Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Nov 13, 2009, 12:06 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Discussion
Fascinated by quadrocopters - can I design one?

Hi,

I'm new to this forum, after having stumbled across the mikrocopter and then the quadrocopter universe just a few weeks ago. Utterly fascinating!

Here's my situation. I'm an experienced embedded programmer looking for a challenge. I don't have any RC experience other than about 6 months back in the 80s trying to fly (and mostly repair) an early RC gas helicopter, no gyros. I see that the RC universe has advanced greatly in the years since then, and I want to play again. And for me playing would be as much designing and programming such a beast as flying it. On the scale of crazy, where does that idea rate?

As I said, I'm an experienced embedded programmer, familiar with, among others, the ARM and AVR families of devices. I gather that the tasks involved in rolling my own electronics would be, at the most basic level:

- Receive and decode servo signals from the RX.
- Combine those signals with gyro information to calculate motor outputs.
- Generate motor outputs in a form that ESCs can understand.

Am I close? Clueless? Any and all comments welcome. Thanks.

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Sign up now
to remove ads between posts
Old Nov 13, 2009, 03:54 PM
Cranky old fart
Balr14's Avatar
Germantown, WI.
Joined Oct 2007
21,413 Posts
It's a bit more involved than that, but you are on the right track. I would recommend this topic for your review; fascinating reading: http://www.rcgroups.com/forums/showthread.php?t=1093510
Balr14 is offline Find More Posts by Balr14
Reply With Quote
Old Nov 13, 2009, 04:38 PM
Registered User
Hoofddorp, Netherlands
Joined Sep 2004
3,917 Posts
Best check out the different threads on UAVP, UAVX, Mikrokopter (see the megathread in the RC Aerial Photography forum), Quadruino / Aeroquad, etc, decide which platform you prefer (ARM [Armocopter, to some degree Mikrokoper], ATMEL [Mikrokopter, Aeroquad], PIC [UAVP, UAVX], ...) and jump onboard on of the projects. Mikrokopter, UAVP/UAVX and Quadruino/Aeroquad are all more or less open source with access to source code and designs. Aeroquad uses off the shelf parts but is less far in its development (you could probably contribute a lot there), while UAVP/UAVX and Mikrokopter are very far along. Joining a project will get you in the air quicker and at lower cost than trying to reinvent the wheel completely.
Arthur P. is offline Find More Posts by Arthur P.
Reply With Quote
Old Nov 14, 2009, 12:06 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Quote:
Originally Posted by Balr14 View Post
It's a bit more involved than that, but you are on the right track. I would recommend this topic for your review; fascinating reading: http://www.rcgroups.com/forums/showthread.php?t=1093510
Thanks, I'll take a look at that. I'm a little overwhelmed at the size of some of the threads here, but that one doesn't look bad.

I'm wondering if this is a good place to ask questions on software and hardware development. For example, questions like, when reading (or generating) servo signals, is 8-bit resolution (4us resolution) good enough or, if not, what is considered good enough? Would a thread dedicated to such questions and answers be appropriate? Seems like it could be useful to capture a lot of information in one place, but I want to conform to the "look and feel" of this particular forum so I'm asking before posting.

Thanks,
Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 12:14 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Quote:
Originally Posted by Arthur P. View Post
Best check out the different threads on UAVP, UAVX, Mikrokopter (see the megathread in the RC Aerial Photography forum), Quadruino / Aeroquad, etc, decide which platform you prefer (ARM [Armocopter, to some degree Mikrokoper], ATMEL [Mikrokopter, Aeroquad], PIC [UAVP, UAVX], ...) and jump onboard on of the projects. Mikrokopter, UAVP/UAVX and Quadruino/Aeroquad are all more or less open source with access to source code and designs. Aeroquad uses off the shelf parts but is less far in its development (you could probably contribute a lot there), while UAVP/UAVX and Mikrokopter are very far along. Joining a project will get you in the air quicker and at lower cost than trying to reinvent the wheel completely.
There's no question that reinventing the wheel is both longer and more expensive than joining an existing project, but I'm just intrigued enough to at least consider it and try to work out a design on paper. After that, who knows? This would not be with the goal of being able to do it better than the existing projects, but rather with the goal of going through the entire intellectual and engineering exercise, if feasible.

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 12:56 PM
Registered User
KAMAX-T's Avatar
Powell TN
Joined Jan 2008
735 Posts
Quote:
Originally Posted by snarflemike View Post
There's no question that reinventing the wheel is both longer and more expensive than joining an existing project, but I'm just intrigued enough to at least consider it and try to work out a design on paper. After that, who knows? This would not be with the goal of being able to do it better than the existing projects, but rather with the goal of going through the entire intellectual and engineering exercise, if feasible.

Mike

Mike you'll see that this stuff is very addictive , and with your expertise ..... I would love to see some simple "plug and play" tri-copter boards ...auto leveling , altitude hold , all the goodies for a solid FPV fling machine ......JMO

Terry
KAMAX-T is offline Find More Posts by KAMAX-T
Reply With Quote
Old Nov 14, 2009, 01:27 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Quote:
Originally Posted by KAMAX-T View Post
Mike you'll see that this stuff is very addictive , and with your expertise ..... I would love to see some simple "plug and play" tri-copter boards ...auto leveling , altitude hold , all the goodies for a solid FPV fling machine ......JMO

Terry
Hi Terry,

What are the advantages of a tri-copter? Is it just the obvious savings of one motor/prop/ESC, or is there more?

I know that quadcopters (and other even-motored versions, I'd assume) use opposite rotations to counter torque rotation (maybe not the correct technical term). How do tri-copters deal with this effect?

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 01:35 PM
destructive testing expert
dcaspi's Avatar
Kiryat Yam, Israel
Joined Apr 2009
348 Posts
Quote:
Originally Posted by snarflemike View Post
Hi,

I'm new to this forum, after having stumbled across the mikrocopter and then the quadrocopter universe just a few weeks ago. Utterly fascinating!

Here's my situation. I'm an experienced embedded programmer looking for a challenge. I don't have any RC experience other than about 6 months back in the 80s trying to fly (and mostly repair) an early RC gas helicopter, no gyros. I see that the RC universe has advanced greatly in the years since then, and I want to play again. And for me playing would be as much designing and programming such a beast as flying it. On the scale of crazy, where does that idea rate?

As I said, I'm an experienced embedded programmer, familiar with, among others, the ARM and AVR families of devices. I gather that the tasks involved in rolling my own electronics would be, at the most basic level:

- Receive and decode servo signals from the RX.
- Combine those signals with gyro information to calculate motor outputs.
- Generate motor outputs in a form that ESCs can understand.

Am I close? Clueless? Any and all comments welcome. Thanks.

Mike
I was in your position 6 months ago. Today I have a working quad with my own software http://caspiquad.googlecode.com. So it can be done, of course. Developing a quadrocopter is mainly a software project. Go over these forums and learn the background. Look at existing designs. You will need some hardware knowledge unless you want to take an existing board (e.g., AeroQuad) and just write your own software. And of course you don't need to write from scratch, take the existing code base of one or more project as a baseline.

Dror
dcaspi is offline Find More Posts by dcaspi
Reply With Quote
Old Nov 14, 2009, 01:51 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Quote:
Originally Posted by dcaspi View Post
I was in your position 6 months ago. Today I have a working quad with my own software http://caspiquad.googlecode.com. So it can be done, of course. Developing a quadrocopter is mainly a software project. Go over these forums and learn the background. Look at existing designs. You will need some hardware knowledge unless you want to take an existing board (e.g., AeroQuad) and just write your own software. And of course you don't need to write from scratch, take the existing code base of one or more project as a baseline.

Dror
Hi, and thanks for the comments. I do have a decent hardware background, so that shouldn't be an issue for me as far as designing any controller and accessory circuits and boards. Right now I've just got to get my mind around the big picture and then start in on the details. Lots of reading to do.

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 06:33 PM
checkout my Blog
otlski's Avatar
Joined Feb 2008
466 Posts
On tricopters the rear motor and thus the prop steer the craft in yaw. Typically the entire boom is rotated by a servo tipping the plane of the prop. To hold a level hover, the rear motor is rotated to a slight angle to counteract its own drag and the yaw influence it exerts; effectively canceling out the yaw imbalance.

Quickly considered, the efficiency losses due to the the out of plane rear prop would made the tricopters undesirable. However, many of the tricopter proponents would tell you that they have huge choice in prop and motor combinations allowing them the opportunity to make up for any efficiency losses.

I'm sure others will chime in and tell you why they prefer tricopters. I fly quads. I like the control symmetry.

Dan


Quote:
Originally Posted by snarflemike View Post
I know that quadcopters (and other even-motored versions, I'd assume) use opposite rotations to counter torque rotation (maybe not the correct technical term). How do tri-copters deal with this effect?

Mike
otlski is offline Find More Posts by otlski
Reply With Quote
Old Nov 14, 2009, 07:08 PM
Registered User
Joined Nov 2009
1 Posts
@snarflemike:
I could've written this exact post myself. I also just recently stumbled onto the world of quad copters, and have been frantically reading up on all of these amazing machines. I gotta build one!

I also have done my share of embedded programming, mostly on arm (cortexm3 from luminary). The reason that I love arm is that we can get a high quality open source toolchain.

I'd love to start with some of the code that is already out there for a base, but the main purpose of my obsessions are to learn, and to put it together myself. My thinking is that I'll use an arm board that I already have, add the 6dof from sparkfun, and start from there. I'm sure that could keep me busy for months!

I think my goal is mainly a UAV. In fact, I don't even know if I'm going to hook up RC controls. I'd rather have a good telemetry system (probably xbee), and control directly from my pc. Just gotta figure out to write a good base station gui. I only write in C, so not so familiar with .NET or whatever to make good gui's. Looks like most people use Labview, which I don't have access to.

Anyway, not to steal your thread... just wanted you to know that your in good company! Any comments or advice is welcome as well!
eccarm is offline Find More Posts by eccarm
Reply With Quote
Old Nov 14, 2009, 08:58 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
I knew I had a twin separated at birth out there somewhere!

Please don't feel you're hijacking "my" thread. The more discussion the better. Your comments about XBee are interesting. I got a development package from them a year ago and the project fell through, so it's all just sitting on a shelf. I hadn't thought of it in relation to this stuff at all.

Yeah, I'm definitely seeing this as an ARM project too. I'm impressed that people have made it work with AVRs and PICs.

If you run across any particularly interesting threads or sites, please post them.

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 09:00 PM
Registered User
Woodland, CA
Joined Nov 2009
66 Posts
Quote:
Originally Posted by otlski View Post
On tricopters the rear motor and thus the prop steer the craft in yaw. Typically the entire boom is rotated by a servo tipping the plane of the prop. To hold a level hover, the rear motor is rotated to a slight angle to counteract its own drag and the yaw influence it exerts; effectively canceling out the yaw imbalance.

Quickly considered, the efficiency losses due to the the out of plane rear prop would made the tricopters undesirable. However, many of the tricopter proponents would tell you that they have huge choice in prop and motor combinations allowing them the opportunity to make up for any efficiency losses.

I'm sure others will chime in and tell you why they prefer tricopters. I fly quads. I like the control symmetry.

Dan
Got it. Thanks a bunch for the explanation.

Mike
snarflemike is offline Find More Posts by snarflemike
Reply With Quote
Old Nov 14, 2009, 10:30 PM
Registered User
KAMAX-T's Avatar
Powell TN
Joined Jan 2008
735 Posts
Quote:
Originally Posted by snarflemike View Post
Hi Terry,

What are the advantages of a tri-copter? Is it just the obvious savings of one motor/prop/ESC, or is there more?

I know that quadcopters (and other even-motored versions, I'd assume) use opposite rotations to counter torque rotation (maybe not the correct technical term). How do tri-copters deal with this effect?

Mike
Mike I personally like them both , but the tri is so much simpler in terms of setup. Like it was stated , the rear motor is tilted to make it yaw. I think that it is more fun to fly than a quad, because of orietation .

Terry
KAMAX-T is offline Find More Posts by KAMAX-T
Reply With Quote
Old Nov 15, 2009, 04:12 AM
Registered User
Hoofddorp, Netherlands
Joined Sep 2004
3,917 Posts
Of course there is also the Y6 configuration for a hexacopter which combines "the best of both worlds":
  • the attractive Y-configuration of a tricopter
  • the mechanical simplicity of a "quad" or even-numbered multikopter (no mechanical servo on one arm which can fail; only motors, ESCs and the FC and program)
  • redundancy beyond either tricopter or quad: if one motor fails or prop breaks or comes off, you should still be able to land in a controled fashion instead of creating a big dent in the roof of that expensive Jaguar driving just by chance at the location your craft is impacting after the failure

On the downside, the coaxial positioning of the motors in the Y6 (or actually Y3x2) config results in some loss of efficiency of the lower motors and does require some searching for the right motor kV and prop combination. Even with the loss of efficiency, you will still gain more thrust and thus have more "cargo-capacity".

UAVX includes the tricopter option in its open source code. Mikrokopter has a neat way of allowing you to choose various quad (+ or X), hexa (6 or 3x2), and octo configurations.

Even if you join one of the existing open project (UAVX, AeroQuad or MK) you will still have an unbelievably steep learning curve. If you join one which isn't as extensively developed yet, you will still get the benefit of being able to get really comfortible with the current source code and design and contribute a lot to its development.
Arthur P. is offline Find More Posts by Arthur P.
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion can you design a skin for an rc airplane? b-29er The Builders Workshop 20 Oct 29, 2009 04:19 PM
Discussion Designing for storms - or - "throwing design concepts to see if one sticks" MCarlton Slope 15 Oct 17, 2009 06:28 PM
Discussion Club One-Design TD Contest Kai@UCSB South Bay Soaring Society 3 Sep 04, 2009 12:37 AM
Yippee! Win $10.00 by PayPal to the one who can best solve this problem Bluecor Aerial Photography 26 Jun 05, 2005 02:10 AM
what do u mean by foam planes?can any one help kishore Foamies (Kits) 12 Apr 07, 2004 01:13 PM