New 6dof Stewart platform for a school project

Hi everyone,

My name is Robin, I'm a 25yrs old student from France. My friend Arnaud and I are building a miniature 6 dof motion simulator for rFactor for a engeniering school project. Isn't it cool? We'll use servos as actuactors. Project have to be completed in 6 weeks. Two types of control will be implemented: manual control with 3 joysticks for the 6 axis. We'll use this for joystick:
http://www.conrad.fr/potentiometre_de_joystick_p_18925_18989_657529_649518_FAS#searchItem

We'll also Control/ simulation for rFactor. We'll use Yoda to extract forcefeedback values to our own simulator program written in C++. We intend to use this to control the servos:
http://www.robotshop.ca/devantech-sd84-servo-controller.html

This controler allows us to control the servos by computer with C++ so it should be easy to use it with x-sim.
Project has begun 3 days ago, I'll try to give updates, pics as it progress.

As it is my first post in this forum, don't hesitate to tell me if something's wrong.

Comments, questions, advices are more than welcome!

Best regards,

Robin

Hey Robin,

I wonder if it wouldnt be better to code your 6axis control software to accept Rx,Ry,Rz,Tx,Ty,Tz data via the Force-Profiler for input, instead of Yoda FF effects.

That way you guys could use XSim2 to make game movement profiles, which would save you time, and you wouldnt be limited to rFactor only. Your work would be to make your app able to process the (already scaled and filtered) Rx,Ry,Rz,Tx,Ty,Tz data from the Profiler, as well as interfacing the servo controller you want to use.

Your code with the basic (stewart) motion algorithm could be standanlone still, but you wouldnt have to care about game interfacing, which would be pretty much work for 6 weeks anyway imho. What do you think?

Have fun with your project :thbup:
Christian

Hi egoexpress thanks so much for your advice. I didn't know that Force Profiler outputs directly Rx,Ry,Rz,Tx,Ty,Tz. I'm not used to x-sim yet, I'll be doing tryouts next week. I'll try to extract the data to a c++ interface.
What you say is exactly how we will do it. Thanks for saving us time !

We'll write the program that handle the movements of the platform taking Rx,Ry,Rz,Tx,Ty,Tz as inputs. I've read an excellent topic from planmix which explains how to do it. I think it's pretty doable in 6 weeks.

Thanks again for your reply,

Robin

Make sure to use XSim2 for 6DOF, as the older XSim1 does not support more than 2 axis.

Regards

Btw, just in case you dont need 84 servo ports, the Polulu controller could be a less expensive alternative.
http://www.x-simulator.de/wiki/Servo_Ta ... _mboRacing

Hi ego tks again for your advices. I'll begin the tests with x-sim2 on the beginning of the week. As we have to wait one week or two for the goods to arrive, we'll have time to make our first steps. The main goal is to export Rx,Ry,Rz,Tx,Ty,Tz to a c++ program. If this works, it will be a great achievement for us.

We need the SD84 because it has analogue inputs for the joysticks (we need 6 AD channels) and some digital out to drive indicators led (we need 2 channels digital out). As the school is paying for the project and our teacher said not to worry about the prices we can easily afford it. Plus SD84 is well documented with sample codes and everything.
I think we may even be able to afford very fast hitec servos so the results should be impressive (if the program is written correctly though). Link to the servos:
http://www.robotshop.ca/hitec-hs-6965hb-servo-motor.html
These are expensive and absolutely not necessary but my teacher insisted on buying high grade stuff. Some teacher spend much more than 1000EUR for a single sensor in this kind of project so I think that's why. It's the first time I make a diy project within these conditions lol!

Tks again ego for your support, it is very appreciated. I'll maybe have some questions later on how to export datas to VisualC++, but I'll try to follow the homepage tutorial and/or read some more documentation first.

Regards,

Robin

Hi all, some quick updates of the project:

This week was very busy because we had to setup the computers, the software, network... from zero. And we also had to order all the stuff needed from 5 different suppliers.
Good news is Arnaud and I found a solution to calculate the degrees of the servos given the position of the platform. We haven't tested the solution yet but I think it's correct. I've read that some 6dof builders were interested by these equations so we shall release them as soon as we finish testing.

I've done my first steps with x-sim2 and I find it awesome! I was able to create some axis but I haven't had enough time yet to export the datas to a c++ program.

That's it for today,
good building everyone!

For a better understanding of the project, here is the general architecture of the 6 dof platform (subject to changes):

looks right. Though the Force-Sender is able to send real telemetric data, gauge and joystic input data to the Force-Profiler as well, and not just ForceFeedback data only.

Regards

Thanks for the correction, I'm such a noobie with xsim :lol: .

Hi everyone, I haven't gave updates of the project for a while so today I've decided to post the software we are writing for the platform.
This software do:

*allow user to design and edit platform's dimensions
*given 6 inputs (pitch, yaw ..) calculate the position of the platform and the angular position of each servos
*draw an openGL 3D view of the platform
-> while 3D window active use numpad 13456789 to move the camera F1 to go fullscreen.
*offers various limiters to limit the movements of the platform.
*works only with windows.

and some more things...
be patient with the optimise button.

The software is still in developpement so keep that in mind while looking at it. In the next two incoming weeks I will try to link the 6dof software to force profiler and use it with xsim. By the way I have a question about that: what is the refresh rate of the datas from force profiler if we use ShareData technique?

Hope I've made my post a little more interesting,

Regards,

Robin

deleted

Hello,

Some feedbacks on the software would be welcome. At first we designed it for our own use but I think its design option would make this software a confortable tool for anyone who desire to build a small 6 dof platform driven by rc servos. Please tell us if we are going to the right direction here.

Regards,

Robin

Same thing occurs to me:



I have already VB6 libraries on my computer and I supppose that's the reason I get different error message from RaceRay...

Can you make a proper installation package for your 6DOF program?


Regards, Thanos

Hi Robin thanks for posting your software .. a lot of guys here would be very interested in this as it once again takes the hard part out of Motion simmimg ... i guess its a commomn problem because i too cannot get the file to run on my XP Pro Machin. This error message comes up so will wait until a fix is available and try again Cheers

Hi guys, thank you for telling me this. I hadn't hit the release option of my compiler (Visual Studio C++ 2008) ha ha ha! Now I get an unresolved link error because of the libraries I used to make the 3d drawings. I'll fix this asap.

See ya

Robin

the 6 dof software -> find new version page 4 of this post

Thanks!

Yes, it appears to be working now. I can open the 3D simulation windows and move the model by the sliders...

But, you should lower the limits and probably allow for alternative output like the Pololu mini servo8 ...


Regards, Thanos