Moderator

DIY Carbon H-Frame Quadcopter

3689479892?profile=original(More construction photos in link below)

I was flying with an APM2 equipped X525 quad frame but there were a couple of thing I didn't like, namely:

if using a GoPro or other wide angle camera the props would show in the far corners of the video.

Also, orientation at a distance was difficult even with two bright neon-orange legs.

Thirdly, it didn't fold as compactly as I would like for transportation.

Lastly, it looked like all the other X525 framed quads.

So after seeing an H-frame quad on the 'net somewhere I set out to build my own. The first frame was from 3mm ply and balsa using the aluminium arms and feet from the X525 but stiffness was a major issue with motors 1 & 2 sitting higher and at a slight angle while in flight. This caused some strange yaw problems... well, there wasn't any at times, the motor angles cancelling out their yaw effect. I tried covering the ply with very light carbon tissue and epoxy and even some carbon tow braces but could never get rid of the twist between the front and rear arms. A fumbled toggle switch at the wrong moment caused a crash and showed another reason why ply is a bad idea, it breaks very easily!

So...

Next was to find something really stiff and strong, either fiberglass sheeting or carbon, prices were a bit unreal both locally and online, while some suppliers showed no interest at all in selling me any (I'm still waiting for some replies).

I eventually found some 1.5mm woven carbon sheet at Hobbyking but it was only 300mm x 100mm.
The 100mm was fine but I needed 600mm (23") long. This is when an engineering friend came to the rescue (HUGE thanks again, Wladek!), he said I could join 2 sheets mechanically with fiberglass sheet and rivets! So 4 carbon sheets were ordered and some plans drawn up in CAD.

Wladek has some nice machines at his work so we were able to do some very accurate drilling, milling and cutting. 12mm aluminium square tube was sourced locally and milled into a C-channel, saving 1/4 of the weight. Two 0.5mm plastic strips were used to make up the extra 1mm thickness (the X525 arms are 13mm). The carbon is riveted with 3208 rivets to the c-channel.

At first we used the carbon in it's original rectangular shape but we have since cut down the sheet to its present form saving 1/3rd of the weight. If I had to do it again I'd use 1.0mm sheet rather than the 1.5mm saving another 1/3rd. Also initially the ESC's were inside the frame but after many unexplained "ins" and other problems in which I started to suspect an overheating ESC, I have now moved them out into the airflow under the props and all seems well.

Anyway, it flies really well, looks great in the air and fulfils all my requirements, especially no props in the video! (See here). Final weight is 1279g (45oz) dry, with a 3 cell 5000mAh LiPo adding 353g (12oz) totalling 1632g (57oz), about 250g (8oz) heavier than the X525 with the same hardware. Flight time is about 13 minutes. Motors are 20-22L's, ESC's are Super Simple 20A, batts are Zippy Compacts (3S 5000mAh), props are 11x4.7 APC SF.

More construction photos here: Picasa Album

Cheers

link to CAD file>H-Frame2.dwg

E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • If Graham or anyone else with experience with DIY H-frames could provide input on my design, I would appreciate it. Thanks!

  • Moderator

    0.8 or 1mm carbon sheet would have been more than adequate rather than the 1.5mm I used and if one could use something much lighter than the 12mm C channel I used for the internal braces AND keep the stiffness then it would be great. The stiffness diagonally across the arms was originally the big problem with my first version hence this attempt.

    So the trick is to design it stiff AND light AND relatively inexpensive...

  • OK Bummer, I wonder if would save weight and still be rigid enough if the arms and frame beams were made with 10 or 12mm square wood, and the frame plates still used CF or G10 sheet.

  • Moderator

    I've actually gone back to the x-frame, but only because of the weight as I much prefer flying the H-Frame. It ended up being about 250g heavier than the x-frame which cut my flying time by 2 1/2 minutes, felt heavy in flight too. With all the FPV stuff on it was 430g heavier. Flight time was down to 8-9 minutes.

    I'll add my last visualization CAD file to the blog post

  • Hi Graham, do you have some dxf files or equivalent for the main cf sheet sections? Are you still using this quad as is, or have you made some improvements to it? I'm quite keen to build something like yours asap! 

  • After the reinforcement with one more layer of balsa for body and arms (+20g) it is flying stable, at least on my basement, outside is rain. See below the outputs for each motors, looks in suynch and I did fly 8.6 min with GoPro camera (with case) and a 2200 3S 20C Turnigy battery.

    Regarding the yaws, I did notice I need to do slow movement as yaw and it turns.

    More testing outside tomorrow when rain stops.

    3692520140?profile=original

  • OK, I got the idea.When I lift one motor (without battery), it is at 2cm from level when the next motor starts to lift. When I have the battery, then this gap is 3 cm. I believe is too much. What I already did in progress is to add one more layer of balsa for the main body (now I have 3 layers, before was 2). And I did add also one more layer for each arm (now are 3 layers), but the 3th layer I did put it perpendicular. It looks more strong now, not yet finish since the glue is still drying. I will let you know my acheaveament. Adding this new pieces of balsa add like 30 grams more as weight, not too much. I will try also to see if tilting the motors will help.

  • Moderator

    If one arm lifts a lot while the others stay put then it'll show how much twist there is, mine now lifts about 1cm before the others lift with the battery attached, before it was about 3cm. As I understand the props can only go one way on each motor so not sure how he achieved the improvement. I haven't tried any other way.

    Putting the washer underneath shouldn't affect the flight characteristics all that much because the APM will compensate, it's just the mechanical effect of the twist that is opposing the yaw effect.

    Experiment and see?

  • Hi Graham,

    One question about your comment: " You can try tilting one or more motors a little with a washer under one side to increase the effectiveness of the yaw, do it in the opposite direction to the twist."

     

    Do you believe tilting some motors will not affect the others flying parameters (not yaw, but roll, pitch, etc)? I was believening all motors should point straight up, is that correct?

     

    Thanks for all your sugestion.

    Last thing: my frame is not foldable as yours, but arms are not "hardcoded"... they are not glued to the main base. I have 2 screws for each arm and I can remove the arms for packing. I wish to have some foldable frame like yours, maybe on next release.

  • Hi Graham, thanks for your comments.

    I will check that lift distance and reply to you, is that important?. When I manually tilt one arm, the frame looks rigid, no twist. Which means if one arm is tilt, the other 2 in diagonals tilt on the same way. Maybe I dont follow your question.

     

    About twisting, I did find some comments, see link below. Do you believe reversing the props as below, could solve the yaw twist issue?

     

    Thanks, Chris

     

    "After reversing the props... the flight characteristics dramatically improved...Yawing now works nicely - in fact, this twisting might now be considered an energy-saving advantage :-)

    "

    http://www.google.ca/imgres?imgurl=http://forums.openpilot.org/uplo...

    Redirect Notice
This reply was deleted.