I just finished building a tricopter today and started setting it up. It is running APM 2.6 with Arducopter 3.1.5 firmware. At the moment it can barely hover at full throttle. It looks like after 1/2 throttle the APM simply ignores the inputs from the radio, as nothing happens to the motors, they don't even try to spin faster after about 1/2 throttle.
With ESCs connected through APM I am getting 13-14 amp draw and full throttle. With all 3 motors connected directly to the receiver I am getting 24-25 amp draw. So, basically, the best I get from the APM is 52% power. I know that the APM needs to reserve some headroom for stability, but not 48% worth of headroom.
Here is what I have tried so far:
- Calibrated the Radio in APM
- Made sure that radio trim levels are set such that min is bellow 1000 and max is above 2000
- Calibrated Servos using APM
- Manually Calibrated the Servos using direct receiver connection
- Set all the throttle related values to maximum setting. Max throttle is set to 1000, curves are off, etc.
- Tested Acro Mode, same thing
- Re-calibrated compas, accelerometer, gyro, etc.
I have no idea what else to try. I saw some discussion about this from 6 months ago, but nothing was resolved back than. Any ideas?
Ok, after fiddling around with the APM with no luck I decided to try and change manually the throttle calibration settings in the ESCs. I used BLHeliTool to set the max throttle to 1900 and multi gain to 1.25. Now the tri is pretty punchy. I don't think this is an ideal solution, because I am assuming that the APM still thinks that the max throttle is 2016, and may do something bad based on those assumptions. However, now it feels like it has good power and can lift it's self up well.
I have had similar issues, seems your experience is similar to mine. Looking at my RCout logs they never go above 1907us. So like you I tweaked my escs (mine are usb configurable), and I gains about 20A draw from the battery (55A to 75A or so). I always check my logs and even at full payload full throttle the value clips at 1907us.
I have seen other people ask similar questions, I haven't seen a good explanation yet.
I still love ardu copter....but this niggles me. Not sure if its a set up thing or inherent in the code.
Have you got RCout enabled in logs?
What value are you getting at "full" throttle?
I could not figure out how to get logs out of APM Planner 2. If some one could point out how to do that, that would be grate. The info in the wiki is for Mission Planner, but I have a Mac, so I have to rub Planner 2.
However, Servo Outputs on the MAVLink show me that at full throttle the servos are going to ~ 1650-1750 out of 2016. When I tilt the tri on it's side, to get it to spin up one of the motors, that motor goes to the full 2016. So that looks like the APM is reserving some headroom for stability, which is logical. However, it feels like my motors develop most of their thrust in that last 300/2016 range. Is there a way to tell the APM to reserve a little less headroom perhaps.
Overall, after fiddling around with some settings and calibrating and recalibrating the ESCs, I got the tri to now fly around for the most part. But it still feels very heavy when going up. If I bypass the APM and just gun the motors directly, the tri will punch a whole in my sealing. It's got the power, but I think the power is curved around the very top of the ESC range, and I can't use it.
I'll try playing around with the thrust curve, maybe I can tune it there. Not sure. Any other suggestions?