The MinimOSD can be powered in two ways. The “normal” way is to give the telemetry side +5V from the APM and the video side -12V directly from the battery. I’m not sure what I did wrong, but like many others, I managed to kill my OSD using this approach.
A safer approach, is to solder two pads on the board, connecting the power rails on the two sides, and then only powering it from a “safe” +5V source at the telemetry side. I have followed this approach and have not had any problems since then.
In order to get the telemetry from the APM to the OSD and connect the telemetry radio you need to connect both the OSD and the radio to the APM’s telemetry TX.
Below is a diagram of my current setup as it works for me.
Two important notes!
- I take NO responsibility if this does not work for you and nukes your components!
- I’m NOT powering the OSD and radio from the APM’s 5V rail. There are known issues with this approach with certain APM versions.
The graphic of the OSD, I took from this web page : https://code.google.com/p/arducam-osd/wiki/minimosd
Took the Bixler out for it’s first flight with the ArduPilot on-board and had few very good flights. I had a bit of issues initialising the airspeed sensor correctly. One init, it showed around 20m/s while on the ground (I don’t think the Bixler can go that fast, ever if it tried to!) and on another occasion it showed zero airspeed throughout the flight.
This is something I’ll have to watch carefully as it does not navigate automatically if this sensor is not working!
However, once that was working it navigated very well. Airspeed and XY position was very good with it hitting each waypoint that I set for it. Airspeed was kept within around +- 1m/s. I’m not totally happy with it’s height accuracy, it lost around 10m on the downwind turn every time. The new speed and height algorithm they use now is new to me and I need to see if I can tune that a bit better. Thought 10m is not THAT bad, considering that it was a very tight waypoint pattern and that it will usually not turn that much during navigation.
See the screenshots below for the recorded telemetry tracks.
I added FPV gear to my Bixler for no-nonsense FPV and had really good results. So good, in fact, that my problem now is that I have great video at ranges that I’m not comfortable flying at without RTH and/or an OSD.
So, all my resolutions of keeping the Bixler no-nonsense have flown out the window and the Bixler is getting modded!
I’m adding an APM2.5 with Minim OSD. While I’m busy, I have switched from the regular Turnigy receiver to an FR-Sky telemetry receiver that is now moving into the tail.
The Bixler’s wing attachment mechanism have always irritated me, so that is changing too. The wing is now attached from the top using big nylon screws.
When I’m done, it should have full autopilot with OSD for FPV, able to carry 2 x 2200mAh batteries in the nose (as all the avionics are now under the wing and in the tail, making more space).
However, I’ll probably have to look at the motor and prop setup before I’ll be comfortable flying with two batteries.
Here are some photos of the build so far.
Radio RX is moving to the tail:
Wing now attach from the top:
Belly cut open and now usable for avionics:
I have been fortunate enough to be part of the ESRI South-Africa Summit 2013 held at the Champagne Sports Resort. I had the opportunity to fly the ESRI UAV there and see it as the highlight of all my efforts in this blog. I could not only do what I love, but also had the opportunity to finally test the results against proven equipment.
I could do a few flights photographing both the front lawn and a part of the golf course. I was then assisted by Ian from Leica to capture a set of control points on the lawn using their excellent GPS hardware. The accuracy of the Leica GPS was in the region of 4 to 8 mm when making use of differential GPS! We also captured a whole host of cross-check points to test the final results against. Without Ian’s help, I would not have been able to get such a good quality set of points to test against.
I then built an ortho photo and a digital elevation model using a trial of Agisoft’s excellent Photoscan software. I imported the GPS measured control points and cross-check points into ESRI’s ArcMap and measured the differences between my Digital Elevation Model and the measured points. I think the trial has made it’s point and I hope to acquire a licence of this software in the very near future.
I was stunned by the results! Height accuracy of the model was between 25mm and 95mm in the control point area. Accuracy was between 40mm and 400mm in the area outside the control points. The worst accuracy was at the edges of the photographed area – as expected.
These areas would typically not be used as there is a lot of distortion there due to that area being made up mostly of oblique photos. Other large errors was due to the stereo process measuring the height of vegetation and not of the ground under the vegetation. This is a limitation of this type of height model generation that is not present using traditional methods and advanced methods such as LAS mapping.
Even though the process has limitations, it should be considered against the cheap cost of the platform compared to traditional hardware. It should also be noted that the results were obtained using low quality runs on the model to keep the processing time low, enabling us to get results in less than a day. Also of note, is the fact that the plane was flown manually and without the assistance of the auto pilot as I am still awaiting the replacement auto pilot after the servo failure and crash I had last year.
The week end yielded a total of 9GB of video and photos taken from the UAV. All in all a very fun and productive week.
Here is a compilation of some of the video taken from the plane:
Throughout the process of building the UAV I had been experimenting with SPAD (Simple Plastic Aircraft Design) planes. Since they are simple and quick to build and cheap material wise, one can build a plane to try out a concept and discard it if it does not work.
I started off with a fairly traditional layout sans the V-tail which worked quite well and flew quite well. The biggest flaws were the too short tail moment which could bite you if you got the GC only a little bit wrong and the too small size to carry any usable size stills camera.
The second design was better. It had a pusher prop, leaving the nose free for camera equipment. Also a V-Tail design, that paid off at least once when I had a servo come loose, while still maintaining enough control for a safe landing. Since the two panels of a V-Tail doubles as both elevator and rudder, losing one, means you only lose 50% of your elevator and rudder control. It’s thus naturally redundant. Due to the pusher design, it had to have a fairly deep section body so that the prop can clear the tail boom. Much the same as the current plane I use for camera work. It had tricycle undercarriage, which made it a breeze to take off on a good surface.
The disadvantages of this plane is that it’s still too small for a stills camera, the wheels are too small for rough field take off and landing. Also, like the current UAV plane, the deep section body doesn’t like crosswinds, and give funny pendulum effects while flying, making it overall not a very nice plane to fly.
My latest SPAD is again a much more traditional plane. Even though I don’t carry the camera in it, it will carry the camera easily. Of all the planes I have, this is the sweetest one to fly. It handles well and is always a pleasure to fly. It has one small flaw that led to the name I gave it : The Lawnmower. It’s a fairly long plane which means that it stands on it’s wheels in a near to level attitude. The plane is a tail dragger which means I cannot increase the angle of attack on the main wing on take off by rotating when I have sufficient speed. This all means that I need quite a high take off speed to get into the air, even though it can fly quite slow once in the air. So, when doing take-offs on grass, it tends to look like a rogue lawnmower as it tries to gain enough speed whilst chopping and throwing up grass with the fairly low to the ground propellor.
I also found that the large tail fin is not nearly stiff enough for high speed flying as I got a scary amount of flutter on the one flyby I did. As luck would have it, I had the keychain camera pointing right at the tail when it happened! (See the video below).
All in all, it’s still an enjoyable plane. It glides surprisingly well and seems to have fairly low drag compared to the UAV I use for camera work. This is quite surprising if you take into account how badly it’s finished aerodynamics wise!
I am already formulating a layout based on an American spy drone as the next plane I will build. I will keep this blog updated once I start with that.
After a hiatus of nearly four months, I’m in the air and taking photos again. I’m happy with the performance of the rebuilt plane and took it out for a test run the week-end. I am still waiting for the auto-pilot, so I flew it manually.
It’s flying nice and stable and the camera mount and remote trigger works well. I took a few hundred photos of the area where I re-maidened it and built up a composite photo from the photos. Results are a bit rough as a human cannot hope to fly as consistent as an autopilot, but I’m happy that the individual photos look good enough. Which means that the new nose is at the very least not causing more vibration to reach the camera than the old nose did.
The trigger also retracted the lens back to safety both flights, so I am much happier about the safety of the camera than I was with the old setup.
In the mean time, I have acquired a Bixler 2 plane that I plan to use for casual FPV flying and as a trainer for the kids.
They are really sweet planes and very popular as a cheap and no fuss FPV platform. I will maiden it tomorrow and plan to have it set up and ready for the summit as an extra example of cheaply getting a video platform in the air. It would not be able to carry the still camera, but it will easily carry the keychain video camera and even a GoPro sized video camera.
I will report back here once I have flown it.
I have finished the shelf on which the camera will rest and cut the hole in the bottom through which the lens will look. I have also managed to get a nasty glue-gun blister on my thumb for my troubles.
After having endless problems with wheels ripping off in thick grass, I decided to leave the wheels off and make the plane a belly-lander. I did however add a single centre line wheel if I have to land on hard surfaces. It’s partially embedded in the body so it should not snag on anything.
One concern with the wheel is that it sits right in front of the camera’s lens hole and can potentially throw dirt right on to the camera lens. I have however built a remote control for the camera using an old servo board. With that I can extend and retract the camera’s lens by connecting the camera’s USB port to a channel on the receiver that is assigned to a switch on my remote control. That way, the camera lens will be safely retracted with it’s lens cover over during take off and landing. Later, the channel can also be controlled by the autopilot if need be. I’ll post details on the camera remote-control later.
I still need to tidy up the servo wires in the non-removable part of the aircraft, but I will always have two loose ones as they have to detach when the wings are removed. I will also have to add a shelf for the autopilot once it arrives. The autopilot will probably not be in time for the summit, so “Mark I Eyeball” flying will have to do for now!
I have some plans for reducing the weight of the plane too by cutting holes in the wings and the body and covering those with film or Vinyl, but I’m still undecided on this.