Required considerations when connecting motors

Dear sim builders,

several G-Seats have multiple motors per flap, like e. g. the GS-4 or @SeatTime's G-Seat.

The motors can be servos with their internal control unit (as in SeatTime's rig), normal motors with a single feedback pot for all of them, or maybe even stepper motors, who knows.

In each case I am wondering how to avoid that the connected motors are fighting against each other:

Servo approach:
Each of the connected servo motors tries to reach its final position, which in theory should be the same if you trigger them with the same signal. But minimal differences in each motor may lead to the result that each position is a bit different from the others. I found several discussions about this topic, like for example this one.
One proposed solution was to make the connection between the servo levers and the moving component (e. g. a G-Seat flap) a bit flexible instead of absolutely rigid.

Motor & feedback pot approach:
You could wire multiple normal motors in parallel and use a single feedback pot for all of them.
Also in this case, each motor will have a slightly different speed, resulting in different lever positions.
But at least the power will be zero for all motors when the pot has reached its destination angle, so that there should not be a continuous fight all the time.

I already had a private conversation with another forum member, and we have been thinking about possible solutions. Now I would like to start a public discussion and ask you to share your thoughts and experiences.

Best regards,
Markus

After doing some tests with 2 Servos without electronics i would like to make this proposal:

The mechanical construction could be as with the original GS-4:

I would remove the electronics of the servos and drive them with Arduino(SMC3) and Monster Moto H-bridge.
Because the servo motors have only 6V, but the MM does not work reliably with 6V, it would be better to connect always 2 servos in series to get 12V.
Only one servo-pot is wired as feedback-signal to the Arduino per flap. So the motors run always in unison and do not work against each other.

Thanks a lot for your investigations and the wiring proposal, @Bernd Manger! This looks promising.

Since this approach does not require the internal control unit of the servo motors, it should also be possible to use other gear motors instead, like e. g. this one, right? Because a servo with a torque of 30 kgcm costs at least between 30 and 40€, and with normal gear motors you could stay below this price.

Or are there other important capabilities of these servos that you would need? For example they are designed for this kind of forth and back movement and thus may last longer than other motors.
Any maybe an advantage of a servo gear is that the holding torque is low, so that you can turn it when it is not powered. I guess this is needed here in order to keep all parallel motors perfectly synchronized(?)

Best regards,
Markus

These are exactly the points. This is an abstract of the motor specs (pdf):

You should consider that scale modelling servos are mass products which can be offered at a low price. They are high quality, have ball-bearings and very low backlash.