** Long post alert **
Rob at FlyingWings and I have long talked of having a go at building an airframe to make good use of the PX4 VTOL branch. Rather than the boring multi-rotor / flying wing fusion seen in many VTOL deployments, this project offers a configuration that uses a dual tractor prop design with a combination of the two contra-rotating props and intricate elevon control to maintain positional stability when in VTOL mode and at the flick of a switch you can transition into level flight, the controls adapt between multi-rotor and flying wing based on the current mode. This offers up the possibility of a VTOL capable craft with long range capabilities; the field applications of this are endless, and instead of going for the low hanging fruit like military contracts or conservation projects we are going straight for the big prize of YouTube humiliation.
From what I can see the only implementation so far is on a TBS Caipirinha, which looks a bit cobbled together and chosen mainly for it's lightness and durability (nothing like having faith in your project...). We've decided to go for it in style and modify the FlyingWings Venturi Evo for a twin prop tractor configuration. The benefit of this airframe is it's relative lightness for its size whilst still having a good amount of space in dedicated bays for flight controllers and other equipment, making a nice neat build which should please Coyote. Rob has done a stellar job and provided me with a Venturi Evo and custom nacelles to take the twin props, there's a lot more to it than that of course as we've had to redesign the spar layout to accommodate the swapping around of the equipment and battery bays, necessary to keep the CG in the right place after swapping the power systems from back to front. There will also be custom winglets which double as a vertical stand when taking off and "landing".
It's always good to work to a brief so we've set ourselves the challenge of a VTOL take-off, 10 KM out and back run followed by a VTOL landing. If we can manage that then we have an interesting capability in a flexible airframe, if we fail then we will undoubtedly have some YouTube gold.
The code itself runs on the PX4 platform, which the Pixhawk is based on and therefore this is the hardware we'll be using to host the code as all the necessary gubbins are integrated into a lovely neat package. The iPeel Technical Centre of Excellence has taken delivery of all the necessary kit to build the project and I'll be updating this thread with the ongoing project. I'm expecting this one to last a bit longer than my usual builds as well as probably making up a few new words as I attempt to get the code running.
So far the parts list is running at:
- Customised Venturi Evo with bespoke nacelles and spar.
- Custom elevons (more details as the project runs on).
- Pixhawk controller + relevant bits running the master branch of PX4 Autopilot software (not ArduPilot).
- MyFlyDream Telefly Pro for tracking.
- MinimOSD running OSDExtra.
- 2x Turnigy 3542/5 1250Kv motors.
- 2x Plush 60a ESCs.
- 2x Hitec 645MG servos.
- 2x APC 9x6 E props (one CW, one CCW).
- Effio-V board cam.
- 1.3Ghz 800mw VTX.
- SJ4000 camera.
- DTF-UHF OpenLRS RX running OpenLRSng.
Battery-wise I'm undecided as to whether we'll use standard 3300Mah 4S nano-techs or use the lighter weight, higher capacity Panasonic NCR18650e cells in our own configuration, the C rating of these cells may not be enough for VTOL though. For the VTOL capability the Plush ESCs will need to be flashed with BLHeli which I've already checked can be done and be responsive enough. The motors weren't my first choice which were SK3 1250KVs but these aren't in stock and the regular Turnigy jobs should have enough grunt I'm hoping. I've also got an extra cam and video switcher on standby to allow a bottom-view camera which can be used as a ground level view when we're in cruise and a forward view when we're in hover.
I'm hoping to make this a nice, interesting project to follow, and will get as many pictures up of the build process as possible along with the configuration steps as very little online documentation exists on this project. Our entry into the VTOL sector has already seen Google run for the hills (http://diydrones.com/profiles/blogs/google-scraps-vtol-design-back-to-the-drawing-board), Amazon watch this space!
More details on the PX4 VTOL project can be found here (https://pixhawk.org/platforms/vtol/start) but what you really want is a nice video showing it off, so here you go:
http://youtu.be/acG0aTuf3f8 (http://youtu.be/acG0aTuf3f8)
Cheers,
Neil.
This is an excellent project. From what you said earlier, I assumed a quad or tri plane into a dual proper normal setup.
The demo looks pretty impressive..albeit possibly the worst camera work ever from that quad. The bi-copter style hover looks very impressive, but tricky in the practicality stakes if you are using FPV to take-off/land.. On the plus side if you can flick a switch an lknow it will come into hover happily by itself, then a very easy goggles off and LOS land.
Looking forward to watching this unfold :)
Thanks Wayne. :)
We've been thinking about the FPV take-off and landing dilemma which is why I have the second camera and video switcher on standby. I'm not concerning myself with that for phase one though as the thought of not maintaining control line of sight with a 1.5 metre 2.5Kg twin prop beast running pre-alpha software at head height isn't appealing.
Neil.
PX4 ... Good Choice :D
Loved the transition to forward flight in the vid..
Looks like a great project !
regards
Andy
Quote from: iPeel on March 25, 2015, 02:39:17 PM
The iPeel Technical Centre of Excellence
I just cant work out an acronym...TIT CE
I look forward to helping you gather the foam from across the field on test day ;D
I think just a normal twin tractor version of the Venturi could be awesome, more efficient and less noise.......... man I'm sounding old! Did I tell you I've just bought a Mondeo! :'(
Sorry Brucey, I would have replied earlier only I was out having an amazing flight while you were stuck working! :P I would have taken you up on buzzing the tower only I'm pretty sure I wouldn't have made it back...
Yes, it'll be interesting to see how the airframe flies in that configuration, shame I'm not putting twin SK3 motors on it as that would have ripped along, as soon as they're back in stock I'll probably find myself ordering a pair.
You're not even a dad yet and you're having your wings clipped mate, I hope you're keeping the beast for old times sake?
Yeah the weather was great... As I looked out of the window... Sob...
Looking forward to seeing this materialize :)
Sounds like an interesting project! The video looked good (well, the actual video was terrible as mentioned, why put a camera with an arm and motor in view?!?) and the transition to forward flight was nice and smooth looking!
Will be great to see it on a big bird like the Evo!
How about a camera mounted on the top of the goggles that you switch to as you transition to hover to give immediate line of site? Would save having to whip the goggles off at a critical moment. Can the pix hawk do auto landings like the APM?
Sent from my iPad using Tapatalk
Quote from: BigT on March 26, 2015, 09:43:08 AM
How about a camera mounted on the top of the goggles that you switch to as you transition to hover to give immediate line of site? Would save having to whip the goggles off at a critical moment. Can the pix hawk do auto landings like the APM?
You mean like the Skyzone goggles? Neat idea and I look forward to getting as far as having that problem to solve! The Pixhawk can actually run the APM:Plane firmware or it can run the PX4 firmware, when running the APM:Plane firmware you get all the features of the original APM ++, as the board is much faster. I'm not sure what the true capabilities of the PX4 firmware are as it's very different to APM:Plane and will report back as soon as I've had a dabble in the configuration screens.
So, a quick update as all the bits and bobs are here:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0001.png)
The Pixhawk is really nicely packaged and the unit itself looks like it's really well built. Size-wise it looks much smaller than it should be but is actually a little bit bigger than an MFD:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0005.png)
There'll be plenty of room in the equipment bay for it and the GPS et al.
The wing has been pre-cut on the wire cutter for the nacelles:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0012.png)
The nacelles themselves look really nice and are reinforced at the sides to take any of my trademark landings:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0013.png)
What I think I will struggle with is the order of the build. Normally I would fit the electronics then laminate over it. I think I'm going to have to laminate the wings then cut away the laminate for the nacelles then glue them in place, quite when I fit in the ESC lines for the motors I don't know.
Another thing I'm stuck considering is when to build the centre section:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0017.png)
I want to leave the ribs which run across the centre section from side to side until last, because only then will I really know how to balance up the batteries and that will determine where I glue the ribs in. I will likely build most of the centre section without the ribs then hack those in last.
Urgh! I'm beginning to realise how big this build is going to be even before I get round to the really cool stuff!
Neil.
Quote from: iPeel on March 26, 2015, 08:03:45 PM
Quote from: BigT on March 26, 2015, 09:43:08 AM
How about a camera mounted on the top of the goggles that you switch to as you transition to hover to give immediate line of site? Would save having to whip the goggles off at a critical moment. Can the pix hawk do auto landings like the APM?
You mean like the Skyzone goggles? Neat idea and I look forward to getting as far as having that problem to solve! The Pixhawk can actually run the APM:Plane firmware or it can run the PX4 firmware, when running the APM:Plane firmware you get all the features of the original APM ++, as the board is much faster. I'm not sure what the true capabilities of the PX4 firmware are as it's very different to APM:Plane and will report back as soon as I've had a dabble in the configuration screens.
Surely you want a camera on the plane that tilts as the plane goes to-from VTOL. That should be pretty trivial to do with PX4 or APM software. Just a question of getting current pitch and subtracting that from your desired camera angle and then applying to ppm output to a servo.
regards
Andy
How big of a prop can you fit on there? Can you sling it in the car and bring it to MD, I'd like to have a look!
I thought about the option of tilting the camera Andy but certainly for the first few runs I'll be LOS only and will probably want to dabble with something elaborate after that.
Bruno, I might pack a few bits of it but doubt I'll have any time tonight to progress any more of the build to have anything like a built up airframe by tomorrow, too many batteries to charge!
I didnt expect you to have it built (lazy or what!), just want to see it in the flesh. I really like the idea of a twin tractor
Wow this sounds like a great project will be keeping an eye on it
Right, some progress eventually made today!
First as this is going to have to do multi-rotor-esque things with the props the 60a ESCs needed to be flashed with something multi-rotor-ish, I used BlHeli as I had planned to do so when I selected the ESC. Flashing BlHeli is pretty easy if you have an Arduino lying around and the BlHeli desktop software has everything you need, except a little lead which marries up with the three pads on the ESC. I had one already for my Plush 12's on my 250 and this was the right pinout for these.
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0004.png)
Next I got the OpenLRS flashed, and as I need analogue RSSI for the OSD there's a little solder pad on the DTF-UHF's which turns the RSSI into analogue without the need for an external buffer:
(http://www.peely.com/wp-content/uploads/2015/03/IMG_0009.png)
Next up, a lick of paint on the centre section and the nacelles, leaving the equipment bay naked so the sticky stuff still sticks. I fitted the motors to the nacelles and then got about channelling all the equipment bays out including also the ESCs, these will still be out on the wing next to the nacelles but as far back as I can bearing in mind all the controllers will be up front. I've also positioned the servos as far back as I can without hitting the spar which I'd asked to be farther back than the standard Evo to make room for batteries in the rear bay:
(http://www.peely.com/wp-content/uploads/2015/04/IMG_0962-e1428167843237.jpg)
You also see the pitot tube positioned on the front in the camera bay, hopefully out of scene. The motors are wired to the ESCs with bullet connectors which can be plugged and unplugged, this is so I can now laminate the wing halves without the nacelles then glue the nacelles in afterwards which should make the laminating process easier, it also means I can upgrade to SK3 motors when they come into stock for more oomph. ;D
The UHF antenna is just waiting for the plasti-dip to go off and then I can wire in the RX, the VTX is already wired in and once they're all complete I can laminate and the physical build should be almost complete!
Neil.
That looks awesome!
Agreed, looks really nice! Good progress so far!
Looks amazing.
A silly question perhaps- but what's the practical use and benefit of a vtol wing?
Is it purely the vertical take off/landing meaning a wing can be launched and landed in a very small area?
Or are you treating this as more a fun exercise and a challenge with a very unique product at the end of it?
Theoretically it offers a more reliable take-off and land capability than a good chuck and a gentle approach, it's also simpler to cater for VTOL take-offs and landings in automated missions. As you say it also gives the ability to operate in areas which are more constrained.
Rob and I were interested in the project some time ago and felt the project deserved more than a butchered Caipirinha! If it all goes well we will have a capable airframe that suits developers but initially we're just having fun testing the limits of what can be done with the Pixhawk.
There's not much update-wise because the airframe is pretty much ready and I'm getting to grips with the Pix and QGroundControl, aside of the fact I'm using the ongoing development branch of the firmware, there's a real lack of documentation which is compounded by the fact most people using the Pix are using it with APM:Plane or APM:Copter which clouds the Google results somewhat. Like a lot of Open Source projects the code is very under-commented too.
Hopefully in a week or so I'll have it all under control.
Quote from: iPeel on April 17, 2015, 11:51:55 AM
Theoretically it offers a more reliable take-off and land capability than a good chuck and a gentle approach, it's also simpler to cater for VTOL take-offs and landings in automated missions. As you say it also gives the ability to operate in areas which are more constrained.
Rob and I were interested in the project some time ago and felt the project deserved more than a butchered Caipirinha! If it all goes well we will have a capable airframe that suits developers but initially we're just having fun testing the limits of what can be done with the Pixhawk.
There's not much update-wise because the airframe is pretty much ready and I'm getting to grips with the Pix and QGroundControl, aside of the fact I'm using the ongoing development branch of the firmware, there's a real lack of documentation which is compounded by the fact most people using the Pix are using it with APM:Plane or APM:Copter which clouds the Google results somewhat. Like a lot of Open Source projects the code is very under-commented too.
Hopefully in a week or so I'll have it all under control.
Fair point about the documentation of Open Source projects. I guess as well as the obvious that writing docs is boring, writing docsit also means less time fixing bugs and writing cool new features. I think PixHawk is better than many and I am sure as it gets older and more stable that docs will get better too..
Anyway Looking forward to seeing this beast go :)
regards
Andy
Interesting. Really looking forward to seeing how it turns out.
Here's another one... ( On kickstarter)
Introducing the Vertex Hybrid UAV (http://www.youtube.com/watch?v=vNFmoqj1wVA#ws)
How is yours going Neil? Any updates
regards
Andy
I finally managed to break away from other stuff and put some time into the VTOL project today. I've been chatting a to a few people on the PX4 forums about my issues and nobody seemed to have much to say about it. I think the amount of people active with VTOL projects are rather thin on the ground.
The problem I was having is that the control surfaces didn't do anything I expected when in fixed wing mode, in multirotor mode the motors spun and the control surfaces seemed to do the right thing, but in fixed wing mode there was no authority whatsoever over the control surfaces, it was if the radio was dead.
Today I managed to get a config together by initially configuring the Pixhawk as a FX-79 wing, then rebooting and going through all the accelerometer, gyro, airspeed config and then once that works (albeit with all the wrong outputs moving) I set the autostart ID for the VTOL config and rebooted and everything seemed to work! So it looks like you have to build and configure a regular flying wing, get everything working nicely and gradually build yourself up to a VTOL - not my style!
So now I have a wing which behaves properly in manual mode and looks like it does some cool stuff in multirotor mode. The control movements are a bit wild in fixed wing mode so I'm not going through all the PID parameters and toning them down to something sensible, unfortunately the PX4 Autopilot doesn't have such luxuries as autotune...
I've got the rat's nest sorted a little inside the bay, but this needs a bit more work. At the moment I'm using a wiring harness from a quad hacked with a couple of extra connectors for powering things here and there. Once I've got everything settled down I'll make up something neater:
(http://www.peely.com/wp-content/uploads/2015/05/IMG_1046-e1433008930619.jpg)
I also popped round to Rob's last week and picked up some oversized winglets, which are reinforced with carbon rod. These winglets are intended to support the wing upright whilst on the ground, but don't look too comical in flight:
(http://www.peely.com/wp-content/uploads/2015/05/IMG_1044-e1433008790207.jpg)
She's almost ready to go although I'm still not quite ready to put some props on her:
(http://www.peely.com/wp-content/uploads/2015/05/VTOL.jpg)
Hopefully if the weather plays ball we should be taking her for a spin towards the end of next week.
Neil.
This design looks much simpler, than any other Vtol project I have seen.
Hope, it works well, and if it does, perhaps, FlyingWings could release the Vtol version of the wing too.
Looks awesome. Let me know if you need some air to air filming!
Wow.. Nice job so far.
Brucey You better go to maiden with him.. You know what he's like. :laugh: :laugh: :laugh:
Seriously .. If it doesnt feel right just put it back in the car and test again.
Anyway Will be fascinated to see how it goes!
regards
Andy
Lol Andy, I seem to recall things don't get much better when you're there. Lampposts seems to magically appear from nowhere. :o
Any news?
'fraid not. I kind of mothballed this for a while when other fun projects came in. I'll put some effort into trying to get the elevons to move a bit more to satisfy my irrational opinion that this is all that's stopping it from killing us on launch.
Rob's really keen to get this in the air too so as soon as the weather dies down and I've got the elevons moving we'll do it.