Dynamixel rear mount point question

[_ADAM_]

First, sorry if I'm using the wrong terminology, hope you understand what I mean.

On the back of the Dynamixels, there is a hole on the same axis as the output shaft that is typically used to mount idlers for "hinge" brackets.

I'm wondering if this mount point is connected to and supports the output shaft that drives the horn. I kind of assume it is but I am not sure, and I don't want to dismantle one. I'd like to know for AX-12, MX-28 and MX-64.

If so, then it will be beneficial to support the output shaft and reduce the lateral load, which is bad for the gears. If not, then it may not be worth the extra work and added weight.

If it is, then it will affect how I will design joints where a hinge bracket is not used, such as humanoid shoulder pitch or hip yaw.

[tician]

No, it is a nut mounted in the rear face plate of the case. The output gear of MX servos is supported by one ball bearing in the front face plate of the case and a second in the central case just before a hole that permits non-contact sensing of the magnet by the PCB+sensor mounted on the opposite side of the central case and enclosed by the back face plate. The AX servos use plastic bushings and electromechanical encoders/potentiometers instead of ball bearings and magnetic encoders, but same general layout. The older RX/EX use bearings like MX and electromechanical encoders like AX.

[_ADAM_]

Ok, thanks. That's interesting. Glad I asked. So I won't bother with adding any extra support there.

[tician]

Basically, the nut is there only to hold external bushings/bearings on the case and keep them co-axial with the horn so that there are no extreme bending moments that will break off the output spline and/or damage the case. In things like the shoulder tilt and hip yaw, where the frame is only attached to the horn and not supported on the back face, the vast majority of the load is an axial force on the horn (pushing it into the servo and pulling it out) so no support is absolutely required on the back side. An additional thrust bearing between the frame and servo mount would not hurt if expecting lots of non-axial forces that might cause the horn's axis of rotation to 'wobble'. In other joints where the loading is a mixture of torque and radial forces, the frame must be doubly supported (front at horn and back at nut) else the frame will twist and the gears/case will break. Still run the risk of shearing off the output spline and/or breaking the case even when doubly supported, but much less likely than when singly supported only by horn.

[_ADAM_]

Ok, I see. I'll take another look at it, and it if it's not too much of a pain, I'll still try to double support. I want it to be as robust as possible without overbuilding.