So there have been some sleepless nights recently as the deadline for the Outback Challenge second deliverable passed this afternoon. I managed to get my report in by the skin of my teeth after some email troubles (still waiting on the confirmation from the organisers :S ).
Each team had to submit a technical report that details the design of their aircraft and their risk management strategies. We also had to compile a video that demonstrates our on field setup procedure, takeoff and landing and how the aircraft handles carrying and dropping the payload.
I’ve compiled a playist of all the D2 videos I could find on YouTube. Of the 53 teams that passed the first milestone I could only find 12. Some teams may have been using private links, while others may not have used YouTube. However I can feel the field shrinking.
My video is included in the playlist above, but if you’re only interested in that one, here it is:
With the second deliverable for the Outback Challenge quickly approaching it was about time I discovered if the Pulsar could even carry the all important bottle of water.
After a successful maiden flight I took the Pulsar home and began hacking away at it with a rotary tool. The idea was to cut the holes I had planned in the fuselage so the bottle could be attached and the camera could see. This is pretty much what I achieved, albeit with a few more slips and scratches then I had hoped for.
With some minor cosmetic damage, but functionality otherwise perfect it was time to take it to the field. The conditions on Saturday were perfect with only a slight wind, no rain and some pretty spectacular looking clouds.
I took a short flight to test some changes I had made to the Pulsar’s controls as well as the epoxy I had just applied to fix a minor wound. The exponential setting on the elevator made a world of difference and the Pulsar is now a dream to fly. Once I was comfy, I landed and attached the water bottle.
The payload is not my proudest moment in terms of design. I just took a 500mL soft-drink bottle, stuck some fins on it and taped some bubble wrap to the front. It’s not pretty, and I don’t expect it to survive the drop. I’ll worry about that if I make it as far as locating Joe.
The bottle also has a plastic pylon strapped to it with some webbing which mates with the mechanism in the fuselage. This includes a servo with metal arms that locks the pylon in place until commanded to release it. Both parts were made from sintered plastic and my tests of the mechanism have been successful so far on the ground.
The Pulsar weighs about 2.3kg with all my gear installed, which is about what the designer had in mind. Adding another 500g also adds a number of concerns. Will it still be able to climb? or will it just crash into the runway on takeoff? If it does climb, will the wings stay attached or be ripped off the fuselage? And finally: if it flies with the bottle attached, what happens when it suddenly drops 18% of its total weight? There was really only one way to answer all these questions.
Launching the Pulsar with a bottle attached is awkward to say the least. Because it needs to be exactly below the center of gravity, it sits exactly where I want to hold it. The first launch was quite hairy as the Pulsar banked hard, directly towards my camera man. I quickly floored the throttle and was very happy to learn it had enough power to climb and clear my camera man by a significant margin.
It actually managed to climb quite well with the bottle attached. It couldn’t quite go vertical as it does without the weight but it still manages to climb quite quickly and steeply. It certainly didn’t fly as easily with all the extra weight, but it still managed to fly effectively. It could still glide and could get up a lot of speed.
With the Pulsar & bottle in the air and with the wings still attached, it was time for the moment of truth. I counted down to give my camera men warning and flicked the switch to drop the bottle. It separated cleanly and fell away without interfering and to my surprise the aircraft barely twitched.
After landing, the bottle was recovered and despite some damage to the fins it still contained all the precious water and had no cracks or leaks to speak of. We completed two drops of the bottle without incident and the bottle itself remains intact.
A very successful day indeed and given the deadline for the next deliverable was extended two weeks, I’m still on track to receive a ‘go’ from the organisers. One objective was also to gather enough footage to prove everything I said above. Thanks to Jan & Uwe I have a bunch which I’ve attempted to cut together into a short film for you all. Enjoy!
So it’s been 18 days short of a year since the Pulsar arrived in a giant box. Since then there has been a lot of time spent measuring and modelling and generally designing to figure out how to fill it with stuff to ready it for the outback challenge.
It took a few attempts at various components to get the fit right. Measuring its sleek, sexy curves proved to be quite difficult. The last few days have been spent shifting things around to perfect the balance and program my recently replaced radio equipment.
So as you can see in the image, the cabin of the Pulsar is pretty cramped. On the right is the stack of avionics including Asity and the controller for the camera. In the middle is the servo that holds and releases the water bottle (under all the cables). Finally on the top left is the receiver used for manual control. Asity is now capable of controlling the aircraft, so once I’ve tested it thoroughly, that receiver will be replaced with a slave receiver and Asity will take over.
Today was the latest in a few attempts, the first since January, to get the Pulsar off the ground. I warmed up a bit by taking a the Paprika for a spin, as I haven’t actually flown anything other than a simulator for about twelve months. The wind was a bit on the strong side but wasn’t gusting too badly, so we decided to go ahead with launching the Pulsar.
The launch was a bit hairy as always but once I pushed the throttle it had no trouble climbing almost vertically. Once I had some altitude I could take a breath and find a comfortable attitude. It doesn’t handle anything like the Paprika as you would expect, so I can’t apply the throttle and do a quick backflip to escape a low altitude stall. It didn’t take much trim to get it flying level but I did leave the elevator controls linear rather than the softer exponential option. Given the elevator is so huge this would have made it much more controllable. So I spent the first flight bobbing up and down with over corrections but generally flying smoothly.
One major issue I had not considered was that the wings look identical from top to bottom. This often left me guessing which way up the aircraft was, leading to a terrifying accidental barrel roll. Given its wingspan it doesn’t flip over very quickly so my usual ‘trial and error’ approach to situational awareness doesn’t leave me with much time to make any mistakes. I’ll probably be sticking some fluro-yellow tape to the underside.
The landing was a typical repetitive stall as I continued to overcorrect with the elevator. Under full flaps, this aircraft can drift incredibly slowly which made the landing look like a slow motion crash. Luckily it actually was slow so it touched down with only a slight bump. Once I’ve sorted out the controls it should be a very nice aircraft to fly.
The Pulsar wasn’t the only plane to take it’s first flight today. My supervisor has recently built a Pilatus Porter which has also been waiting patiently to get off the ground. His son, Jan took the controls and it too had a successful maiden voyage.
The day almost went without incident. I took the Paprika for another run after the Pulsar was done. At some point during the flight, one of it’s stabilisers cracked, and was hanging limply by some balsa threads. Miraculously, or just because of my mad skillz it made it to the ground in one piece. If it didn’t have a v-tail, I don’t think the landing would have been nearly as successful.
So the Pulsar flies! My work now is to mutilate the shiny white fuselage by cutting holes for the camera and payload which I can’t say I’m looking forward to. Hopefully I can get it back up to test the bottle drop in time for my April deadline.
As I’ve mentioned a bunch of times on this blog and to anyone I came into contact with over the last 12 months: I’m building a rather large glider to compete in the Outback Challenge. I recently finished collecting and assembling parts and am keen to see it get off the ground when all the stars align. You’ll notice that it’s big… really big. And no; I’m not short
The Pulsar range of gliders should be familiar to most RC enthusiasts. They are typically a slow, light weight and very well built aircraft featuring enough carbon fiber to make most cyclist jealous. It is not a scale model, so it doesn’t look like a traditional manned glider. Instead it has been designed from scratch to be a very efficient radio controlled aircraft.
The Pulsar 4E is the largest of the Pulsar series with a wingspan of 4 meters. The one I have assembled weighs 2280 grams without a payload, which is expected to be a little over 500 grams of water and bottle. I have fitted a modest 550 Watt Neu motor to the front which I expect will be just enough power to maintain cruise speed while climbing vertically.
So far, we’ve taken the Pulsar to the airfield twice in attempts to fly it. Unfortunately we’ve been foiled by gusty winds and some really buggy firmware for my transmitter. At least it hasn’t been destroyed yet.
I’ll save my rant about Spektrum’s latest DX8 firmware (2.04) being a terrible example of embedded software and something I consider quite dangerous, as it is sometimes in control of very large and fast aircraft. I’ve downgraded to version 2.01 which I will use cautiously for the time being.
I still have to add the avionics I’ve been working on. I had some plastic mounts printed that served as both the dropping mechanism for the payload as well as holding the electronics in place. Unfortunately some of the dimensions were a bit off and it failed at both of those things. I have some more coming in a couple of weeks and will hopefully have written enough code to make that worthwhile by then.
So while I wasn’t flying anything today, I took the time to make this view of airfield (3MB). Enjoy!
I recently completed building the first of my new version of Asity, the avionics I plan to use in all my UAV activities. The main improvements in this version are the new inertial sensor chip I’m using. I previously had a separate accelerometer and gyroscope. Now I’m using the MPU-6000, which does both much better than what I had. It also does some motion processing of its own which I am dubious about, but I’m going to try it out and see if it can save me writing my own.
The top side of Asity
There are also a number a bug fixes I noticed while playing around with the last version. These are silly things like not including a bias inductor on a powered GPS antenna, supplying the wrong voltage to my compass, and forgetting to put decent pull-up resistors on an I2C bus or two. Fingers crossed that there aren’t more I didn’t find.
The most significant improvements are from a manufacturing view. Given that I’m assembling this by hand with a frying pan, I need to put some effort into the design to make it as easy as possible. This involved swapping a bunch of 0402 sized capacitors & resistors for their slightly larger 0603 versions. I also had to consider how I can fix things as I assemble it, so trying not to block access to my soldering iron. I also have a planned out assembly procedure that involves testing at a few stages. It’s much easier to fix some things before the board is complete, so I try to make sure it all works before continuing. Applying solder paste tends to be the messiest part and results in lots of tiny conductive balls rolling around your components if you do make a mess of it. I ordered some Kapton stencil from OHARARP to make this easier.
The bottom side of Asity
The bottom side of Asity is all hand soldered, so I made sure to choose the components and their layout wisely. This means: no 0402 sized components; no packages with hidden leads (QFN, BGA), and; everything nicely spaced so I can get my soldering iron between things. This last point may have been neglected a little, mostly because there isn’t enough space on the board.
I took two attempts to solder the FPGA on the bottom. The first time was a bit of a nightmare and I spent way too long poking at it with the iron to clear shorts etc. I decided that I had probably butchered the chip and that I didn’t really want it flying my aircraft. So I removed the first one completely, cleaned the pads and started again with a fresh one. It wasn’t a cheap part, but given the components on the others side aren’t particularly cheep either, and they were already working, I decided it was for the best. The second FPGA went on with much less fuss and also seems to work fine now.
The daughter board
I’ve also assembled the daughter board that fits on top of Asity onboard the aircraft. This board includes a VHF radio to act as a backup link if the main UHF fails. It also has an interface to the camera in the nose of the aircraft as well as another FPGA and some RAM to process images data.
So far everything has tested well. I haven’t tried the radios yet, as I still need to assemble the other end of the link, but I’m feeling pretty confident about everything else. It all fits together nicely (there’s a massive radio module sandwiched between the two halves) and only weighs 53g assembled. Now I just have to fill it with useful code.
So the massive lack of updates recently has been due to me gaining knowledge in the business management and marketing fields. I’ve been doing some more coursework to complement my PhD and make me generally more entrepreneurial. A fairly painful experience so far, and one which will last for another six months.
Between paragraphs of my business plan I did manage to update the designs for Asity as well as complete the CAD models for the parts I need for the Pulsar. This combination is geared towards completing the Outback Challenge while still being useful to my research. So here’s an attempt to illustrate what I’m building.
This is the layout of the components I intend to include in the Pulsar4E airframe I’ve got lying around. It include the typical motor-ESC-battery you’d find in a modern electric RC plane as well as the avionics and camera components I’ve been working on. You’ll also notice the whopping great water bottle hanging from the bottom. This is of course the all important payload that needs to be delivered to Joe.
The motor, Electronic Speed Controller (ESC) and propeller blades are being carefully selected to allow the whole aircraft to climb vertically with a payload when the throttle is flat out. Once it reaches a comfy altitude, the motor is switched off and the propellor folds back. I’m using a Neu 1110/1Y with 6.7:1 gearbox with a 75A ICE Lite ESC from Castle Creations. I’ve picked a few props with varying pitches to experiment with. I’m expecting the aircraft’s takeoff weight to be around 3 kg, and so aiming for a static thrust slightly higher than that.
The moulded parts shown are all coming from Shapeways. Results so far are promising. The quality is great and my measurements of the curvy fuselage are actually quite accurate. Not so accurate are the dimensions I chose for the servo components in the payload dropping mechanism, so I will need a few bit reprinted.
The payload is suspended from a pylon by two polypropylene webbing straps. I will also add some velcro between the pylon and bottle to prevent it from sliding back and forth. The pylon fits a hole in the bottom of the fuselage and mates with a beefy servo with a metal horn.
The camera is towards the front of the aircraft and also needs a hole cut in the fuselage. The camera I’ll be using is a fairly cheep 1.3MP CMOS sensor; the kind you would’ve found in a phone from 2005. It should be just enough to spot a vehicle from 1500ft, and hopefully Joe from 400ft. I believe the aircraft will be able to stabilise the camera adequately, so I have only included one servo to compensate for the pitch of aircraft. As I’ll be gliding most of this time, I’ll usually be pitched down and will need to adjust the camera.
I designed a small board to transfer data from the camera back to the main processors in the center of the aircraft. The board pictured includes a small LVDS transceiver that will translate the parallel interface of the camera into a high speed serial stream, back up the Cat5 cable to a similar chip which will recreate the parallel interface and feed it into an FPGA. From there it will be appropriately processed and fed into the UHF radio to be sent to the ground.
I’ve also built the first of my latest version of Asity; the avionics I’m developing. I’ll go in to more detail about this in another post, so for now here is a pretty picture:
The couriers have visited frequently in the past few weeks, two of them arriving today. The first brought me the new ADCs I need for the Asity prototype, which I should hopefully have complete sometime tomorrow. The second courier brought a rather large box containing the Pulsar 4E that I plan to use for the Outback Challenge next year. This is a big aircraft! compared with the Paprika I built last year, this has twice the wingspan. It’s 4 meters wide and so far weighs about 1100 grams. I expect it to weigh a little under 3 Kg once I’ve installed all the other components and the payload. I’m still working on sponsorship for the remaining parts so it probably won’t be airborne for a month. Right now I’m going nuts with my callipers to try and build a model of the fuselage so I can plan out the other parts I need. This is one disadvantage of a sleek glider; there isn’t much cabin space to play with.
The Outback Challenge was recently announced as a bi-annual event with the next ‘Search & Rescue’ being held in September 2012. While this does interfere with my plans significantly, I do appreciate the extra 12 months for development.
I may be using this title a little prematurely… I just finished watching Caprica last night. The last few months of my life (holidays included…) have been spent designing the hardware components for the avionics board I plan to use for my unmanned aircraft. A few days before Christmas the PCB designs were sent off to the manufacturer signifying the design was complete and ready to be built. I’ve since spent my time rereading datasheets and conferring with colleagues, praying that there are no significant flaws in the design. So what is it?Asity is basically a processor, a radio and a collection of sensors that cover everything I need to fly an aircraft autonomously and perform some experiments relevant to my research. The PCB is 40mm x 60mm with the odd protrusion and will easily with into the fuselage of a glider. More specifically:
Actel ProASIC3 FPGA
32 Mb Flash
MicroSD Card Slot
3 Axis Accelerometer
3 Axis Gyroscope
3 Axis Compass
Barometric Pressure Sensor for altitude
Differential Pressure Sensor for airspeed
8 Servo Channels
Current Sensing and power enable on each Servo
Flight Pack Voltage Sensing
Motor Current Sensing
Slave Receiver Headers
12 General IOs
I’m also building an USB daughter board to allow a ground variant to attach to a laptop in the field.Why does the world need another autopilot? There are a number of similar boards available covering a range of applications. The main difference with Asity is that the processor is a Field Programmable Gate Array (FPGA) rather than a traditional CPU. I also intend to write the firmware completely in HDL rather than using any ‘soft core’ processor that seem to be so popular with FPGA developers. This allows for verifiable firmware that can run mostly in parallel. Each sub-system effectively has its own hardware within the processor greatly increasing reliability and timing capabilities.
The challenge is now to gather the rest of the parts and the tools required to assemble the boards, as well as the months of VHDL coding I have to do. Only time will tell if this project will actually work .
Having recovered from its tragic accident last week the Paprika was once again ready to fly today. I’m actually still waiting for some parts from the US, so I’ve put a new plastic Graupner prop on it, which is a bit squishier and doesn’t look nearly as slick as the carbon HK I’ve been using. Otherwise it’s the same pitch and diameter. I was also fairly adamant about getting video from onboard the aircraft. The last failure was a dodgey switch so I’ve opened it up to fix it. While I was there, I modded it a little so it can actually point along the fuselage. So the camera finally worked and the result is what I would like to share with you now. It starts a little bumpy while I’m working out the trim, but I eventually manage some low and speedy flybys. I don’t wanna hear anything about my landing! The flaps shed speed a lot faster than I was expecting and I decided that was safer than trying to pull out again.
My recent expedition to the airfield ended abruptly on Thursday with a unfortunate nose-dive on takeoff. I managed to hand launch the Paprika on my own for one flight. After a quick trim, I managed a whole seven minutes of smooth flying and a neat landing with full flaps. The flight only used about 600mAh from my 4S 2600mAh battery, so I could easily stay in the air for over half an hour. The only reason I landed was because the ESC data log was full and I wanted a good snapshot of my flight.
The video camera failed me again. I thought I had just left it on and flattened the battery but after closer inspection I found the power switch had failed. So still no video from the aircraft… The second flight was less successful. Immediately after launch it pitched down and hit the runway resulting in two snapped prop blades and cracked stabilisers. This was probably caused by shifting the centre of gravity forward without removing the tail trim to compensate. The worst part was that this happened immediately after my supervisor arrived. This created some scepticism around the story of my first flight, which was not helped by the lack of video (again).
So another night was spent at the workbench putting things back together. I’m a little relieved that I had the chance to improve the strength in the stabiliser joints. There was not much force involved in the crash, so they could just as easily have snapped off in the air during a hard pull-up. I’ve also ordered new blades and a handful of other things from Esprit Model. They get a special mention as they shipped my parts within 4 hours of me ordering them and their prices are always pretty fantastic. So the repairs are done, I’m just waiting on the new prop and should be back in the air sometime next week, hopefully with a working camera.