• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

My Octocopter Project!

VolksRocket

Rocket Scientist
#1
Hi,

This thread will be for my latest OctoCopter Project! I think you will find it very interesting!

The first discussion point is to determine if I want to build a Octo + or Octo x configuration?

Let me have your thoughts!
 

VolksRocket

Rocket Scientist
#2
Reason for determining layout?

Hi,

Ok, I wanted to add some details to this thread. The reason I ask which layout Octo + or Octo x, is I am trying to determine which will be better for performance, maneuverability, and engine out safety?

What are your thoughts?
 

VolksRocket

Rocket Scientist
#3
Here is the layout pics!

Hi,

Here are the layouts I am looking at, so which one would be the best for safety and engine out situations, which has the best failure mode forgiveness?

Also, is the programming for different engine coordination the same or dependent on the flight envelope specs.?
 

Attachments

cranialrectosis

Faster than a speeding faceplant!
Mentor
#4
I don't think the layout makes a huge difference on an octo. X quads are more popular than + quads because you don't have a motor in front of an FPV camera. With an octo, you likely have the cam below the motors making it a non-issue.

I would be stunned if there is a difference in reliability. A modern flight controller such as an APM really shouldn't care.
 
#5
Hi,
Also, is the programming for different engine coordination the same or dependent on the flight envelope specs.?
When setting it up in Mission Planner or APM Planner2 just select your frame type and that's all the programming that is needed for either + or x configuration. None of the multi's are 100% reliable with a motor out situation. Sometimes they can land safely, sometimes they can't and just flip over and crash depending on how much symmetric thrust is required to control pitch or roll attitude. The Pixhawk (or whatever you use) can handle it with no problems. Many times the mechanical setup of the aircraft can't and it simply runs out of thrust to control attitude and crashes.

So in the end, whether it's quad, hex or octo, + or x configuration, doesn't make one bit of difference. Even with a quad the flight controller will throttle back the motor opposite the failed one and use torque difference with the other two to control descent rate, pitch, roll and yaw. Success rate in the real world making a successful safe landing with motor out with any of them is <50% simply due to a multirotor not being a real aircraft that actually flies, it has no autorotation capability like a helicopter, and is inherently unstable requiring corrections by electronics hundreds of times per second to make one even leave the ground.
 
#6
Are you using the Ardupilot firmware? Otherwise, your motor layout might be different.

I learned this the hard way because the motor map for tricopter is different on Ardupilot & PX4 firmwares.

megabotz
 
#7
Are you using the Ardupilot firmware? Otherwise, your motor layout might be different.

I learned this the hard way because the motor map for tricopter is different on Ardupilot & PX4 firmwares.

megabotz
I assumed the OP is using either an APM or Pixhawk with Ardu, since that's the motor layout he posted a picture of. I've flown ArduPilot in various aircraft, including multi's, fixed wing and helicopters, since the APM2.5 days. I had a 650mm quad rotor multi that had an ESC go up in smoke in flight flying a Pixhawk v1 with AC3.3.3 and the autopilot successfully kept control of it and landed it from 165 feet altitude with no damage to the quad with the RF motor completely dead. But that doesn't happen every time when you get a motor or ESC failure with a multi.

Going from four motors to eight doesn't increase the chances of it being successful either. What it does is double the chances that you will have an in-flight failure.

Multirotors are the worst case possible to discuss "reliability" and "safety". They don't actually fly. If you drop an airplane, a helicopter, and a multi, all from 200 feet, and all with no power to the motor(s), both the airplane and the helicopter can be landed without damage by a competent pilot. No matter how good the pilot is, all you'll get is an epic crash with a multi.

So poring over multi-rotor configurations trying to reason "reliability" or "safety" is pointless. Tri's are the most efficient but have no chance of recovery with a motor failure. Quads are less efficient than a tri, but are popular because of their low cost and simplicity. Hex's are less efficient than the above but can typically lift more than a quad or tri with shorter flight time. Octo's are the least efficient of all of the above and are typically used for heavy lift, very short flight time applications.

You introduce losses with each ESC, motor and prop. There are tip losses on the prop. The more props, the more tip losses you have, and the more props you have running in "dirty" air from the tip vortices and turbulence in front when the thing is moving. Each additional ESC and motor has electrical losses. The more you add, the more of your battery energy gets converted to heat instead of thrust. When you lose a motor on an octo, due the entire thing being less efficient than say a quad, it has no better chances of recovery than a quad or hex. The props are all operating at less than 50% efficiency due to the close proximity of tip vortices from other props and loss of thrust on one side will cause the flight controller to start throttling motors to maintain control of it. But your normal hover power is now adjusted to a higher level, the remaining motors at 90 degrees to the failed one are operating at continuous elevated power to provide the vertical thrust vector. The motors opposite the failed one are operating at very low power trying to maintain pitch or roll attitude. Since the small ESC's typically used on octo's are not designed to operate at elevated power for more than about 15 seconds, you are now in danger of burning out another one. And that's usually what happens with octo's is the domino effect of losing one ESC, and then a second one being overloaded, and the thing crashes within 15-20 seconds of losing the first motor.

There are instances where, just like my quad that had a motor quit, the flight controller can bring it in safely. But it just doesn't happen that often.
 
#8
I assumed the OP is using either an APM or Pixhawk with Ardu, since that's the motor layout he posted a picture of.
I know, just making sure.

If you drop an airplane, a helicopter, and a multi, all from 200 feet, and all with no power to the motor(s), both the airplane and the helicopter can be landed without damage by a competent pilot.
Can you explain about the helicopter? I understand this to be true of airplanes & my most favorite glider is the space shuttle but I imagine a helicopter without power would drop like a stone. I would also think that the rotor would spin, or more likely, cause the fuselage to spin because of the air passing the rotor. I agree that the only thing lifting a multirotor are the props but after building a tricopter, I realized that multirotors only differ from helicopters in a few details. Big details, but not many.

megabotz
 
Last edited:
#9
Can you explain about the helicopter?

megabotz
Well, I'm primarily a helicopter pilot and while multi's and helicopters are both rotary wing, the similarity pretty ends there. Collective pitch helicopters can auto-rotate quite well. They have an advantage over multi's in that you can pull the collective into negative pitch and get the headspeed up (or keep it up) in descent. Then before contacting the ground at high rate of descent, push in positive collective pitch and use the inertia stored in the main rotor to provide lift, flare, and make a controlled landing with both collective and cyclic pitch. In autorotation with a heli you have full yaw control due to the fact that the tail rotor is driven by the autorotation gears in the transmission. The torque is in opposite direction of powered flight. So if you have a helicopter with a clockwise turning head under power the normal anti-torque direction is left. When the helicopter enters autorotation, the anti-torque direction is now right.

Autorotations are one of the first basic things helicopter pilots learn, right up there with forced no-power landings learned by fixed wing pilots. Helicopters are also considerably harder to master than multirotors, and most in RC are not successful learning to fly one without hands-on instruction from an experienced pilot and/or lots of hours spent on a simulator to learn how to fly one. I've found that multirotor pilots with no experience with helicopters will crash a helicopter within 15 seconds of takeoff with some really wild hair raising moments in those first 15 seconds of flight before the crash.

In RC helicopters, 500's and up autorotate extremely well. Here's a vid of a pilot practicing autorotations with a Trex 500. This is something multirotors will find difficult to do:

 

VolksRocket

Rocket Scientist
#10
Wow great answers!

Hi,

Thanks for all the great answers! Helps a lot!

The main question I have is about the layout O+ or Ox, about the programming for control, safety, engine out, and crash avoidance.

So could I get everyone's view on their preference between an O+ or Ox and why? (I may just build both to try)

The actual flight controller has not been selected yet because in the end I will be designing, building, and programming it myself, BUT until then I am planning to use a KK 2.1.5 to get it into the air, then I would move to a programmable flight controller and then to the designed controller.

Note: All this is subject to change during the actual build but I gotta start somewhere.

More specs:

Engine Center Diameter - 30"
Propeller Diameter - 10 x 4.5
Engines - 2830-1100 Turnigy
ESCs - 25A Turnigy
FC - KK 2.1.5, Programmable, Self Designed
Batteries - 11.1V (3C), 2200mah, 25D Turnigy
Radio - TX and RX Spectrum
Camera - 700VTL
Camera TX - 700
Camera RX - Uno
Power Filter - 12V/5V
Parachute - 48" NRN
 
Last edited:
#11
Engine Center Diameter - 30"
Propeller Diameter - 10 x 4.5
Engines - 2830-1100 Turnigy
ESCs - 25A Turnigy
FC - KK 2.1.5, Programmable, Self Designed
Batteries - 11.1V (3C), 2200mah, 25D Turnigy
Radio - TX and RX Spectrum
Camera - 700VTL
Camera TX - 700
Camera RX - Uno
Power Filter - 12V/5V
Parachute - 48" NRN
That looks like all pretty much standard design for an octo. Except for the parachute.

Whether it's + or x won't make one bit of difference. That's only really a consideration on quads where you might not want a prop showing up in your video on your camera, so you use x configuration. Most octo's or hex's will have the camera mounted low enough so that the props don't get into the picture when it tilts to go someplace. The flight controller (at least ArduPilot-based controllers) do not care one bit about that. They fly equally well in all directions. So which way is "forward", sideways, backwards - the flight controller don't care. It's mainly just for your orientation as the pilot, and which way the camera is pointed.
 
#12
Thanks for the explanation, ChrisOlsen.

Looks like the octo x & + motor layout is exactly the same for both firmwares lol

Was different for tricopters, don't listen to me :cool:

Lots of FCs available. There is the pixhawk mini, pixhawk 2.1, Navio 2 among the latest. The Navio isn't as new as the 2 pixhawks I mentioned but is unique because it piggybacks onto the Raspberry Pi, adding many fun options.

megabotz
 
#13
Was different for tricopters, don't listen to me :cool:

megabotz
The only difference between PX4 and Ardu is the output channels. PX4 goes RF-1 LF-2, tail-3, servo-4. Ardu goes RF-1, LF-2, tail-4, servo-7. Those outputs can be mapped in the RC params any way you want. Those are just the defaults.

Also with either PX4 or ArduPilot stack, you can run the motors any direction you want with a Tri and the FC don't care. On my 1 meter Tri/fixed-wing hybrid I run the RF motor CCW, and the left CW with no change in the code. And that won't even appear in the motor maps.

Quads, hexes and octo's, however, must turn the direction indicated in the motor maps. The FC depends on that with those frames for yaw control. A Tri doesn't because the servo actually controls yaw and there is no torque compensation used at all with motor throttling on a Tri - only attitude control.

Lots of FCs available. There is the pixhawk mini, pixhawk 2.1, Navio 2 among the latest. The Navio isn't as new as the 2 pixhawks I mentioned but is unique because it piggybacks onto the Raspberry Pi, adding many fun options.
megabotz
There is also APM2.8, Pixhawk v1 and v2, PixFalcon (basically the same as the mini from 3DR), pixracer, and Earle Brain. Of all of them the APM2.8 and Pixhawk v2 are most readily available from almost anywhere. The mini from 3DR, due to using the third gen ARM cpu, can only run AC3.5 and newer on the ArduPilot side. The Pixhawk Mini is an excellent flight controller for helicopters of 450-500 size class where a full-sized Pixhawk may have limited mounted options.

I would recommend a Pixhawk 2.4.8 with the second gen CPU for an octo for utmost reliability and ability to run any of the code back to 3.2. It is the most tried and tested of all of them. There is supply problems right now with the Pixhawk2.1 and long waits to get one. The batches are sold out before they are even built.
 
Last edited:

VolksRocket

Rocket Scientist
#14
OK which way for camera?

Hi,

Good points, which you made me think, the flight camera will be in the pilots seat view, so do you think the OCTO X would be the best layout from a pilots view?
 
#16
With an Octo I don't think it makes any difference because there's so many arms sticking out of it, no matter how you orient it, if the camera is not low enough you'll have props in the picture. Even a narrow view camera will have a FOV of about 84 degrees. When you have arms every 45 degrees it's kind of a moot point.
 

VolksRocket

Rocket Scientist
#17
Yep, you're right!

With an Octo I don't think it makes any difference because there's so many arms sticking out of it, no matter how you orient it, if the camera is not low enough you'll have props in the picture. Even a narrow view camera will have a FOV of about 84 degrees. When you have arms every 45 degrees it's kind of a moot point.
Yep, I think you're right, but I was just wondering if one way would be better if we were sitting in the pilot's seat.

Well, I have the frame cut out, will post pics soon!
 

VolksRocket

Rocket Scientist
#18
OK, I am back! Sorry, was busy with designing plans for an UL (Ultra Light), SL (Sport Light), and NNA (N Number Aircraft)!

I am done and I can focus on this project now!

There will be a build webpage, and plans when finished!

Going to start a go fund me page, go for sponsorship, and those who just want to donate, to see a VIABLE Personal Multirotor Design, that you can build in your garage!

OK, the FIRST question is (to all the "experts" out there), what is the best layout to use for a manned multirotor? Your choices are: (1) Octocopter Straight, (2) Octo +, or (3) Octo x, layouts which would be best for safety, engine out, and programming?
 
#20
OK, the FIRST question is (to all the "experts" out there), what is the best layout to use for a manned multirotor? Your choices are: (1) Octocopter Straight, (2) Octo +, or (3) Octo x, layouts which would be best for safety, engine out, and programming?
None. Multi-rotors are not safe for manned flight. Period. If you want to go to manned flight, build a helicopter. Helicopters are more efficient than multi-rotors anyway, require no electronic stabilization of any kind, and can be safely landed in the event of power failure. Multi-rotors are very electrically complex machines and in-flight electrical failures comprise over 90% of all aviation-related incidents in commercial and general aviation.