has room to mount the camera gimbal in the middle, batteries in back, and retracts on the side. The retracts are mounted on wedge shaped pieces of wood to allow for more clearance for the gimbal and when the legs are retracted they point up at an angle rather than parallel to the motor arms.

 

[Snarls]

Today I made the last flight for this quad setup. i have been designing and constructing a new gimbal system and I wanted to use the old gimbal one last time to test its abilities. I brought the quadcopter to a nearby field to get some footage. It was slightly windy, but I think most of the shake in the video is due to vibration reducing rubber allowing the gimbal to swing with its own weight. The footage is not very interesting in terms of content, but it provides evidence on what to improve on.

Near the end of my flight time the quadcopter suddenly pitched forward. This was either due to a loss of orientation by me or some other environmental factor. The result was a pile of pieces on the ground. The quadcopter is a quick fix but the gimbal is not worth repairing , especially considering I'm building a new one. The camera took a hard hit and has convinced me that I need a GoPro in its place.

Overall this is the end of my initial quadcopter design. I took on the challenge as a school project and with my limited time I think i was able to achieve quite a bit. I will use my mistakes and discoveries in helping me to build a new, more efficient design. For now, here is a picture and some video from the late quadcopter 1.0.