I have a Tilt-rotor tricoper VTOL, and I suspect that the motor thrust during QHOVER is not linear, particularly below the 50% neutral hover throttle setting. Is there a way to test this? I am using DShot 600 on the 3 ESCs, so I assumed that ESC calibration was not needed.
ESC calibration and thrust expo are two different things.
Calibration is used with analog ESC protocols (PWM, PPM, etc.) so FC and ESCs agree where minimum and maximum throttle are.
Thrust expo is used to align internal motor+propeller thrust model with reality.
You should have adjusted Q_THROTTLE_EXPO in initial setup for your prop size. IIRC Methodic Configurator has the updated formula. Keep in mind that some ESCs already have some linearization curve.
Thanks for the response. Yes, I adjusted Q_M_THROTTLE_EXPO according to the prop size based on the AP wiki.
Perhaps thrust is not the issue. What I’m experiencing is that I don’t have very good descent control when descending from a stable hover. It descends fine at the descent rate that I set with Q_SPIN_DN, but if I raise the throttle to slow the descent rate, it doesn’t respond, and I touch down hard. Maybe I don’t understand throttle control in QHOVER. I thought that either above or below the neutral hover throttle setting the ascent/descent would be proportional to throttle position, up to the max ascent/descent settings. Yes, no?
It should be.
If the issue only occurs near the ground, your baro may be affected by ground effect, otherwise it may be caused by unsuitable Q altitude controller parameters.
First will require setting baro compensation and the second will require tuning Q altitude controller cascade.
I understand the tuning parameters for the altitude controller, and I can tweak them. Can you point me to the barometer compensation? I find that there is an IMU temperature calibration, but is there another type of barometer compensation test/adjustment?
Only this snippet
I would recommend looking at the log and comparing Barometric altitude and GPS altitude as well as vertical velocity. It should match observed behaviour.
I tried to compare altitude from barometer and from the GPS, but the GPS units are in units I don’t understand. They ranged between about 230-250, and I was only at an altitude of 16 feet or so.
Depending on software you use it should be meters, GPS uses AMSL do having altitudes shifted is quite normal.
Aha, Above Mean Sea Level makes sense for GPS units, thanks! For the same flight, though, the barometer output varied from 0-17 meters, while the GPS altitude only varied 10 meters. Is that normal? If the altitude controller is using both of these, how does it get the right altitude?
What really confounds me is that yesterday I performed multiple hovers and tried to control the descent with throttle all the way down for a soft landing. Sometimes that worked fine, other times the plane would descent as commanded until it reached about 3 meters, then it would descend rapidly regardless of a throttle increase to counter it. I’s as if sometimes the Vtol is getting accurate altitude when approaching the ground, and sometimes it is not. I’m thinking of installing a Rangefinder to provide more accurate and consistent elevation data to the FC.