DIY 3DOF, AC Servo Ball Screw Drive

I am working on the design and build of a 3DOF rig intended primarily as a race car simulator, but will also checkout flight simulators. I have just completed the second iteration of a simulator made from wood to insure I have the cockpit dimensions defined. I am tall at 1.9M. "Gabriel the simracer" built a sim rig that I am using as a starting point, and he based his on the DOFReality H3. I will be using 220V 750W 3000RPM AC Servo motors to drive a 1605 Ball Screw, which in turn will drive a lever for my motion.

I have started collecting the bits and bobs I'll need.


Below is a 3D model of my actuators.


It is going to be fun!!!

Dont you need a double leverarm like they do with whormgear motors to get the top frame steable?

Where have you bought the universal joint and the flange bearings?

I purchased the U-joint on ebay, everything else on Amazon:

KFL000 10mm Zinc Alloy Self Aligning Pillow Block Flange Bearing, Zinc Pack of 6

KFL002 15mm Zinc Alloy Self Aligning Pillow Block Flange Bearing, Zinc Pack of 4

Rotation, 0-5V Output, Durable Full-Circle Measurement

F6000ZZ 10PCS Flange Ball Bearing 10×26×8mm Double Metal Shielded Flanged Bearing Chrome Steel ABEC-1(P0) Z2 Deep Groove Ball Bearing for Industrial Machine, Power Tools

200mm Ball Screw SFU1605 Diameter 16mm,5mm Pitch with Metal Ball Screw Nut for CNC Machine

For a few examples

Do you mean this lever arm? It is welded.

No the single one, on the right side of youre mark.
Normaly a 3 dof with heave got 2 leverarms each gearbox (motor) with a bridge to get the top frame steable
look at 3 dofs on this site to know what i mean
regards Ad
sorry for the bad english

Maybe this is what you are asking. I have three of these motor assemblies. Below is my initial mid frame design. the U-joint is on top there and this will attach to the seat assembly.


I will upload more pictures of the compete motion rig when I get the geometry figured out. I am now moving on to the software piece, have downloaded Simtoools and am working on the electro mechanical interfaces. I have a Mach3 board, but I don't think I can use it, thinking Arduino Every board.

Looking at youre drawing i think youre making a 3 dof with roll pitch traction loss not heave

Correct-a-mundo. Roll, Pitch and Traction Loss. Where is the best place to put the pivot point for traction loss? Under the driver seat? this might be best for "yaw" for an airplane sim, but it might be better to put it further forward to get the feeling of the rear of the simulated car swinging around. Recommendations?

The pivot point is one of these two ,its youre choice i cant tell.
verry strong motors for youre sim. the drivers get aevery ting to do in a small part of his range

I have completed the build of my Ball Screw based Motor Drive Test Bed. All of the fabricated components have been 3D printed and I have added some additional telemetry for testing.


Above you can see two sets of limit switches, the top set trigger off the extreme limits of the ball screw and are to protect the mechanism. the second set of switches are triggered by a cam on the shaft driving the lever arm. The white lever arm will be attached to a rod end and shaft in turn connected to the seat.

Close up of the Cam driven limit switches.


closeup of the ball screw "emergency" limit switches

Also connected to the main lever drive shaft is a hall sensor potentiometer that will provide actual positional feedback for the lever arm. In the final design these limit switches and pot will not be needed, but may remain, we will see.



Below is a closeup of the position potentiometer:




I will now turn my attention to the software to drive this test bed. I will use an Arduino Mega2560 and Cubexxx' code as a starting point. I am trying to get my head around the time scales. At the limit our Servo motor can spin at 3000 RPM, or 50 RPS, or 20ms per revolution. The motor needs 10,000 pulses per revolution so the time between pulses worst case is 10 uS. If I use the 50KHz timebase for the pulse generations in the Arduino, this is a 20uS base tick. I know there are dividers and multipliers in the AC servo motor controller than can be used to modify the number of pulses per revolution if needed. I will use PTfly Mover for my initial testing. Here the calculation time is 2ms, so a position message is sent to the Arduino at this rate and we interpolate the pulses to give us the motor position we need.

I think it will become more clear once I can see it on the oscilloscope and then in the motion of the motor and mechanism.

I am making some progress with the 3D model of the sim rig, see below, and I have setup the software development and integration environment. I am excited to get to welding the rig, but I have learned the hard way to develop a good 3D model and excersize all the motions before making sparks. This morning I have been working on the seat assembly, adding a wheel and pedals for the fun of it. Next I will turn my attention to the mid frame. I need to decide how adjustable I want to make the rig. the U-joint position will be adjustable, but I am thinking the pedals to seat and wheel will not be adjustable. I want it to be rigid and light.

. Much better to find issues at the design stage, than try to rework a welded rig.

I have now received most of my parts from aliexpress, bearings, ball screws, servo motors etc. and am continuing work on the CAD design. I have moved the traction loss pivot point further forward and have worked out how I want allow for adjustability of the center of gravity. My thought now is to build the chair assembly first, which includes the pedal and steering wheel mounts, and check the CG, then adjust the design of the mid frame accordingly. I am still playing with how high I want the seat to be mounted, this is limited by the position of the pedals as you don't want the front of the rig to hit the ground in a full forward tilt.