I just did a quick demo video to show Arducopter 3.0.1 flying a large electric heli. This is my new GreyHawk Maxi development frame. Loosely based on TRex 550 mechanics, stretched to 700 size, and with a direct drive tail rotor motor. The DD Tail gets rid of a lot of unreliable, noisy, vibrating geartrain. The heli is very smooth and you can hear the blades instead of gears screaming. In person, it's actually eerily quiet, if there's a wind out, you can barely hear it from a few hundred feet away.
This video is pretty tame. But I wanted to show how stable the and precise Arducopter can be. Normally I shoot test video in worst-case conditions, but most marketing videos are done to show perfect performance, so here you go.
This is only my 4th flight on this new heli, and I'm still getting it dialed in. Alt Hold has absolutely no problem at all with the Asymmetric rotor blades.
This video is also the first appearance of the DD Motor control feature which should be coming out with 3.1. The motor control for the tail motor comes out of Ch7. The setpoint can be changed in MP, independently of the main motor setpoint. And it can be started before the main motor, or held on after the main motor is off (auto-gyro flight). One very nice feature of this system is that you can have different flight modes on the main rotor, low speed and high speed, but the tail motor retains a set speed, even with a very low main rotor speed, so you retain tail authority.
Comments
I should point out however that, when talking about efficiency... This thing weighs about... well I should get it on the scales but it is somewhere around 5 kg. The math works out where it seems it hovering at about 750W.
I have an H-8 that uses the same batteries. But I arrange the 4S 5000's in parallel rather than series for the heli. So same exact batteries. If flies for about the same 12 minutes. But it only weighs 3.6kg. And if you stick extra batteries on it, it really suffers. The 700 heli would hardly notice the extra load.
Josh, this is with the Asymmetric blades on. I'm not seeing much improvement. However, I can feel that they have more thrust at maximum collective, so that tells me I should be able to slow them down for a more efficient hover.
That also brings other challenges however. With the lower headspeed, there is a problem with cross-coupling. Basically, you commonly hear heli guys say that if they run the head too slow, they see a "bounce" on the pitch control. This is due to a complicated aerodynamic/gyroscopic effect called Acceleration Cross-Coupling. Joly has been working on this and his code is now in my fork. The only problem is it probably won't run on the APM. It really needs PX4. We're just at the limit of what we can do on APM.
Whatever ends up in 3.1 will probably the end of major development for APM on TradHeli anyway. I'm fairly well happy with it. I hope to be moving on to PX4 in the immediate future.
Matt, you are taking the official firmware out of MP?
I've also been having a problem with MP not recognizing my copter as a copter, but I assumed it was something to do with the code that I compile. I even tried to fix it yesterday but this doesn't work either.
https://github.com/R-Lefebvre/ardupilot/commit/45c399b3c47617ae782f...
But if the official code also isn't doing it, that's not good.
Wayne, I'm getting around 12-14 minutes. I may have gotten 16 on one flight but I'm not positive about that one. There is some opportunity to run the main rotor slower in a hover for more efficiency. But I doubt it'll break the 20 minute mark in any case.
Keep in mind though, that this is a VTOL that is perfectly happy cruising at 70-100 km/h. In fact, cruise at 50-70 km/h should use less power than a hover. Also, it can take a payload. As it is right now, I have 1.5 lbs of ballast on the nose for balance. The machine is laid out to carry a camera and gimbal on the front. The battery pack for this thing only weighs about 2lbs, so that ballast could just as easily be a second battery pack. It should almost double the duration, as it's not affected by weight the same way a multirotor is.
I think there's some more efficiencies to be gained too. I bet the motors are much bigger than they need to be. And smaller motors running at a higher % load might be more efficient.
I don't know that this dual-motor system is more efficient than a single motor, but it's definitely simpler, smoother, and much, much more quiet. It's actually kind of freaky seeing it fly because it's so quiet. It's hard to tell on the video what the sound level is, but trust me. I flew it around at speed last night and it was really neat. The tail rotor is easily the loudest part.
Further developments I want to make now on the code, is to create a second "flight condition". FC1 would have a lower main rotor RPM and higher gains for hovering. And FC2 would use a higher main rotor RPM and lower gains for flying at speed. I find that I often have to compromise on my gains between high and low speed flight regimes. The Rate P terms need to be lower for high speed flight.
That's what I thought :-)
Euan, they were not outside.
Darrel, actually, like most people I don't think my neighbors realize the danger, and come out to watch. I've had to tell them that I'll only fly it if they're not outside, and if ever I'm testing and they want to use their backyard just let me know and I'll land. I wanted to fly on Saturday but they were having a party so I didn't.
As a rule, I try to keep this to the bare minimum. I'm usually just checking it'll get off the ground, and then head to the club field.
Josh, the DD tail code is something I added, but it should come out with 3.1 I hope. It allows for either either a variable pitch tail, so it has an output for a governed ESC and a servo. Or you can use it with a fixed pitch tail where it outputs a signal for a variable speed ESC.
Very Nice!
homes *or gardens...doh.
-
1
-
2
of 2 Next