No, not kidding at all. There's even a company selling similar vehicles except they call them T-copters (just FYI, I would stay away from that company - they have a history of less than honorable practices). The motor on the tail tilts for yaw control.
Flipping a propeller over does not change the thrust direction. If you change the direction of shaft rotation then it would produce thrust in the wrong direction.
You need two pusher props and two regular props. If all motors produce the same thrust you have a stable hover. In order to yaw you need to increase thrust on two motors that turn the same direction and decrease the other two. If they all turned the same direction you would not be able to yaw - assuming that the motors were all parallel to each other.
Why do you need pusher props? Isn't the motor spinning in the "normal" direction so that you can use standard propellers?
As a side note, you should be able to use regular propellers by just flipping them around and changing the direction of shaft rotation.
A pusher-prop is probably the wrong term, but counter-rotating prop is probably the kind we use. The Aeroquad store calls them "pusher props". I really don't know what they're called.
Flipping a propeller over generally doesn't work unless the design was made with this in mind - the shape of a typical propeller on the top is almost like an airfoil - a semicircle if you were looking at it from the edge. Turning it over wouldn't have the same effect except maybe at crazy higher RPM. But, I have seen some cheaper propellers at hobby stores that are asymmetrical and could be flipped over. Generally, with multi-copters you wouldn't use low-cost propellers because they have other problems related to balance, material, etc. There are so many points of failure already.
Thanks Cody. Hope you're doing well - didn't get to chat last time we ran into each other.
Phil: I am not sure about counter-rotating or angle control by a servo. But that video of the tri flipping is incredible. It is extremely manoverable. However, a lot of that is the pilot of course!
Ken: They called the counter rotationg props pushers because they were first used on planes with the engine and prop at the rear, pushing the plane. There are a number of quads that tilt the motor(s) (some fixed tilt, some by servo control) and just use normal propellers.
I didn't realize that two of the rotors spin counter to the other pair. A glance look at the wikipedia page has quickly humbled me...
And now I understand the confusion. In most "pusher" propeller setups, the air is flowing over the motor and then past the propeller, while a quadrotor will use a "pusher" propeller in the opposite fashion by having the air flow from the propeller and then past the motor. So I guess the pusher propeller is mounted in a reverse fashion to the way it was intended...
As a side note, I wonder if there are any ESC's that have built in and runtime reversibility... Seems you could get some really fast flips in then, if your motor can handle the abrupt changes in direction.
Aircraft ESC's won't do reversing, though vehicle ESC's will. Those tend to be sensor based (not sensorless) because of the torque required when starting. They'd probably work, but they're tuned for use by wheeled vehicles. If you're trying to get inverted, or just -really- tight control, variable pitch props are the way to go. You lose efficiency, but not nearly as much as you'd lose by trying to rotate uni-directional props the other way.
If you google tricopter you can find theory and explanation of the flight principle. Some are T shaped, in place of Y. Anyway in all the models I found on the web the tail rotates along the axis with a servo.
Massimo
The quadcopter is a neat project and the videos were fun to watch. The Hoverfly boards are really expensive compared to other quadcopter avionics out there. Besides using a propeller chip (which obviously would appeal to Ken) what features do they have that cause them to be more expensive?
Martin_H,
I don't know what they have that all of the others don't have (except the Prop), It's frustrating to me how expensive their board is. On top of that their source is closed, so you can't add anything to it or change it to suit your specific build. The Ardupilot Mega includes everything that their Hoverfly Pro does plus has GPS for $250, while the Hoverfly Pro is $450 and the GPS option when it becomes available will be $450 also (so $900 total). HoverFly does have the Sport board for $225, but it lacks features compared to others.
Their website boasts about their HoverCore technology fusing the inputs from the Gyros and Accelerometers and utilizing the props multi cores to do it better. They say if you want to write it yourself then go ahead, but after you have you'll appreciate their stuff more... seems kind of arrogant to me. It's highly likely that all they are doing is a Kalman filter fusing of the inputs like everyone else does, then feeding a PID setup to drive the motors. Makes me want to get together with Jason Dorie and show them up with free open source software!
The quadcopter is a neat project and the videos were fun to watch. The Hoverfly boards are really expensive compared to other quadcopter avionics out there. Besides using a propeller chip (which obviously would appeal to Ken) what features do they have that cause them to be more expensive?
Martin, the cost is justified because their customers are using Hoverfly boards to solve a problem: aerial photography, surveillance, movie films, etc. In fact, the same electronics fly around a 5-lb camera so the code has got to be rock solid. I can attest that the code is bombproof, at least from my observation. Even on Sport version - it'll get you up and flying quickly.
I guess you have to ask yourself about the mission when it comes to multi-rotor craft - do you want to fly one or program one? Even if it's the latter, having a board like the Hoverfly will speed your project because (a) you could program it as your own and (b) you can use their firmware for in-flight testing to validate your hardware.
Programming these from the start is going to be a challenge - I've noticed that any little problem with propeller balance, ESCs, vibration can compound through the system.
We will have the Hoverfly Sport boards in stock - with an open schematic, too. You are all welcome to make your own PCB, of course.
Ken,
Sorry if I am a little dubious. Maybe it's because I know software pretty well. The code I end up making for my quadcopter will be at least as stable. Pretty much every quadcopter board that, has 3 axis gyros and accelerometers, out there is stable. In fact many are fairly stable with just gyros. It's not magic anymore. Also, the amount of weight it can manage (when the weight is mounted to the center, especially) is more a mater of the motors and props than the software. The Kalman Filter and PID logic handle it.
I guess I need to get my quadcopter flying to prove my point.
Ken,
Currently, I have the frame I had back at UPEW except with the center bit changed out to some round plastic parts Kevin gave me when I was there, and I also have a set of the props you guys have that Kevin also gave me. I have 4 hacker clone motors and a set of ESCs. My frame setup is pretty heavy, and it's using those aluminum U rods for the 4 arms (which are supposed to be bad with vibrations). It's something that will be tougher to get flying stable, which is a good thing, but having a known good frame/ESC/motor/prop setup would be handy to get started. I did order that frame I linked before from hobbyking (the cheap wood/fiberglass one) and some more props. It's supposed to be very light and seems like it would be easier to get flying then my current setup, we'll see.
Today (Saturday) I soldered up my new control board (prop proto board) with little daughter boards that interface to the receiver, ESCs, and the sparkfun IMU3000/ADXL345 board (digital gyro and accelerometer). I've got code to read the receiver and drive the ESCs, but I need to write the code to talk with the IMU stuff via I2C (should be fairly easy). Then I dive into the Kalman filter and PID code.
Anyway, if you are willing to send me a full frame/ESC/motor/prop setup (I'll pay of course), then that would be great. I plan to spend a few hours tomorrow (Sunday) working on code for IMU board. I've been getting back into the swing of things this week with all of my projects (including the compiler). The next couple months should be much more productive.
Just received my pcbs with servos which will plug into various prop pcbs. I will post a pic shortly. The pcb is 30x45mm and supports up to 16 servo headers. Each 3pin servo header supports a series resistor from/to the prop pin, and the 5V can be linked (because you should not parallel your ESC 5V connections). Now just waiting for my Prop pcbs to arrive.
While NATO countries fly unmanned aerial vehicles (UAVs) high above Libya, none of these UAVs, or the vital intelligence they provide, was available to the Libyans fighting to free their country -- they were fighting blind. So, they got one of their own. The Libyan rebels have been using the Aeryon Scout Micro UAV to acquire intelligence on enemy positions and to coordinate their resistance efforts. This video gives sample photos and video from both the Scout's daylight and thermal payloads.
Ken,
You could be on your way to a new lucrative enterprise...
Interesting share, Ron. I've seen stories about how the rebels are making all kinds of arms on their own. This technology is certainly not out of reach for them.
Here's what I want to achieve with the QuadCopter project:
- increase the Propeller-head participation in this area;
- get people flying using the HoverFly boards and a simple kit;
- have our kit be a community-inspired project rather than a formal Parallax product
- share the design, drawings, etc. so people can make their own, fully open-source
The relevance of the last point is that although we'll stock all the parts required to build one, there's not enough profit in the frames/kit to justify making it an official Parallax product (at least for now). For that reason I wanted to offer it through the forums, not distribution [which changes how we do things].
Don't care about making money from this project. If we start to focus too much on quadcopters then other more important business will be neglected. Consider it a special treat for forum members!
One part that may concern customers is the approach with the Hoverfly SPORT. For those who want to get a tested platform up and flying, this is a great solution. It simply works. For those who want to build their own open-source, DIY effort it is probably not the best approach. Hoverfly will open-source the schematics. This will allow you to write your own firmware or re-load their factory firmware, but does not include access to their source code. If you are happy with their circuit design perhaps this won't be a problem. I'd like to hear your thoughts on this part. Cluso seems to have a solution coming, too.
Also, I should advise some people that quadcopters aren't for everybody. They're amazing to watch and the first thought of many is "I must have one!". Consider the task of flying alone - not even the more difficult part of programming - and I'd suggest pilots really need some prior time on other R/C devices. It takes a lot of attention to detail to assemble the wiring professionally, too. Exposed wires are a problem, as are cheap parts of any kind. Putting the motors together is a careful process too: c-clips, set screws, Loctite, prop balancing, etc. And these quadcopters can crash into walls or people (I flew this at the convention this week with many people around - probably not a good idea). You also need a radio and LiPo charger. But I must say=>
This is the most fun I've had in a long time! Didn't do any programming, but I will in time for the October Propeller Meetup Group!
Feel free to let us know your thoughts on these details about the plan. Would you folks be interested in a simple kit?
Also, I should advise some people that quadcopters aren't for everybody. They're amazing to watch and the first thought of many is "I must have one!". Consider the task of flying alone - not even the more difficult part of programming - and I'd suggest pilots really need some prior time on other R/C devices.
Ken Gracey
I've got a couple of electric R/C planes and I am terrible at flying them.
I've got a couple of electric R/C planes and I am terrible at flying them.
Ron - it's easier to fly a quadcopter than an R/C helicopter, especially if you keep your point of view the same by holding the heading of the quadcopter in the same direction. Then it's a matter of using the right stick to move around and the left one for throttle. And, it's also easier to fly this than an electric profile plane. I think R/C trainer airplanes with plenty of dihedral are easier to fly than a quadcopter, however.
What a fantastic video. I wonder if would be possible to even think about how to program a microcontroller to do this. I can't even understand how some of those moves (dancing on its tail) are even physically possible?!
Comments
Why do you need pusher props? Isn't the motor spinning in the "normal" direction so that you can use standard propellers?
As a side note, you should be able to use regular propellers by just flipping them around and changing the direction of shaft rotation.
You need two pusher props and two regular props. If all motors produce the same thrust you have a stable hover. In order to yaw you need to increase thrust on two motors that turn the same direction and decrease the other two. If they all turned the same direction you would not be able to yaw - assuming that the motors were all parallel to each other.
A pusher-prop is probably the wrong term, but counter-rotating prop is probably the kind we use. The Aeroquad store calls them "pusher props". I really don't know what they're called.
Flipping a propeller over generally doesn't work unless the design was made with this in mind - the shape of a typical propeller on the top is almost like an airfoil - a semicircle if you were looking at it from the edge. Turning it over wouldn't have the same effect except maybe at crazy higher RPM. But, I have seen some cheaper propellers at hobby stores that are asymmetrical and could be flipped over. Generally, with multi-copters you wouldn't use low-cost propellers because they have other problems related to balance, material, etc. There are so many points of failure already.
Thanks Cody. Hope you're doing well - didn't get to chat last time we ran into each other.
Ken Gracey
Ken: They called the counter rotationg props pushers because they were first used on planes with the engine and prop at the rear, pushing the plane. There are a number of quads that tilt the motor(s) (some fixed tilt, some by servo control) and just use normal propellers.
And now I understand the confusion. In most "pusher" propeller setups, the air is flowing over the motor and then past the propeller, while a quadrotor will use a "pusher" propeller in the opposite fashion by having the air flow from the propeller and then past the motor. So I guess the pusher propeller is mounted in a reverse fashion to the way it was intended...
As a side note, I wonder if there are any ESC's that have built in and runtime reversibility... Seems you could get some really fast flips in then, if your motor can handle the abrupt changes in direction.
I may just have to buy one of them.
Massimo
That's a strange claim to make when the props aren't even included.
I'm curious how you did this with a stamp rather than a commercial RC controller.
It's based on that kit being an upgrade of one of their "full" kits that does include the motors, ECS, props, and so on.
Computer-controlled, too...
http://www.thequadshot.com/
http://www.kickstarter.com/projects/jkg/quadshot-an-aerobatic-blend-of-rc-helis-and-planes?ref=NewsAug1911&utm_campaign=Aug19&utm_medium=email&utm_source=newsletter
(kickstarter site, for funding the mass production and preordering kits)
Think flying wing with VTOL and acrobatic performance. and uses Xbee comms...
I don't know what they have that all of the others don't have (except the Prop), It's frustrating to me how expensive their board is. On top of that their source is closed, so you can't add anything to it or change it to suit your specific build. The Ardupilot Mega includes everything that their Hoverfly Pro does plus has GPS for $250, while the Hoverfly Pro is $450 and the GPS option when it becomes available will be $450 also (so $900 total). HoverFly does have the Sport board for $225, but it lacks features compared to others.
Their website boasts about their HoverCore technology fusing the inputs from the Gyros and Accelerometers and utilizing the props multi cores to do it better. They say if you want to write it yourself then go ahead, but after you have you'll appreciate their stuff more... seems kind of arrogant to me. It's highly likely that all they are doing is a Kalman filter fusing of the inputs like everyone else does, then feeding a PID setup to drive the motors. Makes me want to get together with Jason Dorie and show them up with free open source software!
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=1029059&viewfull=1#post1029059
Martin, the cost is justified because their customers are using Hoverfly boards to solve a problem: aerial photography, surveillance, movie films, etc. In fact, the same electronics fly around a 5-lb camera so the code has got to be rock solid. I can attest that the code is bombproof, at least from my observation. Even on Sport version - it'll get you up and flying quickly.
I guess you have to ask yourself about the mission when it comes to multi-rotor craft - do you want to fly one or program one? Even if it's the latter, having a board like the Hoverfly will speed your project because (a) you could program it as your own and (b) you can use their firmware for in-flight testing to validate your hardware.
Programming these from the start is going to be a challenge - I've noticed that any little problem with propeller balance, ESCs, vibration can compound through the system.
We will have the Hoverfly Sport boards in stock - with an open schematic, too. You are all welcome to make your own PCB, of course.
Ken Gracey
Sorry if I am a little dubious. Maybe it's because I know software pretty well. The code I end up making for my quadcopter will be at least as stable. Pretty much every quadcopter board that, has 3 axis gyros and accelerometers, out there is stable. In fact many are fairly stable with just gyros. It's not magic anymore. Also, the amount of weight it can manage (when the weight is mounted to the center, especially) is more a mater of the motors and props than the software. The Kalman Filter and PID logic handle it.
I guess I need to get my quadcopter flying to prove my point.
Do you have frame parts, motors, ESCs? What do I need to provide to get you in the air? Willing to help!
Ken Gracey
Currently, I have the frame I had back at UPEW except with the center bit changed out to some round plastic parts Kevin gave me when I was there, and I also have a set of the props you guys have that Kevin also gave me.
Today (Saturday) I soldered up my new control board (prop proto board) with little daughter boards that interface to the receiver, ESCs, and the sparkfun IMU3000/ADXL345 board (digital gyro and accelerometer). I've got code to read the receiver and drive the ESCs, but I need to write the code to talk with the IMU stuff via I2C (should be fairly easy). Then I dive into the Kalman filter and PID code.
Anyway, if you are willing to send me a full frame/ESC/motor/prop setup (I'll pay of course), then that would be great. I plan to spend a few hours tomorrow (Sunday) working on code for IMU board. I've been getting back into the swing of things this week with all of my projects (including the compiler). The next couple months should be much more productive.
Roy
You could be on your way to a new lucrative enterprise...
How Libyan rebels got drone technology
http://news.yahoo.com/blogs/envoy/libyan-rebels-bought-miniature-surveillance-drone-internet-213029799.html
Interesting share, Ron. I've seen stories about how the rebels are making all kinds of arms on their own. This technology is certainly not out of reach for them.
Here's what I want to achieve with the QuadCopter project:
- increase the Propeller-head participation in this area;
- get people flying using the HoverFly boards and a simple kit;
- have our kit be a community-inspired project rather than a formal Parallax product
- share the design, drawings, etc. so people can make their own, fully open-source
The relevance of the last point is that although we'll stock all the parts required to build one, there's not enough profit in the frames/kit to justify making it an official Parallax product (at least for now). For that reason I wanted to offer it through the forums, not distribution [which changes how we do things].
Don't care about making money from this project. If we start to focus too much on quadcopters then other more important business will be neglected. Consider it a special treat for forum members!
One part that may concern customers is the approach with the Hoverfly SPORT. For those who want to get a tested platform up and flying, this is a great solution. It simply works. For those who want to build their own open-source, DIY effort it is probably not the best approach. Hoverfly will open-source the schematics. This will allow you to write your own firmware or re-load their factory firmware, but does not include access to their source code. If you are happy with their circuit design perhaps this won't be a problem. I'd like to hear your thoughts on this part. Cluso seems to have a solution coming, too.
Also, I should advise some people that quadcopters aren't for everybody. They're amazing to watch and the first thought of many is "I must have one!". Consider the task of flying alone - not even the more difficult part of programming - and I'd suggest pilots really need some prior time on other R/C devices. It takes a lot of attention to detail to assemble the wiring professionally, too. Exposed wires are a problem, as are cheap parts of any kind. Putting the motors together is a careful process too: c-clips, set screws, Loctite, prop balancing, etc. And these quadcopters can crash into walls or people (I flew this at the convention this week with many people around - probably not a good idea). You also need a radio and LiPo charger. But I must say=>
This is the most fun I've had in a long time! Didn't do any programming, but I will in time for the October Propeller Meetup Group!
Feel free to let us know your thoughts on these details about the plan. Would you folks be interested in a simple kit?
Ken Gracey
I've got a couple of electric R/C planes and I am terrible at flying them.
That's why I am amazed when I see guys like this:
Ron - it's easier to fly a quadcopter than an R/C helicopter, especially if you keep your point of view the same by holding the heading of the quadcopter in the same direction. Then it's a matter of using the right stick to move around and the left one for throttle. And, it's also easier to fly this than an electric profile plane. I think R/C trainer airplanes with plenty of dihedral are easier to fly than a quadcopter, however.
Ken Gracey