6 DoF platform pre-design advice

Hello all,

I've been considering my 6 dof DIY build project since I joined the community 6 weeks ago. And I'm really overwhelmed wiht the bounty of knowledge I found here.

So far, reading a lot about Stewart Platfomrs, writing inverse kinematics python script to calculate motion ranges. And then about trajectory generation, servomotors, gearboxes, ball screws, etc etc.

I do have access to mechatronic devices but servomotors and their drivers are not cheap in a bundle of 6. Even less when you add up the reducers and the PSUs in case I'll use 24/48V system.

Roughly I plan for an initial version with rotary actuators, leaving ballscrews for later. I reckon a 'linear stroke' of around 300 mm, meaning the arms around 150 mm. Then enough torque to allow each actuator to move 1/6th of the total weight at around 3g max, and a maximum linear speed of 2000 mm/s max.

Then I wonder wether is best to place the cdg of the platform. I don't see actually a revolve-symmetric platform ( i.e. all angles and legths equal) , as long as I plan to have a sort of cockpit shaped platform.

I will post geometry as soon as my work allow me to complete first design round, jus to seek for advice.

But for the moment I have no idea if I'm going straight to a wall, or my thoughts so far are sensible. I don't know of the max/min motion ranges I should aim for. I have been reading FlyPt web and I like the motion cue classic filter algorithms proposed there. They talk about simulating, but I don't understand how to do so. I wonder whether could get some tips here on how to simulate with FlyPt or SIMTOOLS (don't own a licence yet). If necessary I will reead the 34 pages long thread I saw.....

I seriously consider to build first a 1/6 scale platform with RC cheap servos to understand whether my design assumptions are correct, at least in terms of geometry. And then use FlyPt to test.

I wonder whether there's a way to record from a static version of a run with, say Assetto Corsa (A software I'm familiar with), all the cues given to a hypothetical 6 dof platform. Meaning: to record in a file all the platform poses in function of time given by a short (or long) run. That will help me, through a binning algorithm to better calculate design parameters of my platform.

I'd appreciate any help from knowledgeable people here.

Of course will continue reading most of the posts here .

Build a model 6dof first using rc servo motors

You will learn a lot and get in the journey sooner

i skipped the 6dof mini model and do it directly ;D i go for rotary as well and its enough for me at themoment. ballscrew would be nice but i dont see any advantage in VR because less is more in VR i think:

take a look at my rig building threat:
https://www.xsimulator.net/community/threads/just-started-my-6dof-rig.17437/

Thanks for nice replies!

Sorry for late reply. As part of my job I was in Italy to an interesting automation show, where btw I saw a couple of nice driving simulators, using top notch stuff.
Will hava a look to your thread, for sure.

Just gave a quick look to your thread. Really nice job!
I missed wich kind of board or other electronics you use to drive the DC motors. Was the same equipment for the 140W first set of motors than you used when doubling the power to 240W? Thanks for sharing

i use the Sabertooth 2x32 boards (3 of them) + 3 Arduino Boards

also 3x 750Watt PSU so its sufficient for the motor upgrade, i just swapped them out and tweak the SMC3 Settings.

of course a ballscrew simulator is better but needs a lot more space and other stuff, but i want to keep it simple in the first place and a bit more compact.

Thanks for replying with those data. I do appreciate,
Well, yes, ball screw can probably be better. I work for an automation company and have sold some of our best servos for ball screw used on professional simulators.
I dream of buiding a driving rig with them, but aside of the space you mention they are way too expensive. Nothing to do with wrom gear reducers and DC motors. So in my private life, I will start slowly and cheaper.....learniing from everybody here