HELP NEEDED!! Calling all PROGRAMMERS, ENGINEERS, and any HOBBYIST Who Scratch Builds

Nerd Ninja

Reinventing everything!
Hello Flite Test Forum!
And mods. Really sorry if this is the wrong spot! Looked like the best area in my opinion.

Um, I'm new here so first off hello! My name is Trevor and I've been binge watching FliteTest videos for a few weeks now since I found the channel. I've always been a fan of anything that flies and especially RC since a young age, and just recently I've started a super ambitious project with two other friends for a competition called SkillsUSA, specifically this competition: http://www.skillsusa-wi.org/wordpress/?wpfb_dl=316

My goal here is to build a Quadcopter, now I know, what's so difficult about that right? Well this isn't your standard quad. The goal is to combine three different aspects of other existing quads into one to achieve a very unique machine. Specifically those of the Stingray 500 by Curtis Youngblood, the OFM Swift TR260 by AliShanMao and a concept of my own that I personally have not seen applied to quads before. Wish I could give out all the details but I can't at the moment. If you're interested in seeing this come to life then continue reading!

Now to the point of this thread. I need help! This competition is in March 2016 and my team and I have that time to do what took Curtis two years and add even more! I'm talking about the programing side of things. There is no flight controller that I've seen that will be able to handle this crafts abilities. Therefore I have decided to use a Pi or Arduino and hope to have a custom program to control this thing and unlock its full potential. My friends and I have almost no experience in that regard and it would take us many years to get it done alone. I'm asking for anyone's help, especially if you have exceptional knowledge in programming and flight physics.

Please PM me if you're interested. I will be checking until the final product is complete, contest or no contest. I want this thing to become more than an idea.

Thanks to all who read to the end!
~The Nerd Ninja, Trevor
 

Nerd Ninja

Reinventing everything!
are you making a whole new chassis or using the stingray as a base?

I'm designing a entire chassis. I have 3d modeling experience and access to a 3D printer for prototyping.
But that's not the hard part. I can do that all by myself, no sweat. The programming is where I need help.
 

Snarls

Gravity Tester
Mentor
What part of the Stingray are you going after? Is it just the 3D capabilities? If so you might be better off just using reversible ESCs. It won't be as efficient in 3D, but it will be way less mechanically intensive and a lot easier to implement.
 

Balu

Lurker
Staff member
Admin
Moderator
I'm sorry, but I think your best option is to start with something ready and modify it. It took years and a lot of peop people to have the different open source firmwares we have now. Starting absolutely from scratch will not be done in three months unless you have a programmer that has all night to code.

It'll help if you could say what's the idea behind your design. I've seen a lot of different things being done just with mixers on a simple KK 2.1 board. Not that I understood what and how they did that, but they did somehow :)
 

Nerd Ninja

Reinventing everything!
I'm sorry, but I think your best option is to start with something ready and modify it. It took years and a lot of peop people to have the different open source firmwares we have now. Starting absolutely from scratch will not be done in three months unless you have a programmer that has all night to code.

It'll help if you could say what's the idea behind your design. I've seen a lot of different things being done just with mixers on a simple KK 2.1 board. Not that I understood what and how they did that, but they did somehow :)

*Note: Had this as one post, split it up to be able to tag you. I don't know how :? so some things might not line up. Just read the post below this and it might help

As far as the idea goes, it's one that I've never seen but could have some amazing flying potential. Imagine a Quad that could tilt each prop individually 90 degrees both forward and back and also Left and Right 90 degrees. Allowing for a full half sphere of movement. The acrobatics would be insane! It would also be extremely stable, smooth, and quick. The Variable pitch comes in to 1. Allow the use of one motor, and the option for a upscaled simplified version, and quicker response when going from one extreme (-90 Degrees) to the other (+90 Degrees) and allows for the craft to not have to rotate the props 360 degrees to achieve any sort of reverse thrust.

I realize that this has a bunch of engineering obstacles and mechanical ones as well but to be honest I believe I've already got it mostly solves. The issue is programing. And that's where I'm asking for help!
 

Nerd Ninja

Reinventing everything!
What part of the Stingray are you going after? Is it just the 3D capabilities? If so you might be better off just using reversible ESCs. It won't be as efficient in 3D, but it will be way less mechanically intensive and a lot easier to implement.

*Note Read note from post above

How the heck does the multiquote here work?? Whatever, The reversible ESC's would work for ease, but I'm more after the single motor aspect. For the contest one of the criteria is "How can your innovation be applied in the real world?" And a one motor system could be the answer to full scale personal quadcopter aircraft. More mechanically intensive? By golly yes, but I think I've found a way over that hill already.

And modifying something else is definitely an option and what I'll probably do, but neither I or my friends have the programing knowledge and physics knowledge to modify stuff, and how to even code things like, if I do this on the stick tell the aircraft to do this!
 

bitogre

Member
What I am seeing here sounds interesting but may be a little overly ambitious for a 3 month project, especially if you do not have any coding experience and are trying to do this all from scratch. I graduated from CalPoly with a BS in Computer Engineering (which includes programming) took all 3 quarters of control system theory while there and would not try to take on this project in 3 months. That may be because I may be overly cautious. If it were me, I would use the Seriously Pro Racing F3 controller as it can control up to 12 servos/ESCs (if you use SBus or CPPM in) and has all the sensors you should need. Also, it can run CleanFlight and many other Open Source flight control firmwares. Use the existing Open Source flight control firmware so you are not writing everything from scratch (especially the PID control loop). You can get CleanFlight source code from GitHub.

My next concern is how much do you understand PID control loops and control system theory? The big problem I see here is you are trying to prefect the Tricopter tail control but multiplying that difficulty by 4 as you are now having all 4 arms need to do that level of control. Most flight controllers have problems with a simple tricopter for good reason (CleanFlight is still trying to get theirs so that it consistent and reliable). Magnifying that complexity in 3 months is awfully ambitious (though theoretically possible) but if you are able to pull it off, that would be extremely awesome.

Good luck. Feel free to contact me if you need help.
 

Craftydan

Hostage Taker of Quads
Staff member
Moderator
Mentor
The big problem I see here is you are trying to prefect the Tricopter tail control but multiplying that difficulty by 4 as you are now having all 4 arms need to do that level of control.

Read again. he's not replacing each boom with a single axis tilt, but instead a dual axis -- Assuming there's some sort of throttle and pitch control, a total of 13 degrees of freedom in that control system (2 axis per boom + 1 prop pitch per boom + 1 common motor throttle). There's a fun concept called "the curse of dimensionality" that comes into play here, although at 13, this example is small . . . I wouldn't call the math trivial.

I agree, 3 mo is foolhardy at best with an ace programmer, and from the sounds of it, Ace programmer doesn't even have hardware yet to play with. In my experience, having software lead hardware always makes for an "interesting" integration period, and devastating when the overly-clever-but-never-been-done-before hardware slips schedule.
 

Balu

Lurker
Staff member
Admin
Moderator
I have actually seen something like this. Ralph Kayser who built an amazing model of the Avatar Samson Duocopter was also working on a frame for the Avatar Dragon. The basic multirotor frame was already flying great and used tiltable motors. I'm not sure if it was two tiltable axis or four though.

If I recall correctly he just used a modified KK2.1.5 board, but I have only talked with him for a few seconds, so I don't know the details.
 

bitogre

Member
Read again. he's not replacing each boom with a single axis tilt, but instead a dual axis -- Assuming there's some sort of throttle and pitch control, a total of 13 degrees of freedom in that control system (2 axis per boom + 1 prop pitch per boom + 1 common motor throttle). There's a fun concept called "the curse of dimensionality" that comes into play here, although at 13, this example is small . . . I wouldn't call the math trivial.

I agree, 3 mo is foolhardy at best with an ace programmer, and from the sounds of it, Ace programmer doesn't even have hardware yet to play with. In my experience, having software lead hardware always makes for an "interesting" integration period, and devastating when the overly-clever-but-never-been-done-before hardware slips schedule.

You are right in that I missed the two axis of independent direction control on each arm. And with 13 degrees of freedom, it sounds like they are going to need something a lot more powerful than a Seriously Pro Racing F3 flight controller, both in numbers of outputs and processing power.

I also agree that it is very rare where you can have software lead hardware. Usually, those are cases where the software is effectively trivial and the hardware problems trying to be solved do not impact the software (for example, shrinking an existing solution to meet new size, power, and weight constraints). This is definitely not one of those cases.

Is it possible to do what they want to do? I think so but the odds are stacked against them.
 

jamboree1

Active member
Whoah Nelly! A very ambitious plan. Just an idea I had for ya, look into some of the 3D heli controls and programs, might allow you a fresh perspective and shortcuts.