Kunteng sensorless control doesn't work
With this article I would like to draw attention to the fact that I would like to use the Kunteng controllers for my projects, but that Kunteng does not respond to complaints via email that the controllers do not work on some common BLDC motors.
BLDC motor controller with sensorless back EMF detection
Sensorless brushless motor controllers, as the S06P, eliminates the need for built-in Hall sensors, which were normally required for commutation. These controllers utilize a technique called Back Electromotive Force (Back EMF) measuring to determine the rotor position and control the motor's operation. However, sensorless BLDC motor control does not always work well as with the Kunteng motor controllers:
Kunteng motor controller problem with the Q85 hub motor
In the past, I had had good experiences with the old Kunteng KU63 sensorless brushless motor controller. It works very well with my Q85 hub motor, and I have been using this combination for 10 years without any problems. But this controller is discontinued. The new Kunteng controllers have unfortunately proven to be incompatible with my Q85 motor. I have made a video of this issue. The motor exhibited loud squealing noise when accelerating (fast or slowly) and stopped working than, revealing a malfunctioning back EMF circuit within the Kunteng controller as the cause:
The back EMF measurement circuitry of the new Kunteng controllers differs from the old KU63, the new controllers use a separate board for this:
Kunteng sensorless control doesn't work
Kunteng Back EMF Detection problem
There are others who also have problems with the S06P controller in combination with motors such as the Q75, Q85 and Q100 see the forums:
https://endless-sphere.com/sphere/threads/q75-loud-squealing-noise-with-acceleration.81852
https://endless-sphere.com/sphere/threads/q100c-cst-noise-then-cuts-out-under-load.72169
https://endless-sphere.com/sphere/threads/pairing-q75-with-lsw-675-not-s06p.87916
Kunteng sensorless back EMF detection board KTE-QHE
I have redesigned the sensorless back EMF detection circuit. The circuit contains a separate STM8S003F3 microcontroller:
The circuit of the Kunteng sensorless back EMF detection board KTE QHE
Normal operation
Hub motor test bench with wheel on trainer
Kunteng S06P back EMF detection measurement
The operation of the back EMF Detection circuit can easily be derived from the 4 signals measured with the Rigol MSO5104 oscilloscope:
S06P back EMF detection circuit normal operation
The circuit of the sensorless back EMF detection board shows where the four probes are connected.
The cause of the Kunteng motor controllers problem
To investigate the cause of the failure, I mounted the Q85 motor on a test bench, which was created with a wheel-on trainer. The failure must be initiated by temporarily overloading the motor with approximately 400W. This was achieved by using a cloth to brake the steel flywheel of the wheel-on trainer. Using a Rigol MSO5104 4-channel oscilloscope, I determined the moment at which the S06P commutation began to fail, and through this, I identified the cause of the failure. The oscilloscope screenshot below displays the signals at the moment of the malfunction:
Kunteng motor controller oscillation problem
To the left of point C, the operation is normal. To the right of point C, the controller starts oscillating at a fixed frequency, resulting in a loud squealing noise, as seen in the video. The motor controller no longer recovers from this unstable state and remains stuck in oscillation. Starting from point C, ch2 exhibits a sawtooth signal, see point B. This signal is distinctly different from the time preceding point C, where the operation is still normal. The cause of the first sawtooth pulse is that the motor controller incorrectly drives the motor once. The reason for this is unclear, but it is likely that the initial cause does not lie in the back EMF detection circuit. Due to the first sawtooth pulse, the STM800 processor generates an incorrectly timed commutation signal. From this moment on, the commutation of the controller becomes disoriented, leading to system oscillation.
The problem can be solved by making the software of the back EMF detection tolerant for erroneous signals generated by the comparator.