To reiterate what others have said, and for posterity, be careful with the concept of “throttle” in ardupilot. The absolute throttle value that ardupilot uses is a function of several variables, which might be different than what is used by eCalc.
For instance, lets say you have MOT_PWM_MIN/MAX set to 1000/2000 and ardupilot reports that hover throttle is 50%. For the sake of argument, this means that the average real PWM sent to all motors is 1500. But, if you configure ardupilot to lower MOT_PWM_MAX to 1750, then when ardupilot is sending 1500 pwm to the motors for hover, the throttle percent will change to be (1500-1000)/(1750-1000) or 500/750, or 66%.
This is why you need to dial in your motor parameters (and not just those) before you use the throttle value for things like notch filters, because the throttle is an “abstract value”
Yea, sure it should. The fact that your average outputs at hover are ~1700µs, which is close enough to what eCalc predicts, is indicative of insufficient thrust:weight
As a rule of thumb thrust at 60% throttle on charged battery is max what you can safely use.
You should keep in mind that throttle increases when battery discharges, let’s say for liion battery it can discharge from 4.2 to 2.8v (or even lower), that means than if at start of flight you have 60% throttle as the battery discharge you you will get 60*4.2/2.8=90% throttle, which is close to the edge for attitude control.
You can try bigger prop pitch, it can increase a max thrust a bit but in reality you need bigger prop, like 10" that will be able to handle such load.
On that note, I currently have a misconfigured (connected other way around) current sensor. The doc says, voltage compensation uses an estimated battery resistance. Since I don’t have the current sensor, is there a resistance paarameter I can change manually so that the compensation is more accurate? Or should I not touch anything since the drone has flown pretty well?
I need to flash the Bluejay firmware to support BDShot. How is the stability of this firmware? Also, I would like to confirm if the F405 V4 supports telemetry for temperature?
EDIT: It would be good if you can tell me a general test to see if the drone really feels good to fly. Else, how do you gauge it as not flying “pretty well”?
I’m glad I finally got my plane fixed and managed to shave off a bit of weight. However, during today’s flight, I noticed some yaw issues. How should I optimize this?
Thanks so much for your input! I’m still working on optimizing the aircraft’s weight—though I know this will take some time to get right.
Also, I’m really in need of guidance and tutorials for tuning the rate PID parameters right now. Would you happen to have any tips or resources to share on how to approach this?
On a related note, I’d also like to ask: What is the actual goal of PID tuning? What should a “perfectly tuned” PID look like in practice?
Hey Dave, could you share a solid reference curve that I can learn from for the tuning? Or even a flight log from a well-tuned aircraft would be awesome—it’d help a lot to have a practical example to follow!
A very practical, linear and step by step example to follow is the ArduPilot methodic configurator software.
It is already as simple and as fast as it can be. Moving any faster than that will give you an incorrectly configured vehicle.
The learning curve of it is also the best one. You only learn necessary stuff. All that is fluff is left out. If something is included, there are very good reasons why it is there.
There is no one reference curve, it’s a multi-step process with a series of adjustments building on the previous ones. As you work thru the process this will become clear.
