Showroom Compact AC Servo 6DOF Rotary Stewart platform

This is taking longer than expected, mainly because I have a motorbike & car project running as well... here's my progress.
The base was based on Smittys' shape, but went with more or less your arm design because I couldn't justify CNC costs vs making them myself.

I also saw one of the guys on your thread put the servos on the base which I also liked, and plan to do the same.







I'm considering placing all the servo control gear underneath but for problems with any maintenance in the future so not decided on that one yet.

So thanks again Chris, all made possible by your generosity in providing decent open source info.
Regards
GB

OK will do, I did wonder whether I should or not, just didn't want to steal Chris' thunder.

nice work my dude!

Hi Chris, Got my rig powered up, and am trying to config it. Can't do too much yet as I haven't wired the E-stop buttons yet, and feel after re-reading your posts, you make a point of fitting them, and I know I really should.

With respect to the numbering of the motors, I've just installed Simtools v3, and the image in the setup window shows motor No.1 to be approx 1 o'clock , with motor No.2 at 3 o'clock, No.3 at 5 o'clock, and so on. Is this how you laid yours out? Looking from the rear, forward.

Then these should marry up to the motor numbers on the control circuit board?

I have never got SimTools v3 to work, I was given early access back in the day but could not figure it out and more or less was offended that it did not work backwards compatible with my old game plugins, so I abandoned even trying. From your description no, the motors in simtools do not align with the how they did it in V2, where v2 was sending generic scaled x,y,z, rx,ry,rz values to the PCB, the PCB then was doing the inverse kinematics to figure out where the motors should go. Do we have any V3 users in here??? how can we emulate v2 outputs in v3?

Thanks for the speedy reply.
Oh, I found V3 to be reasonably intuative. The 3D virtualisation suggests all is moving correctly, but the actual base is not moving in recognisable directions.

But Ok, I still have V2 installed, I'll go back to that for initialisation.
Would you be kind enough to advise what motor the No.1 IDC plug (on the circuit board) connects to, on the motor on the basebase? and IDC No.2 so I can deduce direction. I'm assuming they'll be in order?
I need to start debugging from somewhere...this is a nice easy step.

front right is 1, then clockwise till 6

Cracker. I'll get on it.

Ok, all sorted. Thank you for that.

I'm having a problem getting any pitch output from Assetto Corsa, it seems to output pitch when running over kerbs. But almost nothing for acceleration. Id like the chair to tilt back somewhat when accelerating.
Is this just how Assetto is? or can I edit files?

you need to blend in the idea of lateral /longitudinal forces into the pitch/roll axis, here is an example setup, look at axis 4/5, i add/blend the sway and surge forces into the pitch and roll axis.

Hello @knaufinator. Thank you so much for gifting your design and support to this community. Of all of the designs I have seen on this site so far, yours is my favorite. I'm considering jumping in and making something heavily influenced by your design. I had a few questions that will help me decide.

In several posts and videos you talked and demoed just how overpowered your rig is (and that its not even working at its full power/potential). With that in mind, I have to trains of though I'd love to hear your perspective on.

Option 1: Save Cost
Q: If you were looking for a way to knock down the hardware cost by about 2K or more, it sounds like the project could afford to get away with a lower wattage motor and controller. Perhaps not something as low as 300 watts (like many other systems out there) but something lower than your current 750watt motors (perhaps tow winch motor? and lower wattage controllers? If you agree, can you offer any suggestions?

Option 2: Use the Extra Power and Make it a 2 Seater
Q: Created a custom frame to allow dual/side-by-side seat arrangement. Modify the frame so that a single seat could slide over to attach a second seat when needed. How far each seat slides would be determined by the two riders combined CG. Perhaps add in 3 gas struts for extra support?)

Just a couple quick final general questions
Q: It seems like the linear actuators are All noiser than your leverage arm design. Iam I correct in believing that any similarly powerful and motion range linear system will Always be noiser than your leverage arm system?

Q: If you were to do this project again with what you learned was over kill, not required or with the new motors and controllers available now (5 years later), would you do anything noteable differently?

Thank you again!

Good day, tough questions,
1) it really comes down to subjective compromises, I would not want to reduce the power output, my quest was/is fastest/strongest. I dont think there is 2k worth of savings and still having the same capability..... my rig is good because its overengineered. AC servos are essential in this as well, they speak the language that my PCB was designed for relative positioning

2) you COULD do this, but the other part of the equation from above is to make the chassis as Light as possible, humans are heavy, mechanisms for seats heavy and will induce more slop in the movement. light and ridged is the goal, but again, ... subjective to the users goal.....

3) yes, my rig is silent, the hum of the ac servo drivers is louder than the movement. it also has similar articulation as the linear actuator versions., ~11 inches of vertical heave movement,

4) I am trying to restart the project in my spare time, I am making the controller faster and more precise, as well as some other safety related PCB changes. but summer is around the corner and i am knee deep in building a race car, so we will see how much progress i make.

I appreciate your answering all my questions. I thought you might find this new DIY 6dof project. While not nearly as powerful, he did pull it off for under $1000, which is impressive in it's own way.

Thanks again @knaufinator. If you had to make a estimated guess, what do you think your max dynamic payload would be? I'm curious if you believe that your current setup would be able to support a total weight of up to 550lbs (two reasonable sized riders + all peripherals/seats moving and then coming to a quick stop).

I estimate the rig—including all gear (monitor, wheel, seat)—to weigh around 100 lbs, with my own weight adding roughly another 200 lbs. I’ve safely tested riders up to about 300 lbs, although at that point, the primary issue I noticed was that the base plate began to flex during heavy maneuvers.

This plate is a 1/2-inch thick steel disk with a 30-inch diameter, and under load, its flex becomes noticeable. While it might be possible to support an additional 200 lbs, I wouldn’t recommend doing so without reinforcing the structure. Key improvements would include:

• Reinforcing Structural Components: Use thicker steel tubes or plates.
  
  
• Upgrading Hardware: Implement larger heim joints.
  
  
• Adjusting the Drive System: Employ different gearing along with larger swing arms or shorter arms to reduce force on the motors.
  
  
• Increasing Motor Capacity: Consider stepping up to a larger AC servo, such as a 90 or even 130 series, to achieve a stiffer, more robust setup.
  

Ultimately, weight is the main enemy here—the lighter the rig, the higher the dynamic frequencies it can safely achieve.

Wow thats shocking that it was bending that 1/2" steel base plate. Wouldnt expect that. I would assume it was happening on the vertical hard mounts attached to it. Really appreciate the inside info so that I don't head down the wrong road.

@knaufinator I was planning on using 40 x 80 aluminum extrusion for most of the framing of the upper plate and cockpit structure. Light weight and versatile. Do you think I should go larger for up to 550lbs? (worse case scenario would be Rally racing)

its been used, and should work if you construct it ridged enough, the actual steel chassis though on mine only weighs around 25 pounds. the cost savings of just welding it from steel tube was not insignificant. ... though i did have to learn how to.

A "sandwich" plate made out of two 1/4" plates with some rectangular pipes in between is much stiffer than a single 1/2" plate. Of course, welding them together is much more work but the sandwich approach can help to reduce overall weight.