As some of you might have understood I’ve done some experiments with multirotor helicopters for the last couple of years. Mainly because I think it is extremely interesting what you can do when you combine these machines with cameras. Stuff like this:
And if you want the best images you need to be at the right place at the right time. And for images like this you need to be there with your copter. So the copter needs to be easy to carry.
I wanted to build a FPV quad that I could carry in my ground station case together with my radio and some batteries. I have built several mini-multirotors that fits in that space, but now I wanted to build a bigger quad that should be able to carry a bigger video Tx, gopro and even a Sony RX100.
I ended up building a foldable quad. So far I am very pleased with the result.
I’ve built several multirotors during the last two years. Quad, hexa, Y6, X8, tri, octo. I’ve built in wood, alu, carbon fiber and combinations.
But i like wood. It gives less jello in my recordings because it is slightly flexible so it absorbs some vibrations. And the flexibility also makes sure that it survives a lot of crashes. But it is not too flexible, so it is stable. And it is cheap. And environmentally friendly. Doesn’t block radio signals. Doesn’t transfer current. And it is easy to work with.
So, here is my wood based foldable quad:
There is a bit of jello in the recordings here. It is 2.7K @ 25p with no ND filter. Will probably get rid of it with some tweaking of the mount.
First FPV test flights. Now with GoPro Hero2 at 720@50p and a horribly misplaced ND filter. But: absolutely no jello!
The arms are tightened with four screws. Folding and unfolding takes approx 30 seconds.
Motors: Sunnysky 2212 980 kv
ESC: HK F30A SimonK
FC: DJI Naza GPS
Props: 8 inch
Motor-to-motor: 58 cm
GoPro mount with Tilt
First of all. If you want to know what I’m up to next, follow me on twitter: @eirikso
So, I posted this video on YouTube two days ago:
And it have gone completely viral. It was made in a hurry and leaves a lot of things that might need an explanation.
The place and the moose
The video was shot just outside of Oslo. We have a lot of moose roaming around the woods here in Norway. And some of them very close to the city. They’re shy of people, dogs and other living creatures, but are pretty used to the sound of cars, trains, helicopters, planes etc. To see a moose is not very uncommon in Norway.
This moose is probably a 1,5 year old female. She seems more curious than afraid. And we spotted her again from far up in the sky later. Still calmly walking around eating in the same wood.
The technology in general
One of my hobbies for the last year have been to build and fly multirotor helicopters. When I fly them I do it either by flying Line-Of-Sight (LOS) or First Person View (FPV). When flying LOS you simply take off and fly the thing around while watching it from the ground. For quadcopters this limit the range because the copter is symmetric and you loose the orientation pretty fast. However, as I’ve gained experience you start to feel the direction based on speed and the way you control the copter, making it possible to fly further away. I fly LOS when I want to do acrobatics and train accuracy and pilot skills. Here is an example of some acro flying with a very small quadcopter:
But for the video with the moose I fly FPV. Meaning that I have a dedicated camera connected to a video transmitter on board the copter. I feed the video signal into a pair of video goggles and navigate the copter like I was sitting inside it. Based on the video signal. When flying this way I always have spotters beside me that keep an eye on the copter and what is happening on the ground.
In addition to the goggles I have a screen and a digital recorder on the ground station. That recorder also has a mic and records the audio from the spot where we are standing. Ensuring that we got the recording of the quite enthusiastic spotters and my rather nervous comments about battery status and the fact that it is darn difficult to navigate this thing between and under the trees.
On the copter there is also a GoPro Hero3 to record HD video on board. The video downlink is only SD-quality. In the video I cut to the footage from the DVR at some points to show how it looks from my goggles. Here’s a screen shot from that feed:
You can see the battery status in the top left corner. We spotted the moose at the end of my flight. At 13,2 volts the batteries will be harmed and the copter will descend automatically or simply crash.
Because the HD footage is done in the GoPro on board the copter it was pretty important for us to get it back in one piece. On this copter the GoPro is mounted “naked” without the water proof box. To save weight and make the mounting easier. So I don’t want to crash it in the snow.
The recording in the GoPro is 2,7 K @ 25 FPS. It is slightly stabilised in FCPX and rendered at 1920×1080.
It is a quad rotor helicopter. Based on a X468 frame that makes it possible to wrap it together for easy travelling. It is controlled by a DJI Naza flight controller that has gyros, accelerometer, barometer and GPS to assist the stabilisation of the copter.
I am using an old Futaba 9C remote with FrSky Tx and Rx modules.
After my previous post about the copter I have changed to three blade, nine inch props (9x5x3). They give me less jello and vibrations in my recordings. In general I find it easier to get rid of vibrations with smaller props running at higher speed. So I run the copter on 4S batteries. Usually two 2200 mAh in parallel. That gives me up to 12 minutes flight time with GoPro and FPV equipment. But 12 minutes is with completely new batteries and when I am pushing the limit. Usually I fly around 8 minutes and land when my batteries are close to 14V. In this video I push it all the way down to 13,2 V when I land.
I have also changed the ESCs on the copter. From Turnigy Plush that I mention in the article about the copter to HobbyKing F30A with SimonK-firmware. That change made the copter way more stable.
I am also flying this copter with a Sony RX100. It is heavier than the GoPro, giving slightly lower flight time. But the RX100 has excellent image stabilisation built in and you get rid of the GoPro fisheye-look. Here is a RX100-video from the same copter and same area.
Multirotors can be dangerous. The props are spinning fast. And they might cut your skin if you get too close when they are running. On my copter I have pretty soft and cheap GWS props. They’re not as dangerous as the carbon reinforced props that many people use. But if you are flying high the pure weight of the copter is dangerous if it falls down. And they do. That is why we are flying outside of the city over fields with no people. And always have spotters on the ground when we fly FPV.
The lipo batteries are very powerful. And can explode or catch fire if they are damaged.
In addition to this there are rules regarding remote controlled devices. They are different in different countries. In Norway you can fly up to 400 feet. And you need a special license to be allowed to put a camera on anything that flies.
If you want to to professional AP work you need an extra license from the civil aviation authority.
The fun and the problem
Lots of people immediately say “I want one” when they see how fun this is. And I can really recommend this hobby. But you need to be patient. First of all it is very difficult to fly these things. It has taken hours and hours of training to be able to fly the way I do in this video.
And even more hours fixing the broken copter after crashing. That is why I build them myself from the ground up. For the first couple of months I never returned with empty batteries. I always returned with a broken copter.
To start training I can really recommend this small and harmless copter: WLtoys V929. It is a toy, but it is four channel and is controlles just like the bigger copters. If you learn to fly this thing it will be easier to fly bigger and more dangerous copters.
This is one of the nicest multirotors I’ve built so far. Lifts my GoPro with ease. Flies very stable and powerful enough for acro. And I love the fact that you can fold it in seconds and fit it in a suitcase for travelling.
But to make it perfect you need to hack it slightly.
Props: Have been experimenting with different ones. Best experience with 10″ Graupner, GWS 9x5x3 and carbon 8×4.5 Slowflyers.
The frame in general is very nice, but the tail servo mechanism is not good enough. Fortunately you can fix that by adding two thrust bearings and a longer 3 mm screw.
The problem is that the original tail mechanism is nearly impossible to get really tight and still smooth to move. If you add two thrust bearings this problem is fixed. I used these.
Then you need a 3 mm screw that is longer than the original. Approx 30 mm plus head. The original is 25 mm plus head. I used a 40 mm screw that I cut down to 30 mm.
Now you simply assemble like shown in the images. Screw with thrust bearing, mount, original bearing around the screw, thrust bearing, another of the original bearings, motor mount and then a 3 mm locking nut. Tighten till you have it completely without slop but still smooth to move.
In addition to this I’ve mounted the motor plates the opposite way of what’s indicated in the manual. In a serious crash it broke one arm exactly where it’s weakest: where it has a slit to fit the folding mechanism. I hadn’t had the time to buy extra arms, so I fixed it by fitting a 12mm carbon tube covered with epoxy inside the arm that broke.
Just recieved my KK2 Flight controller from HobbyKing. I planned to put it in one of my existing quads, but figured I had four ESCs and four motors lying around. So I could just as well build a new quad for it. The only thing I needed was a frame.
So I found a bottle of Amarone, made a good meal, had a nice dinner and made a quad frame out of the box where the bottle of Amarone used to be:
Using 500 mAh and 1300 mAh 3S batteries. Gives 7 minutes of pretty agressive flight on a 1300 mAh pack.
I crashed the Amarone-frame so many times that I ran out of spare parts. I have now moved the parts to this carbon frame.
I also managed to destroy the KK2-controller. While waiting for a new one I’m experimenting with a baseflight-flashed FreeFlight controller and a Rabbit on this small quad. Both seems to work well. The FreeFlight is not available anymore, but with baseflight/MultiWii it is basically a very bulky Naze32 with less sensors. So, buy the Naze32 if you want a very good MultiWii based FC.
Conclusion about the KK2-controller
Very easy to set up. Love the fact that you can debug and adjust right there on the FC thanks to the LCD. Very stable rate mode but too sluggish auto level. I really hope Rolf Bakke have the time to implement the AHRS algorithm that utilize both the accelerometer and the gyro when in auto level mode. Right now the auto level mode is only good for slow hovering and maybe some help if you loose orientation.
But as you can see from the video in this article. That $30,- board makes my micro quad built from cheap parts steady as a table and loads of fun to fly!
And the Amarone-frame? Takes a lot of beating, but I managed to crash it bigtime.
Luckily I had some spare parts from the same wine bottle box so I managed to fix it and make it even better…
This is Eirik Solheim's prize winning experiment. I use this page to share information about media, marketing, technology, photography and stuff I find important. I have been running this page since 2003. And since I started using Statcounter in 2005 more than 2 million people have visited my site.
Eirikso on the web: Follow me on Twitter - My images on Flickr - Videos on YouTube