Movement platform evolution (2DOF to 6DOF)

Attention!!! Important addition
I didn't like the low speed of the linear actuator. After measurements, it was only 110mm per second. At first I was a little upset, but after carefully studying everything, I found my mistake. ESP32 has logical units (high level) 3.3 volts. I thought the driver of motor would understand this signal. But it turned out, judging by the scheme, both digital and analog signals can be received on the driver pin.

It turns out he leveled my PWM signal with a level of 3.3 volts and perceives it as an analog signal of 3.3v (5v maximum). Apparently this is done for the versatility of the driver control (it understands both digital and analog signals). At home, I did not find transistors in order to make keys. Found the PC817 optocouplers. Below is the connection diagram.

I soldered one optocoupler for the sample and it worked. The maximum speed became 320mm per second !!! plus galvanic isolation for pin ESP32. I think the torque is now okay. If you understand what I wrote here, then be careful

Good work,

But that vibration...
The cause might be on the pots. They always have noise.
In the ESP32 code you can make an average of the read value to soften them.
Might get a bit "spongy", but you might remove most of that vibration.

I'm using level shifters for the 3,3 to 5V.
And added some batteries, they give a good punch, but at a cost €€. Not sure it's worth.

Yes you are right. It's all about the potentiometers. At first, the wires were in the same casing as the engine wires. Now I ran the wires from the potentiometers in a separate shielded cable. Before that, the value of the readings jumped very much. The median filter in the firmware helped a little. The mean arephmetic filter could not cope with such large outliers in the readings. Now there are no such jerks! But it is still possible in the future I will switch to a hall sensor or, like yours, I will take readings from the hall sensors of the engines themselves

I'm tired of these potentiometers! I delete everything and do it like on the internal Hall sensors of the motors. I redid everything in a couple of days and this is what happened. I now have two ESP32 boards that control six drives. Signals from the Hall sensors through voltage converters are converted from 5 to 3.3V and transmit information about the rotation of the motor to the ESP32. I removed the potentiometers themselves and instead soldered a voltage divider from two 10K resistors. When the maximum limit switch is triggered, a voltage of 3.3V is issued (or 4095 in a 12-bit ADC), with a minimum limit switch - 0V, and in the middle position - half of the voltage from 3.3V. All information on code and everything else from the @pmvcda project https://www.xsimulator.net/community/threads/flypt-6dof-brushless-diy-actuators.10799/. The only thing that seems to me is that the voltage divider is shown incorrectly. It didn't work right away until I redistributed the voltage divider.

Hi Michael!
I completely missed your build.
Congratulation! Well done and good documentation!
I am wondering, why you changed from the sunken-platform to an on top-platform? Did you have some interferences?
Wickie

I switched from sunken-platform to an on top-platform for the following reasons:
1. Mounting points for linear actuators are closer to the center. This increases the load on them, but with the same stroke length gives a greater angle and speed of the platform tilt.
2. The structure is more rigid, less connections.
3. To tilt the platform to a certain angle, a smaller stroke of the linear actuator is needed. And that means reducing the noise from the ball screw