R_Lefebvre
Arducopter Developer
Hi guys,
I was recently trying to help somebody tune a Carbon Cortex with a Pixhawk on it, equiped with... 16" T-motor props and I'm not sure which motors (assume appropriate). It had T-motor ESC's, I think maybe 40A?
I had more difficulty tuning this machine than ever before, Arducopter Autotune completely failed on it, and it's taken a lot of work doing a manual tune, still working on it. I have to look at high-rate data logs as it's impossible to tune heuristically as I normally do.
The reason being, it seems that the powertrain combination is very unresponsive. ie: Normally I manually tune by increasing Rate P until it oscillates, then backing off a bit, and doing the same for Rate D. But I tried this, increasing both terms until they were double our typical maximum, and it still wasn't oscillating. Looking at the datalogs, I could see the motor output was just a complete mess. Just pure noise, filling up the PWM range from 1200 to 1800 on all the motors. It's a wonder it was flying at all.
That's when I realized that something was slowing or heavily filtering the motor response. I don't think it's the propeller inertia, as I have an octo with 15" props that isn't like this. I really wonder if it's the T-motor ESC's filtering the inputs. This is no SimonK!
I've got the thing flying well enough, but it's still not quite done. I'm not sure this thing will ever be perfectly stable. I've had to turn down the angular acceleration lower than I ever have before. And it still bobbles every once in a while.
Is this a known issue with T-motor ESC's?
I was recently trying to help somebody tune a Carbon Cortex with a Pixhawk on it, equiped with... 16" T-motor props and I'm not sure which motors (assume appropriate). It had T-motor ESC's, I think maybe 40A?
I had more difficulty tuning this machine than ever before, Arducopter Autotune completely failed on it, and it's taken a lot of work doing a manual tune, still working on it. I have to look at high-rate data logs as it's impossible to tune heuristically as I normally do.
The reason being, it seems that the powertrain combination is very unresponsive. ie: Normally I manually tune by increasing Rate P until it oscillates, then backing off a bit, and doing the same for Rate D. But I tried this, increasing both terms until they were double our typical maximum, and it still wasn't oscillating. Looking at the datalogs, I could see the motor output was just a complete mess. Just pure noise, filling up the PWM range from 1200 to 1800 on all the motors. It's a wonder it was flying at all.
That's when I realized that something was slowing or heavily filtering the motor response. I don't think it's the propeller inertia, as I have an octo with 15" props that isn't like this. I really wonder if it's the T-motor ESC's filtering the inputs. This is no SimonK!
I've got the thing flying well enough, but it's still not quite done. I'm not sure this thing will ever be perfectly stable. I've had to turn down the angular acceleration lower than I ever have before. And it still bobbles every once in a while.
Is this a known issue with T-motor ESC's?
