stenzer
Advanced Member level 3
Hi,
my XIAOMI vacuum cleaner robot complained since a while that it's left wheel got stuck although there was no obstacle. He even some kind jumped when I was able to observe this issue. Below a solution for my problem can be found.
First thing I tried was to clean the wheel and the related contacts as recommended here. Everything was quite clean on my side so that did not do the job. So i disassembled the vacuum cleaner, to have a closer look on the electronics, especially everything related to the left wheel driver. Here I could identify the A4950 PWM motor driver. From optical perspective everything looked realy nice and clean. The electrolytic capacitors look fine, but I have not desoldered them to check them. This I avoided because of the conformal coating. The 0.033 Ohm sense resistor (see A4950 datasheet) is fine as well.
So here I suspected a problem with the supply voltage of the A4950 or its reference voltage. To solve the problem, I populated two non installed 0603 capacitors capacitors (470 nF / 50 V) the left wheel (C56 & C58), which are creating a PI-filter with a ferrite bead (region is marked in red below). For the right wheel the layout is somehow nicer and cleaner, and shows an actual ideal PI-filter "layout". For the left wheel the layout is not that nice, and even the smaller capacitor (0402) which was already installed, was farther away from the VBB pin of the A4950. I also installed a 200 µF / 35 V electrolytic capacitor in parallel to the SMD one, as I didn't have a suitable SMD replacement. Further, I installed a wired 100 nF ceramic capcitor to the reference voltage input pin of the A4950. Next to the A4950, a LMV358B from 3PEAK is located, which I assume is creating the required reference voltage. Due to the black solder resist and the conformal coating I did not spend much effort in analyzing the circuitry in detail. The LMV358 at least mentions "it is capable of driving heavy capacitive loads" (nothing more), further the 100 nF capacitor is isolated by a resistor if directly soldered to pin #4 of the A4950.
With this modifications the vacuum cleaner robot has been tested for several hours and it works so far. Before it was "stucked" already after a couple of minutes.
I hope this information might be of use for someone else.
BR
my XIAOMI vacuum cleaner robot complained since a while that it's left wheel got stuck although there was no obstacle. He even some kind jumped when I was able to observe this issue. Below a solution for my problem can be found.
First thing I tried was to clean the wheel and the related contacts as recommended here. Everything was quite clean on my side so that did not do the job. So i disassembled the vacuum cleaner, to have a closer look on the electronics, especially everything related to the left wheel driver. Here I could identify the A4950 PWM motor driver. From optical perspective everything looked realy nice and clean. The electrolytic capacitors look fine, but I have not desoldered them to check them. This I avoided because of the conformal coating. The 0.033 Ohm sense resistor (see A4950 datasheet) is fine as well.
So here I suspected a problem with the supply voltage of the A4950 or its reference voltage. To solve the problem, I populated two non installed 0603 capacitors capacitors (470 nF / 50 V) the left wheel (C56 & C58), which are creating a PI-filter with a ferrite bead (region is marked in red below). For the right wheel the layout is somehow nicer and cleaner, and shows an actual ideal PI-filter "layout". For the left wheel the layout is not that nice, and even the smaller capacitor (0402) which was already installed, was farther away from the VBB pin of the A4950. I also installed a 200 µF / 35 V electrolytic capacitor in parallel to the SMD one, as I didn't have a suitable SMD replacement. Further, I installed a wired 100 nF ceramic capcitor to the reference voltage input pin of the A4950. Next to the A4950, a LMV358B from 3PEAK is located, which I assume is creating the required reference voltage. Due to the black solder resist and the conformal coating I did not spend much effort in analyzing the circuitry in detail. The LMV358 at least mentions "it is capable of driving heavy capacitive loads" (nothing more), further the 100 nF capacitor is isolated by a resistor if directly soldered to pin #4 of the A4950.
With this modifications the vacuum cleaner robot has been tested for several hours and it works so far. Before it was "stucked" already after a couple of minutes.
I hope this information might be of use for someone else.
BR
