I've been working on this quad for about three weeks. I based some of the frame design on the arducopter. The flight was done with the ardupilot mega+gps+compass in stable mode. The start was a bit disappointing at first, but it was soon resolved and possibly caused by a faulty RSSI indicator. As can be expected, it's way more stable than the tricopter. The only required nudges are throttle and where the quad is going. Soon, I'll test GPS Hold mode and a mode for navigation that I put in there. I removed some of the arducopterNG code and added some OSD telemetry code to interface with an OSD board I'm finishing.
More info about the build and some close-up pictures:
Comments
It feels heavy-ish, but the props I have are 10x4.7 with about 600-700g thrust. in terms of stiffness, the 0.5mm should already be enough. I considered U-tubing once, but that might be too much subject to torsion. The best I could find was at an online-shop for metal in Holland: http://metaalwinkel-metalen.nl/index.php?item=5&aitem=70. But I didn't like the prices and just went for the 1.5mm one. I'd say it'd be very difficult from the looks of it to get it so thin. Maybe some specialty online modelshop would have it for the fine work.
Terms of weight/flying time... I think I had it in the air for 5 minutes and that burned 2200mAh. My battery is a Turnigy 3.6Ah and another 5.0Ah. That should give me 8 and 11 minutes respectively. Doesn't sound like much, but it's enough really. I just wanted something finished for now and alu tubing doesn't cost much in comparison to the rest. If I want something else in the future, I can always replace the tubes.
I didn't pay attention to 6063 7075 or anything like that. Just didn't go for "anodized" stuff, as that would be heavier I think?
goed gedaan!
15x1.5mm square tubing, 30cm long for each arm. Screwed together on the sides using 20x2 alu strip bent at 90 degree angle. The tricopter I built before had close to 45cm armlength from center to rotor. I shortened this down a bit to get more stability. Props closer to the center travel less up/down to rotate the model in the xy plane, so should be quicker in regaining the set attitude, allowing some more aggressive PID settings. I decided on 30cm more as an intuitive choice rather than anything scientific. Getting the props too close increases the influence of one over the other due to turbulence at tips and downward air flow. When going forward, front props may leave more of a turbulent wake for the aft props if they're too close together. It's probably something of a tradeoff between transport, acrobatics, weight and turbulence interactions.