Still waiting

Propellers are sitting safely in the box they were shipped in, cluttering my work desk. But the big package from Singapore is still en route. Tracking says it left for the USA yesterday evening, local time.

Meanwhile, my minor concerns about whether the cheap and simple ESCs and flight controller board will really work are starting to bloom into a touch of paranoia because I can't verify until I actually build.

I guess I should just blot that out with lots of hours of Skyrim.

Or I could pester people on Quora about their understanding about how privacy will be affected by UAVs.

No, it's going to be Skyrim.

FAA to (finally) publish rules on drone use?

This weekend the Los Angeles Times did an article on commercial use of drones, which is not currently allowed. (Article can be found here)

The article lists several civilian uses, such as law enforcement and agricultural uses. The comment section is full of misunderstanding and paranoia. It seems that people think these devices can just hover in the air indefinitely. I would guess that they would probably have a different opinion if they knew that flight time was limited to 20 minutes.

Constructing a military drone. Photo by Pierre Gazzola

 Obviously, I am all for clear and usable regulations for RC Multicopters and UAV drones. But I am also aware that existing laws governing expectation of privacy and photography still protect people, even if there is new technology. I can't climb a tree in your yard now and shoot pictures through your bedroom window. Tomorrow I won't suddenly be able to take the same pictures just because I use a drone instead of climbing the tree myself.

 It is unclear to me why privacy concerns are among the considerations of the FAA's regulations. Federal law touches very rarely on privacy matters. Instead states legislate these issues. Outside of surreptitiousness wiretapping and medical privacy, citizens are generally just left with state laws covering Peeping Toms, recorded phone calls and other privacy issues.

Obviously cameras can't be banned on drones and RC: FPV piloting depends on having eyes built into the device. Yet paranoid scenarios sort of miss out on the idea that aerial surveillance (police or civilian) still falls under existing laws that limit what people can do. I guess we'll know in less than two months, but I am worried now that the money I have spent so far will be wasted it my intended project is too heavily regulated.

Launching a UAV drone. Photo by Larry Zou.

Waiting for the mailman

The bulk of the multicopter parts are on order from HobbyKing USA. But since this is just a local shopping site for an Asian manufacturer and distributor, the parts will be taking quite some time to get here. Tracking lists the items as processing at the sorting center in Singapore, awaiting shipping to the USA. Interesting, since HobbyKing has been loudly congratulating themselves on the opening of the US warehouse. That explains why shipping was so expensive, I suppose.

 Meanwhile, UPS estimates that my props will arrive from Aeroquad on Wednesday, the day before Thanksgiving. I supposed I can show them to the family, thus avoiding awkward or dull conversations for at least 15 minutes.

 The Beginner Makes a Multicopter blog will be taking a holiday hiatus, as you might guess. Not only am I awaiting parts, but I will be working on location shooting junior football tournaments in Auburn, WA in between holiday meals with the family.

What other people are doing

Another Friday link, photo and video post.

Links
A point-by-point article discussing the differences between learning to fly an RC helicopter and a quadcopter, focusing on use for aerial photography: See this blog.

Google apparently experimented with aerial photography using a "Microdrone". Wonder if this project is still in action. Article on the Registry.

Photos
There are some great photographs on this German language blog.



Video
Here YouTube user Andrita35 installs a panoramic camera mount on the bottom of his hexacopter:



At this site, you can see a Flash-based display of the resulting panorama:


Commercial video for swimwear utilizing a Droidworz Hexacopter:


This is the sort of work I hope to be doing next year. Finger's crossed....

See ya next Tuesday.

Get out your credit card...

With the airframes at HobbyKing now in stock, it's time to try and acquire all of the components I will need to build this quadcopter.

Photo credited to  www.seniorliving.org; found on Flickr.

Here is a list of the parts that I will need, along with a link to an adequate source on HobbyKing:

Flight Controller (HobbyKing Quadcopter Control Board V2 (Atmega168PA) $24.79

  [If you recall, I am using a different flight controller from Aeroquad. But Aeroquad is on back order and this one is cheap, although not modifiable.]

Airframe (HobbyKing Quadcopter Frame V1 (plus an extra for repair parts) $29.98

Radio Transmitter/Receiver (Turnigy 9X 9Ch Transmitter w/ Module & 8ch Receiver) $54.99

4 ESCs (Hobbyking SS Series 15-18A ESC) $26.00

  [Why these simpler controllers? Because they are in stock, cheap, and the motors I have are rated 14-16amps.]

4 Brushless Motors (hacker Style Brushless Outrunner 20-22L) $58.88
  [Fairly powerful 924 kv motors, good for upgrading to octo later]

LiPo Balancer Charger (HobbyKing Simplex 1~4S LiPo/LiFe 12,110~240v charger) $27.59

LiPo Battery (Turnigy 3600mAh 3S 20C Lipo Pack) $21.13

Battery Monitor (HXT Lipo Monitor 3S) $2.10

Servo Leads (10CM Servo Lead (JR) 32AWG Ultra Light (10pcs/bag)) $1.99
  [These connect the Rx to the flight controller]

3.5mm Bullet Connectors for the motors (3.5mm 3-wire Bullet-connector for motor) $4.26

4mm Bullet Connectors for power and ESCs (HXT 4mm Gold Connector w/ Protector) $7.28

Shipping from HobbyKing: $71.98   [yikes!]



Propellers (plus two spares) (APC 10x4.7 plus and equal number of  APC 10x4.7 Pusher) $34.00

Shipping from AeroQuad: $14.91


12 Gauge stranded wire (black and red) [Local store]

Heat Shrink Tubing [Local store]


Grand Total for the first project toward the final goal: $379.87


Remaining expenses:
AeroQuad flightcontroller (about $200)
Components to handmake carbon fiber/aluminum Octocopter frame (about $80)
Four more motors, ESCs and propellers (about $120)
Camera mount and servos (??)

Time to start saving while I wait for shipping to get these parts to my home.

Radio controller or Rx/Tx

Hobbyists use a shorthand for receiver and transmitter in radio control: "Rx" for receiver, "Tx" for the other, and "Rx/Tx" as a short of shorthand for the process of doing radio control. And then there is just "RC", as in an "RC car".

Back in the day, radio transmitters were used to send analog pulses which were interpreted by receivers and turned into movements of servos. These servos were then used to manipulate the controls of the car, boat, airplane or whatever.

Nowadays, transmitters are computerized packages that bundle several signals together (channels) into one multiplexed signal. The receiver then takes this signal and breaks out the individual channels into separate PWM (pulse width modulation). Then the individual signals can be turned into meaningful commands.

Photo by Tony Podlaski

So what decisions go into selecting a radio control system?

Channels
The number of channels that your receiver can receive is the number of different controls you can operate on the RC device. For an RC car, you need steering and throttle, both of which can be controlled through a range of motion. For a plane you need at least three channels (aileron, elevator and throttle).

Multicopters with just about any number of motors have a computer (the flight controller) that manipulates the speed of each motor to produce movement along various degrees of motion. Run all of the motors faster and the airframe goes up in the sky, slow them all and it descends. Multicopters can also tip in any direction, causing forward movement in that direction. And spinning clockwise moving motors faster than the counterclockwise motors (or vice versa) causes the vehicle to rotate.

So since all of these movements are controlled by the multicopter, the radio controller just needs to give movement directions to the flight controller, rather than having a human make individual control inputs for each motor and function, as one would do for a single rotor helicopter.

With three directions of movement, three channels are needed as a minimum. Additional channels are used for other commands. In fact, the Arduino-based flight controller I settled on is built around receiving from an eight-channel receiver. (Details on that board here.) This leaves plenty of control channels for later operating a camera mount.

Frequency
There are a few choices you could make on frequency, such as Rx/Tx that broadcast in the range set aside for only model aircraft (72Mhz), but the "new" technology is the 2.7Ghz Spread Spectrum. It's one drawback is that smaller radio waves--gigahertz are smaller than megahertz--do not bounce off of objects and terrain very well. This means you have to keep line-of-sight with your model.

But the benefits are many: Tx and Rx pairs are assigned unique numbers so you will never conflict with another operator. In fact, the Spread Spectrum technology means that you are bouncing from frequency to frequency every few milliseconds, so it is hard for any radio interference to send your aircraft into the ground.

Decision
With all of the bells and whistles that transmitters can come with, it's hard to pick one without experience. In the end I'll just go with the Turnigy 9ch model recommended by Frank in his Instructable (see links on right). It comes with an 8ch receiver and can be programmed in various ways, should I ever advance to that stage. Most importantly, it's only $55 at HobbyKing.

Everything seems to be on back order...

The flight controller I selected is out of stock. The airframe I picked for the first stage is also not due back in stock for as many as 30 days.

Looks like I'll just have to wait.

NEXT POST: I try to wrap my head around radio transmitter/receiver sets and pick the right one for this project.

Great examples of multicopter photography/video

On Fridays I intend to list links, pics and videos from other people that have been flying multicopters. Following the long tradition of posting time-consuming media for the destruction of productivity at the end of the week. That's what blogs are for, right?

First a couple of blogs and articles:
From June 2008, a short blog from an unnamed author documenting his build process
An amazing article about preparing and then field modifying a quadcopter intended to track bats in flight.

Flickr user sjardine has taken a gallery of aerial photos with his quadrocopter:

Photo gallery found here.

Video taken by a hexacopter flying in Cameroon, Africa:

And a pro aerial photographer's demo real:


Copter Kids Aerial Reel from Trent Palmer on Vimeo.

What's in a name?

I have found the following generic names used for various configurations of multicopters:

Tricopter
Quadcopter
Quadracopter
Quadrocopter
Quadcopter
Quadrotor
Hexacopter
Octacopter
Octocopter
Octorotor
Multirotor
Multicopter

And then there is the E-Volo. Sixteen moderately sized motors provide enough lift and redundancy to lift a person into the air. This video was just filmed last month.

Deciding on an airframe

There are several key factors when deciding on an airframe, but as I see it, these are the three main factors:

• Weight
• Strength
• Cost

There are other factors, but many of them are not very flexible once you have set other elements of your design. For instance, if you are going to be lifting a heavy load, your motor size and prop radius are decided by the lifting power you have calculated. At that point the size of your airframe is more or less determined by the spar length, which is determined by the prop radius (and a few other factors).

A almost complete Arducopter kit, photo by Ole Bendik Kvisberg

It should come as no surprise that weight, strength and cost are all directly related. Choose components that alter one factor and you have surely changed one or both of the other factors. Using solid aluminium spars for the motors instead of hollow tubes increases all three factors, using carbon fiber decreases weight, increases strength and cost.

It is possible to get the most efficient design if you are some crazy engineer that can calculate the strength needed for operation or crashes. But from what I'm reading on the web, people just guess. Plus one needs to normally use easily available parts, rather than custom ordering the exact thickness and material that an engineer might determine is the most efficient.

But then, you also have to build it.

Photo by Kevin Vertucio

A kit is far easier than a handmade frame. There cost of the kit is certainly more expensive because the value of the design and identically produced parts increases the price vendors can ask for the kit. Sometime this cost is, er..., rather high ($1095.00 Droidworx AD-8 HL Airframe Kit).

Hand-making your frame does save you money, but takes more time since you will need to custom manufacture some parts. Alternately you can buy some of the frame and make the rest yourself (example: a frame plate sold individually by Mikrokopter US).

Here are there resources I am using to make my decision. A couple of handmade frames and a couple of kits:

• Cheapest kit on the planet. As used in the Instructables quad (see link on right). $15
• 1000mm Quad Copter Design (by "Old Man Mike"). About $35.
• Step-by-step frame build. Price not listed, but carbon fiber plates suggests $60+.
• AeroFPV Frame. Open source maker of the Aeroquad, costing $200
• Mikrokopter Okto XK Frame Set. Commercial frame costing $280
• Carbon SCARAB Quad/Octa. All carbon fiber frame by MultiWiiCopter, for a bling price of $350

Decision

For the prototype, which will be small anyway, no reason not to go with the cheap plywood frame from HobbyKing for $15. I can get two and thereby have spares arms for the inevitable crashes. This kit even comes with a camera mount tilting camera mount!

For the heavy lifter I believe I am going to go with a custom frame using carbon fiber. Keeping the frame as light as possible enables more flight time or more lifting power for the camera. Hand-building will keep the costs down.

Selecting a flight controller

What is a flight controller?

A multicopter uses several motors to spin multiple rotors, somewhat like a helicopter.

A Quadcopter.  Photo by no.

The spinning of any single rotor is going to introduce a spin to the motor too (Newton's Third Law). Helicopters deal with this by using a tail rotor to counter-act the spin. Multicopters deal with this by using an equal number of engines and having half of them spin in the opposite direction.

Flying an RC Helicopter
Photo by Deep Frozen Shutterbug

But there is a whole balancing act beyond keeping the torque of the motors (called yaw) under control. Multicopters will flip if the force of the engines is too far out of balance, but balance it just so and the aircraft's pitch tips in the direction of travel.


No human would be able to remotely balance the motor speeds of three or more engines while standing on the ground. Instead multicopters depend on a microcontroller on a circuit board to read position sensors and make minute adjustments to the motors according to a program. This bit of electronics is called a flight controller.


So, the brain of every multicopter is a flight controller. In my case I expect to upgrade the frame and motors over the course of the project, but keep the same flight controller, updating the software as needed. So I had better start there when making decisions about what to build.


There are a variety of flight controllers and software to choose from. It took me several days of reading bits and pieces off of websites to reduce my apprehension over which to select. Here are some of the websites I had to consider:


The Multicopter Table. A comparison, somewhat out of date, listing dozens of open source and commercial options for flying a multicopter. This lists the base capabilities of hardware and software. DIY-types can program more capabilities themselves, in theory.

Frank26080115's Instructables step-by-step on building a quadrotor helicopter. Here is shows how to take an Arduino microcontroller and build a flight controller board from scratch. He goes into depth about the inner workings and has code examples. Perhaps too advanced for me now, but a great reference.

Aeroquad has flight controller kits and open source software. For about $200 you can have a full flight controller, with sensors and expandable options.

Multiwiicopter actually has pre-built flight controllers, again with all of the sensors you might initially need. These are sold as "plug-and-play." These flight controllers are built around the Nintendo Wii accelerometer, which is apparently a highly sensitive, but low cost sensor.

And then there are commercial vendors, like Mikrocopter (US sales site). Certainly fully capable, often built-in with the code to add camera mounts and GPS already. But very much in the pricey range, starting at $460 and going up from there.

Decision

After looking at all of these options I have elected to go with an Aeroquad kit. It comes with the four main sensors. And it has a shield to make installation of add-ons easier. The shield also has ten PWM outputs for motors, ten servos outputs for cameras and stuff, and pinouts for GPS and video display. When compared to the best board by Multiwiicopter, it is obviously heavier, but clearly has more motor controllers and extra servo outputs.

I do find it interesting that pretty much none of the suppliers list the weight of their flight controller. I guess from their point of view, you don't really have a choice. You need one (theirs) in order to get started. But the pictures give you a general sense of the size and that is a primary indicator.

Okay, so now to order the Aeroquad v2.0 kit...

Aw crap, it's out of stock. Well, I set up an e-mail alert for when it is back in stock and will wait a week or two. Meanwhile, time to figure out what I want to do for a frame.

NEXT POST: An airframe. Do I go with a kit, hand build or something in between.

What is the design goal?

Having never built anything like this, it is probably helpful to consider what my goals are and what steps I plan on using to get to the end result.

Main Goal: Sustained flight with 4 lb payload
Payload: Canon 60D (775g / 1.6 lbs); radio controlled, stabilized camera mount (abt 600g / 1.3 lbs); lens (abt 1 lb)

Hey, this French guy brought a hexacopter to
 Burning Man! I'm so jealous.

First step: Build a working multicopter and practice piloting
At first, a working RC quadcopter at a smaller and more affordable size. Considering the process of upgrading in this project, the flight controller card should be capable of working first the small size, but then upgrading components and software to become a camera-controlling hexa/octocopter.

Second step: Improve working multicopter
After successfully building and flying the first vehicle, add a light mount for my ContourGPS helmet cam (170g / 6oz.). Then look at upgrading motors, installing controlled camera mount, upgrading programming.

Third step: Build full-sized frame and servo-controlled camera mount
After the "prototype" is successful, look at building a full sized frame with 6-8 engines (either a hexacopter or an octocopter). Upgrade programming to handle additional engines. Also consider increasing sensors (GPS?) and auto-flying options (self-landing), before testing at full weight. Then test fly under the safest conditions and verify flight time at over 10 minutes maneuvering.





Needed for this project:

• Flight controller board that "does it all" and grows with the project.
• Initial simple frame. Something cheap most likely.
• Big, strong and light "primary" frame (carbon fiber or aluminum?).
• Motivation. This is going to take at least all winter.

Other factors to consider:

• Existing code that is modified is much easier than writing from scratch. Code samples from the community are even better.
• The flight controller board needs to be a kit or an completed hardware product. Onboard soldering is allegedly very difficult. Making something from scratch is probably not an option for me.
• Keeping costs down would be really nice, you know, so the wife doesn't think I've gone insane.

NEXT POST: start off by figuring out which flight controller to order?

Who am I and what do I want a multicopter for?

My name is Todd Gardiner and I am a both a professional photographer and a software test lead. I like to work so much I have two jobs. I also go to Burning Man; ten years running now.

I specialize in documentary photography and shooting events. You can see my photography here:
http://thebes-inc.smugmug.com/

I want a hexacopter (maybe an octo) to lift my Canon 60D into the air for aerial photography and video. Hopefully I can even turn that into a business model, but that's going to take a while. The laws on commercial use of RC aircraft are still in limbo, although I am allowed to shoot things for my own purposes and even sell that footage. But operating for-hire could become illegal.

I don't know much about electronics. I have done some simple circuits, rigged lighting for things, played with EL-wire, but I have not worked with microcontrollers, radio control (RC), or other electronics. I'm not skilled at soldering, but I've done it before. I don't know the programming language and tools that Arduino uses, but most of my coding skills have been learned by playing with existing code; since there is already code for controlling an Arduino-based multi-copter, I should be okay here if I go slowly and test as I go.

So, if you are like me, this might be a good blog for you. I am about to embark on reading up on multicopters, selecting hardware, and then building my device. One day you too might be able to get your camera into the air.

--Todd