When I test my drone for the first time and I run the throttle up form my remote with the props off the motors keep spinning even when I bring my throttle back down. Will my quad flip out if I put the props on and try to fly it? Why does this happen? It seems fine when I spin the motors up from the motors tab.
Why do the motors on my drone keep spinning up and max out when I run up the throttle on my radio on the bench?
This is a result of I term windup from the PID controller. It's completely expected and normal with props off, and will be fine when you put props on and go fly! Almost everyone new to RC multirotors asks this question on their first build, so don't feel bad!
The detailed answer to this question is related to how multirotors maintain stability in flight. Multirotors use a control system called a PID controller. This is a feedback loop control that takes information from the sensors onboard the flight controller and then executes change via the motors based on how the commanded input is different from what it is sensing. For more information check this white paper from the national instruments lab.
To understand this problem, the component of the PID controller we need to examine is the I constant, often referred to as Ki. Essentially this is based on the integral calculation between the sensed value and the setpoint or target value. What this means is that I get stronger the longer there is a difference between the sensed value and the target value. Hence change over time. What you're seeing as the motors slowly spin-up is the buildup of that Ki value getting stronger and stronger as flight controller is trying harder and harder to execute a change that it can sense, but can't impact. No props means no force to execute change, hence the windup. This is often referred to as I term windup.
Here is an excellent video from Joshua Bardwell on the topic:
1$\begingroup$ Excellent response! I wonder - is this effect as pronounced with Airmode disabled? I know that is the case when simply idling, but I’d be interested to see if it is the case with active throttle. $\endgroup$ Apr 15, 2020 at 0:33
$\begingroup$ Yes, definitely Airmode aggravates the problem, as it gives more strength to the PID controller at low throttle. When you drop the throttle to zero the PIDs still have full authority. Without Airmode enabled in Betaflight, when you drop your throttle to zero, the PIDs relax, which reduces the I term considerably and can minimize the windup. I don't want to necessarily mention it in the answer though because beginners will often hear "Airmode" and just turn airmode off, which is not a necessary solution. $\endgroup$ Apr 15, 2020 at 0:38
$\begingroup$ excellent, and very interesting. Thank you for your response! $\endgroup$ Apr 15, 2020 at 0:39
The reason when you bring your throttle back down the motors don't return to the previous idle speed is that you have triggered Air mode this is perfectly normal.
$\begingroup$ While technically true, this answer can be somewhat misleading. Airmode's only function is to allow PID controller to have full authority at low throttle. The problem is a function of the PID controller, not a function of Airmode. In fact Airmode is technically not a mode, it's the disabling of a "feature" that was programmed into early firmwares that prevented the PID controller from taking full effect at low throttle in order to prevent spazzing out on takeoff and landing or low-throttle bumps for beginners. So techincally Airmode is turning off a feature, not enabling a new one. $\endgroup$ Apr 15, 2020 at 1:42