Hello,
Today I noticed that the maximum output I am getting from the Orel 20 is significantly lower than the maximum output I get from a traditional PWM ESC with the same battery/motor/prop setup. I am reading only 3 amps with full throttle on the Orel and over 12 amps with full throttle using a PWM ESC.
I also tested this combination using the UAVCAN GUI ESC Panel. The same results were obtained setting the output to 100 on the panel.
Is there some kind of limitation with the UAVCAN protocol or the Orel 20 that is causing this discrepancy?
Thanks,
Jon
Orel could be reaching the RPM limit. See the figure 2.10 in the Sapog Reference Manual. Consider increasing the PWM frequency and see if that affects the behavior. If it doesn’t, try decreasing mot_pwm_dt_ns by about 20%.
Indeed, that was the case. I am surprised I hit the RPM limit since I am only using 2300 KV motors at 12 volts. Would swinging a larger prop help mitigate this issue? I’m thinking I could get higher thrust at lower RPM. The current draw should scale with the thrust and as long as I am under the 20A, I should be good, right?
I increased the PWM frequency and the results were better. With default settings, I was getting 0.5 lb of thrust from the motor compared to another ESC which gave me 1.0 lb of thrust. By increasing the PWM frequency I was able to achieve 0.7 lb of thrust, so the results are better, but still not to where I expected.
I am curious for the reason there is such a frequency-RPM curve. Why is the output RPM limited by that?
Thanks,
Jon
Would swinging a larger prop help mitigate this issue? I’m thinking I could get higher thrust at lower RPM. The current draw should scale with the thrust and as long as I am under the 20A, I should be good, right?
Correct.
I am curious for the reason there is such a frequency-RPM curve. Why is the output RPM limited by that?
Sapog does not support high-speed operation because it is rarely needed in industrial/professional UAV applications because of high iron losses associated with high electric frequencies. The technical description is provided in the reference manual, but the overall idea is that Sapog collects several samples within a PWM period and then solves a linear regression problem to find the most likely location of the zero crossing. If the motor is spinning too quickly, there is not enough time to collect the samples, therefore Sapog has to limit the speed.
Thanks for the help and the extra information. It definitely helped me in my design. It looks like I need to buy some bigger props.