After my Talon build part 1 & 2 (here), you'll find herefater my first experiments with the installation of Pixhawk and other components in a fixed wing UAV. What's great with this Xuav Talon model is its huge internal bay. So I started to draw on paper the location of all of the components of my setup, as shown in top picture.
I imposed to myself some rules/principles:
-to keep different antennas and receiver as far apart as possible.
-not to use of the wings to install any components. This is a compromise to keep transport/assembly/unassembly easy (wings can be removed with two screws).
-contrarily to certain builds, keep the gps with a clear sky view (eventhough it would probably still work buried in foam)
-have backup power on pixhawk: 3DR power module + SBEC on RC rail (+ required to power servos)
-to plan wiring before installing components; to draw a wiring scheme as shown above. This to avoid bad/wrong connections and ground loops. On the latter principle, Pixhawk has a weak point : if you power , as a backup, via a SBEC the output rail, not only do you have to mess with the installation of a 5.6V zener diode to protect the board, but you are forced to connect the ground wire too (creating a ground loop with the ground coming from the power module). This should be adressed in a next card's version : integrate the zener diode in the PCB and make it right like APM was: a common shared ground for all circuits of the FC.
Let's see what was done,
1. Customized electronics platform to host Pixhawk:
I don't know what these foamy airplane manufacturers think about when they design their model : they always forget to plan for a leveled horizontal surface, inside the bay, to host the flight controller board (maybe they are old school and never use a FC ?).
As a consequence comes a first question: where to place the pixhawk board so that it is firmly held in place (with room to use an anti-vibe platfrom underneath) , at the center of gravity and perfectly horizontal when the plane is horizontal ?
My answer is to add a diy custom plywood plate held horizontally in between the two existing plywood pieces that are used to reinforce the wing attachment points on the fuselage:
Side view of the added black painted plate:
As can be seen on this last picture, Pixhawk sits on the custom platform. It leaves a big gap underneath which is useful to pass cables and/or position other components on the bottom below the platform. It gives also the advantage to move pixhawk far away enough from possible electromagnetic interferences of cables/electronic components that would be positionned underneath, on the bottom of the bay.
2. RC receiver and antenna's positionning:
I decided to start with a FrSky 2.4Ghz rc control (a long range opensource Hawkeye DTF UHF alternative is ready and tested but will wait for now).
I use a X8R FrSky receiver with the PCB antennas. It is supposed to give a 20% range increase versus the simple coax wire antenna's. To be efficient they have to be placed perpendicular to each other (to cover horizontal and vertical polarisation. Especially useful when the plane banks, in order to not loose the signal).
Question : where to place the receiver, pcb antennas and how to have them perpendicular ?
Answer : eat a popsicle; then insert and glue the popsicle stick in the fuselage on the back of the bay. This location is far away from the motor/ESC, far from the telemetry antenna and FPV antenna. One pcb antenna is then velcro'ed vertically on it.
Using a solder iron, I melted a hole through the fuselage to pass the antenna wires. The receiver is velcro'ed on the side of the fuselage as seen on picture.
The second antenna is place perpendicular to the first one on the fuselage body.
3. 3DR GPS and compass module installation:
I wanted an external GPS installation. I did not want it on the wing, nor drill a hole in the top foam cover.
Furthermore it needed to be close enough to Pixhawk because of these damned DF13 cables that are delivered way too short in the 3DR kit... (=> request for 3DR, in your next FC version, replace these cables by longer ones and use connectors that DIYers can modify easily)
The GPS + compass module is installed on the back of the bay where the air inlet (cooling the motor) is located. It provides a small horizontal surface that is perfect for the GPS to sit on. However a "cable pass-through" hole has to be melted through the foam, to reach to Pixhawk (again a fined tip solder iron makes the job).
4. Telemetry module installation
Current 3DR telemetry modules are delivered in a plastic casing that makes it easy to velcro to a supporting surface.
I decided to locate the module in the front of the bay where this telemetry casing fits perfectly on the internal side of the fuselage.
An iron solder was again needed to melt an antenna hole.
5. Airspeed sensor (Pitot tube) installation:
The airspeed sensor consists in a pitot tube that ideally should be installed on the front (nose) of the plane. However that will be reserved for the installation of the FPV equipment. Also, the sensor's silicon tubes and I2C wires are too short to go all the way from the nose to pixhawk.
So i chose an alternative installation on the side of the fuselage,
I found an easy way to attach the pitot tube to the fuselage so it is maintained fixed and horizontally about 1,5 inch away from the fuselage. A zip tie is used around the metallic tube. The "plastic tail" of the zip tie is glued with CA on a CF rod. Hot glue is used to reinforce and fix the rod perpendicular to the tube.
The cf rod is pushed through the foam on the side and hot glue on both sides seal the rod firmly.
When travelling, the metallic tube of the sensor can be slided out of the zip tie.
6. Side note on servo control horn and control rod installation:
It is said by xuav talon experts on rcgroups forum that the maximum deflection of control surfaces should not exceed 20 degrees. Indeed only very small surface movements are required to fly the Talon (and people tend to move their sticks too much).
So I decided to mechanically limit the surface's deflections by positionning the control rod attachment point outward on the control horn. That way surface's movements are limited to about 20 degrees.
The consequence of this is I do not get perfect 90 degrees angles between horns and the control rod. I hope this will not create more adverse effects than not limiting the amplitude of the surface's deflection...
This ends part 3 of the XUAV Talon build.
Soon to follow : FPV components installation.
Build update : I copied Gustav's pitot tube positionning in the front of the fuse. Therefore I had to diy make a 4 wire DF13 cable; for that I used a UTP shielded cable as shown below:
To be honest I never opened one of those X8R antenna plastic containers so I do not know what's inside. I am also pretty sure it radiates 90 degrees to the plastic enclosure.
Hugues, I don't mean to suggest changing antennas, just the orientation of the one that is in parallel with the fuselage. I'm not sure what type of antenna is contained in the plastic enclosure but if it radiates at 90 degrees to the plastic enclosure then during a bank the received power from that antenna isn't going to change barring internal obstructions in the aircraft. Because the received power from that antenna doesn't change during a bank, it isn't going to improve its reception, but of course it won't get worse either.
@Doug, a good dipole (or pair of) would certainly be better than the stock PCB antennas.
I will move my antennas away from the GPS but will keep at the moment the X8r stock antennas (not planning OOS flights for now anyway).
@ Hugues, I used slightly larger tubing, obtained from my local hobby shop.
Your position of the second antenna of your receiver are you sure the radiation pattern during a bank is going to improve your reception? A normal dipole type would be better if it was also 90 degrees to line of flight wouldn't it?
@Mark, thx I will move the X8R away then
@Gustav, where do you buy tubing extensions of the same diameter?
Thx for sharing your setup.
We really should get all the Talons together in one place :-)
You can have quite long runs of the tubing to the actual sensor without any problems.
I prefer not to have to worry about extra connections during setup.
I lost my last drone, by forgetting to connect the tubing to the pitot on the wing.
Per chance my setup on Talon didn't require any calibration for airspeed.
I did find that the 3DR power block was very inaccurate, I'm running 12000mAh, 3S, @ 9A for 60 Km/h.
After checking with a accurate power meter.....