Soooooo...

After my adventures with a little foam plane I started looking for something slightly larger where I could use the same battery and motors as my 4" kwad. I quickly found out about the LIDL Glider and after watching a gazillion videos I knew it would be perfect because it's super versatile, there are many ways to turn it into an RC glider, wing, speeder,...

I wanted my twin engine airplane to be reasonably fast but not scare the shit out of people at the park. Coming from sbang-y freestyle kwads I also wanted it to be as agile as possible. And last but not least I wanted a detachable wing because even though it only has a 86 cm wingspan it would take too much storage space in my small house if it would all be glued fixed.

The glider is usually only sold by LIDL during spring but I found a reasonably priced "Jamara Pilo XL" alternative on Amazon here in Europe: https://www.amazon.de/gp/product/B08JVG4RP3 . I think it's the same quality as the LIDL glider, but it's the design from last year with a narrow fuselage and blunt rear edge of the wing. That didn't bother me too much though, and at least it has a nice color scheme and print.

Like most other builds I found online I took the dihedral out of the wing so I could cut nice and large ailerons. The best way to flatten the wing is to pour boiling water over it and bending it carefully. Then I put it on the flat shower tray, covered it with heavy towels and drenched it all with boiling water. This made it absolutely perfectly flat.

I reinforced the wing with a thin 10 x 0.5 mm carbon fiber spar. This makes it very strong and stiff, and lighter and more aerodynamic than using a 4 or 5 mm tube. The ailerons are attached with tape hinges like most people do. I put the servo's on the bottom to keep the top of the wing clean. To prevent the links from getting damaged while landing I glued some wing tip protector skids right next to it.




I spent a lot of time engineering my motor mounts so they would be strong, light and dismountable. They are made from a 15 x 15 x 1 mm aluminum corner profile, slide into the wing from the bottom, and are held in place with two M2 screws that go into 20 mm standoffs that I glued horizontally in the wing. There's also a thin 1 mm carbon rod glued to the foam in front of the motor mount, to prevent the thin piece of foam from breaking off.








To make the wing detachable I cut a piece out of the top of the fuselage. It's designed such that the wing can be slid into place from the back, and is held in place quite nicely in the front. There's a 50 mm M3 screw that goes through the front part and through the wing to an M3 nylon standoff glued in the bottom part. A large washer hidden inside the top prevents the screw head from ripping through. A 20 mm M3 screw holds the back of the wing in a similar way.

Because cutting away the top of the fuselage and hollowing out the center part to make room for the wires and connectors made it pretty weak, I reinforced it with some 1.5 mm carbon fiber rods, right where the wing would rest. Together with the 4 mm carbon tube on the bottom (which goes all the way to the elevator) everything turned out to be very strong.








The elevator servo is tucked away within the stabilizer, this is the neatest way I saw it done in one of the YouTube videos.






The 850 mAh 4S battery goes in the front at an angle, to keep it as close to the center of gravity as possible. In the end I still had to add 10 g of weights to the tail.




The electronics are mounted to a 30 mm wide aluminum plate which acts as a heatsink for the 4-in-1 ESC (of which we're only using two outputs) and the Caddx Tarsier board in the front. This is what it all looked like for the maiden, some parts are still held in place with tape but will get a permanent place after testing:




Some random things I learned while building this:

• Cyanoacrylate glue works great to get a permanent strong bond between foam and carbon fiber
  
• Add baking soda (bicarbonate) powder to cyanoacrylate to get an instant hard filler
  
• Use normal foam glue (like Uhu Por) to glue large parts that need time to position correctly (like the wing spar)
  
• Hot glue is a mess and it's very heavy compared to other glue, but it's ideal for parts that may need to be disassembled later because it detaches easily with the use of isopropyl alcohol. I used it to glue the servos, horizontal stabilizer and tail reinforcing tube (which has the elevator servo wire underneath).
  
• Gemfan 5126 biblade props are too heavy for 1408 motors. Smoked one of them within 30 seconds on the bench. 4032 props are also a bit too demanding. Using the safe 4023 props for now and ordered HQprop 4X2.5 biblades, curious about those.
  
• Quality control on Caddx Tarsier sucks. I broke not one but TWO Tarsier cameras because of a faulty cable with a crooked connector which probably shorted something crucial. They stopped showing analog signal but they still record digital, so I may still find some use for them.
  
• Creating a plastic canopy is not easy. Been messing with plaster moulds and PET bottles and stuff but it turned out ugly. May try with some balsa or just keep the electronics "al fresco"...
  

For this build I got a new Mamba F405 Mini MK 3.5 flight controller because it fits the flight stack, has a barometer,  a nice USB C connector (kinda hate the flimsy micro connectors...) and is very affordable.

To make good use of the flight controller's capabilities for a wing I decided to use INAV instead of Betaflight. Tiny problem though... INAV only allowed me to use two servo outputs (motor output 3 for the elevator and motor output 4 for both aileron servo's) so there would be no way to configure flaps. INAV doesn't support remapping pins like Betaflight does, so I couldn't use the LED_STRIP pin to drive the elevator servo like I did before with my little glider.

So I took some time to dig into the source code and figure out how to create a new firmware target where the LED_STRIP pin would be defined as a servo output. In the target.c file where the timers are defined I simply had to change the line where it said TIM_USE_LED to TIM_USE_FW_SERVO (the "FW" stands for Fixed Wing by the way). I put that line before the definition of the two other servos to keep the order logical. In the target.h file I removed two lines related to LED_STRIP and decreased TARGET_MOTOR_COUNT from 4 to 2. This last change took me a while to figure out but without it, INAV was really confused about the motor and servo count I wanted to use.

Here are the changes compared to the MAMBAF405US target. It's not an actual diff but it's more readable like this for anyone interested in the code;



I've attached the full source code and compiled firmware for the new "MAMBAF405US_I2C_3SERVO" target. Not sure if I should try to get it added to the main INAV repository? I'm probably the only one crazy enough to use a quadcopter controller for a wing? 

Oh that's awesome! It's on my to-do list as well. Going to use folding props to minimise drag when gliding?

(12-Dec-2021, 11:49 PM)skipper_is Wrote: Oh that's awesome! It's on my to-do list as well. Going to use folding props to minimise drag when gliding?

I don't really intend to glide a lot, I think it's a bit too heavy (440 g now) for that. Also, I'm not sure if there are (backwards) folding props that are small enough? Plus I'm afraid they would damage the wing while folding.

First tests show that this plane can keep its altitude with about 3 to 4 amps worth of throttle. With average flying around 5 amps and short bursts to 20 amps I got about 10 minutes of flight, which is plenty for me because I'm comparing to barely 4 minutes of flight flipping around with my freestyle quads

Nice build!

For a plane, you want a smaller diameter prop with a high pitch speed. That will give you much better efficiency, and a higher top end if you want to push it (since you seem to have built quite a sturdy frame). Something like this would be ideal for those motors:
http://www.gemfanhobby.com/show.aspx?id=162&cid=45

(13-Dec-2021, 05:19 AM)V-22 Wrote: Nice build!

For a plane, you want a smaller diameter prop with a high pitch speed. That will give you much better efficiency, and a higher top end if you want to push it (since you seem to have built quite a sturdy frame). Something like this would be ideal for those motors:
http://www.gemfanhobby.com/show.aspx?id=162&cid=45

Thanks!

I thought a bigger prop disc area is always better for efficiency? I do understand that a higher pitch might be better for an airplane though because there's usually lots of positive air inflow, totally different to a freestyle quad where the props regularly handle static of even negative air flow.

Now luckily I actually have quite a few of those Gemfan Flash 3052 props lying araound (from before I turned my 3" Gecko into a 4" Gecko) so I'll definitely check the difference it makes! Heck I might even reconsider trying the 4032 props - they made the motors fairly warm during a bench test but it might be ok in flight...

Here's a video of my second flight after some initial testing and tuning:



I'm really happy with how it performs. The roll rate is about 450 °/sec and it can climb with about 12 m/s. Logs show acceleration forces over 8 G on the Z axis.  Not sure how fast it is, guess I'll have to install a GPS module for that.

I'm also impressed with the INAV performance, Acro mode works great and Horizon mode is nice for cruising. In Manual mode there's a lot of wind wobble to handle and I constantly had to fight an imperfect elevator trim.

I intended to follow the process outlined in https://inavfixedwinggroup.com/guides/tu...n-the-air/ but I had way too much fun to bother tuning it right now  

I've saved the Blackbox log of this flight here https://drive.google.com/uc?export=downl...uQaTg0Dd6b so if anyone can help me with some general tuning guidelines that would be great.

Here's the current INAV config. If you find anything that could be improved I'm very eager to learn!

Apart from the general tuning tips I asked above I have some specific INAV questions now that I haven't found the answer to yet. I'll number them for easy reference and hope someone knows the answer:

1. You can see in the config I posted that I had the roll rate set to 270 °/second, because that's what I thought to be the maximum this plane is capable of. But in the video I posted (at about 20 seconds in) and in the blackbox log (around 38 seconds) it's clear that it reaches roll rates over 450 °/second while being in Acro mode all the time. Why does this happen? Shouldn't the PIFF controller limit the rate to the maximum I set?
   
2. When I did loopings it was important not to pull the stick all the way because that would lead to high speed stalls. So I know I'll have to limit the elevator throw to prevent those stalls. But I was wondering if it's a good idea to mix in some aileron action to achieve a higher pitch rate? Like 10 or 20% flaperon action when pulling the stick back, and 10 or 20% spoileron action when pushing down? I know how to set it up (on a switch, for 0, 10 or 20% mix) but I just wonder if it's a good idea to begin with?
   
3. The ESC makes the motor beep irregularly (like once every 3 to 30 seconds, without any pattern) when the plane is not armed. Is this intentional? Did I set it up wrong? I'm simply using DSHOT600 and everything works fine once it's armed.
   
4. After disarming, the OSD stats screen shows 1/2 -> in the upper right corner. Is this like a page counter and if so, how do I go to the second page?
   

(Never mind that last question, figured it out, you just push the right stick to the right that's all. For some reason that didn't seem to work the first time I tried it)

Thanks!

I've been having sooooo much fun with this airplane!

The biggest change was to add a BN-180 GNSS module, which enables me to fly far away without having to worry about failsafes. Return-to-home works very well and gives me a lot of peace of mind.

I've also been doing some waypoint missions. I used to think that kind of functionality is boring but I was so wrong about that - it's pretty amazing to let the autopilot do its thing and now I can even fly beyond the range of the ExpressLRS 2.4 Ghz transmitter and FPV system... Long(ish) range flights sure are thrilling and it's always a relief when I start hearing the soft buzzing motors approaching again!

I'm super impressed with the cheap foam platform and I can't wait to get the new, more aerodynamic version once they are available again. I'm especially happy this airplane can do a fair amount of fast acrobatics, but is equally capable of flying slow and stable, unlike most wings that need a bit of speed to stay in the air.

Here are some record numbers I gathered so far:

• highest air speed (diving): 142 km/h (88 mph)
  
• highest ground speed: 123 km/h (75 mph)
  
• lowest ground speed: 0 (with a fair bit of head wind!)
  
• longest flight time: 13 minutes (and I wasn't even trying - I'm sure it can do 15+ minutes flights, and even better during summer)
  
• longest total flight distance: 8 km (5 miles)
  
• farthest ELRS range before failsafe: 1.8 km (1.1 miles)
  
• highest flight: 1.5 km
  

Enough words and numbers, here's a calm flight over a beautiful piece of nature 

Nice!
What props are you running now? And where do you have your ELRS antenna mounted? Something is seriously wrong if you're getting failsafes at 1.8km.

75 mph is pretty impressive for a glider!

Right. Actually I've tested quite a few props;

Gemfan 3052 works fine but doesn't give me a high top end speed.
Gemfan 4023 is slightly better and also a little more efficiënt.
Gemfan 4032 gives me more thrust but higher current draw at the top end.
Gemfan 4052 wants to BBQ the motors... Didn't fly them.
HQ 4x2.5 provide decent thrust at medium throttle and reasonable current draw at the top.

All of them give me a top speed around 110 km/h at level flight. The HQ biblades have the best efficiëncy at every throttle input so that's what I'm currently flying with.

I agree that 1.8 km is not a great result for ELRS but the main reason it failsafed was probably because I was respecting the maximum 120 m (400 feet) flying height and some trees were blocking the direct path. The mounting of the antenna is definitely not ideal because it's on the right side of the fuselage and partially blocked by the battery depending on the direction I'm flying in. But I got a Matek R-24D on its way which will allow me to put an antenna on both sides of the fuselage, and I can angle both differently to avoid the null zones.

Still, the video gets pretty bad at those distances so I'd also need a better patch antenna if I want to fly further, I only have a Realacc TFP-Mini now...

But I'm not too focused on long distance anyway because it's hard to find nice open spaces where I live in Belgium...

This little airplane takes you to Heaven and back!



It has always been my dream to fly above the clouds and I'm thrilled I finally got a chance to do so!

With the 850 mAh batteries I'm using, the airplane barely has enough energy to make it up there though. So I'm looking into powering my next airplane with some 3000 mAh 18650 or 4000 mAh 21700 Li-Ion batteries.

V-22, I think another reason for the limited ELRS range was probably because the Caddx Tarsier emits some horrible interference. I've had nothing but trouble with it on my Apex 4" also, where the receiver sits right on top of the Tarsier board. I'm now eyeing a Caddx Peanut (which I'm happy is NOT manufactured by Caddx) but I wish it wasn't as expensive as an entire kwad/airplane, especially considering you could lose it in the blink of an eye... Ugh, this hobby is going to ruin me!

My latest (and last) flight...