I am building an x8 setup, and I wanted to try and test a design with the front booms almost straight to the side, similar to an H setup, and the rear booms back at a large angle, so the setup is asymmetrical if looking at it from the side. Picture below. I have it setup with the mixer settings from the mikrokopter wiki X8 seen here: http://gallery.mikrokopter.de/main.p...geViewsIndex=1 Testing it out, everything is spinning the correct way, and with enough throttle to test left/right/forward/reverse, everything responds correctly. However, when it lifts of the ground, it yaws hard counter clock-wise if looking at it from the top. Any thoughts on what this might be? I have ensured the props are all balanced and aligned. The rear booms give a little extra weight to the back of the copter, but its AUW is perhaps 2.5 kilos tops with landing gear (brushless gimbal with the bottom removed) so I don't think it will play too large a role in the yaw.
My first thought is check that all motors/motor mounts are level with the frame hub. Next I would triple check that the mixer settings are correct and all motors are spinning the in the correct direction.
Thanks for the reply Howard. I have gotten the booms as level as I can and double checked the mixer settings. Would any of the settings need to be altered to accommodate the motor layout? I see see mikrokopter has the bottom props set higher than the top. Any MK mixer experts out there?
I'm also wondering if the distance between the front motors are the main cause for yaw. My none scientific guess is telling me that front motors 1,5,3 & 7 should be set to spin at a faster rate in the mixer than the rear motors because of the unequal distance. I know larger diameter frames yaw much slower than small diameter frames. Now just imagine if there are unequal distances between groups of motors. There's bound to be some weirdness.
Getting that setup balanced might be a mission all by itself! Especially if you're running a 3 axis gimbal.
I agree. I'm no engineer, but I believe the way yaw works is by centrifugal force, and if there's not an even distribution of power around the CoG, it's not going to work. Maybe. Shaun? We need you! Ironically, you might be better off building it as a Y6 instead of an X8, but then you're totally on your own, and maybe should think about one of the existing frame designs. 3DRobotics have a Y6 that people like, although it's a way smaller bird.
My understanding is that the yaw is produced by differential torque -- which is imparted as rotational force to the copter as a whole via the booms acting as levers. You'll need to brush off your vector and matrix math, though. Check out this page which explains how mixer tables work. http://paparazzi.enac.fr/wiki/RotorcraftMixing I think the problem with your configuration is the asymmetry -- the "front" appears to be the place where the largest gap exists between booms and that, I suspect, needs to be reflected in the mixer table. Andy.
There is also this tutorial that's a little bit more approachable: A very laid back Australian, I'd say! Andy.
Thanks for the input, everyone. Andy, I learned a lot about MK from your DVD guide, so I am glad to have your input! That video will certainly be a great help in going further in my understanding of the physics behind it, which is my goal, to understand the behind the scenes workings rather than just the mechanical action of putting together off the shelf items. There is nothing wrong with that, of course, just a personal endeavor. Though it is an odd design, I don't think it is unflyable, just have to find the correct mix. I will report back after reviewing this information.
After reviewing the information, I have a much better understanding of how custom mixing works in regards to prop location in relation to the FC/center of gravity. I am able to articulate this design under the MultiWii custom mixing, but I am unclear as to how to translate this into MK Tools mixing which does not look at relation to the FC but rather the Gas, Nick, Roll and Yaw. Any ideas from the MK pros out there? Here is how the MultiWii mixer would be represented to my knowledge, though I could find no information as to how it expresses the coaxial placement of motors. I am placing them on the same position, only changing their rotation: #ifdef OCTO-X8-NEW motor[0] = PIDMIX(+48.3,-12.93,+48.3); //FRONT_L_TOP motor[1] = PIDMIX(+48.3,-12.93,-48.3); //FRONT_L_BOTTOM motor[2] = PIDMIX(+22,-12.93,+22); //REAR_L_TOP motor[3] = PIDMIX(+22,-12.93,-22); //REAR_L_BOTTOM motor[4] = PIDMIX(-48.3,-12.9,+48.3); //FRONT_R_TOP motor[5] = PIDMIX(-48.3,-12.9,-48.3); //FRONT_R_BOTTOM motor[6] = PIDMIX(-22,+12.9,+22); //REAR_R_TOP motor[7] = PIDMIX(-22,+12.9,-22); //REAR_R_BOTTOM #endif
I redid the motor mixing and got the yaw removed, however it is still very unstable, as the asymmetry needs to be taken into account in the mixer. I have looked further into the nick and roll, but I am hung up on how everything needs to equal zero. For example, if you place the FC at the center of an xy grid, and base your figures off of that, say the top motors are +10 and the bottom are -50 based on their location, you would end up with a total of -40 and it won't fly. Any thoughts?
I'm afraid this is at the limit of my knowledge without doing additional research. I've got to believe someone out there has documented how mixers work in relatively understandable language. (Glad the DVDs helped, Eric!) Andy.
Thanks Andy I have posted on the mikrokopter site as well, hopefully the folks there have some additional insight into how the MK specific mixer functions in this scenario. I will report back if I get it figured out.
I upgraded to firmware 2.0, recalibrated the acc and compass and took it out again. Lifts off smoothly, I can hover in the air, but the yaw response is still messed up, which makes sense as it isn't yawing around the FC as the center, but rather the center of the props. This seems like something that could be corrected in the mixer, but I will have to find someone who has some insight into the MK mixer for this type of setup.
These guys look like they've got the asymmetrical layout nailed. Give them a holler. http://www.sky-hero.com/Sky-Hero_Spyder6.html