3DOF - C1500´s first Sim Seat

OK, i got it. It´s easier than expected.
The next arduino uno is programmed, a test with a 12V fan shows me that it will work.
A 5V fan is ordered.
So update will follow after delivery

Many thanks to @pmvcda for his excellent work and @noorbeast for the information

greetings
Peer

Hi

i got the rig running, but everything in wrong direction.

If i turn the heave slider to the left in FlyPT-Hexapod the rig is moving down
If i turn the heave slider to the left in FlyPT-Mover the rig is moving up.

But both use the same settings.
3d looks also the same and shows the correct movement:

FlyPT-Hexapod:


FlyPT-Mover:


When i use "output swap swapped" it moves yaw instead of heave.

@pmvcda Can you help me please?

Thanks
Peer

hola, me gusta tu trabajo, comencé a pensar en un 3 dof como tú, y comencé a prepararlo, ya tengo smc3 en Arduino y conectado a simtools con lfs en la demostración, hice las pruebas con motores pequeños, funcionó y Lo di para siempre, ahora con tanto tiempo "gratis para el covid-19" pasan las horas mirando proyectos, ayer estuve viendo el tuyo y me gustó la evolución de 3 a 6 dof, soy flojo para hacer un plataforma para el asiento, otra para 2 dof y otra para la pérdida de tracción, cuando con solo dos plataformas, puede obtener más y mejores sensaciones, con un motor Arduino y un motor Monster SHIELD más, (que es la configuración que elegí ) en este momento he preparado dos Arduino hasta que finalmente decido, he visto los gráficos, pero ¿qué medidas tiene su base? Gracias

Hi Daniferias
Sorry, i only speak english and german
Please translate your post
reganrds
Peer

hello, I like your work, I started thinking about a 3 dof like you, and I started preparing it, I already have smc3 on Arduino and connected to simtools with lfs in the demo, I did the tests with small motors, it worked and I gave it for good , now with so much time "free for the covid-19" the hours go by looking at projects, yesterday I was seeing yours, and I liked the evolution from 3 to 6 dof, I am lazy to make a platform for the seat , another for 2 dof, and another for the loss of traction, when with just two platforms, you can get more and better sensations, with an Arduino and a Monster SHIELD motor more, (which is the configuration that I chose) at the moment I have prepared two Arduino until I finally decide, I have seen the graphics, but what measures does your base have ??, thanks

Sorry, in the other post, there are no more characters, (you almost have it ready) and your last question, I can not help you, but I will investigate, thanks

Hi Daniferias
The main-frame has 84cm on the long side and 17cm on the short.
Ist looks like this:


It was OK for the woody-upper-frame cause it wasn´t so "deep".
But now with the steel-upper-frame i think i have to expand the short side for 10-20cm cause in "yaw" the part with the paddles easily touch the rod.
Then i will also be able to reduce rod to ~60cm.

kind regards
Peer

Hello, and thanks for the answer, it is big, in the images it does not give that impression, (it seems smaller), I think that I will not be able to reach those measures, (my wife would take us both out of the house), I will imagine the way to raise it, because I do not have any material at home, what I want to get back to normality to go to work and start building, thanks again and good luck with the modifications.

Hi
the Problem with the wrong direction is found.
I use 180°-hal-poti with 5V output instead of normal potis
This hal-potis have one big disadvantage: you cant change polarity by switching GND and +5V
So they only work correct, wenn you put them in the front of the motor.
I placed them in the backside and so they give back the inverted direction.
No idea how to solve it in hardware.

I optimized the harness of the motors.
its now ~30cm shorter and 6mm² instead of 4mm² between PSU and IBT_2
I also prepared the harness for the super-caps and attached 12v 100A relays to shut off the super-caps

Now it looks like this:


In the first test-movement still without the super-caps i was very positive impressed.
No stucking or overcurrent.
It will hopefully work nearly perfect with the assambled super-caps.

I keep you informed

Peer

FU*K
I killed one PSU
I was testing and playing around with the SMC3 software workaround for the inverted problem, so I disassembled the arduino.
Unfortunately I accidentally put the ardunio on the fan of the PSU.
Because of this the PSU got no more ventilation and smoked up.


New PSU is ordered and will be delivered hopefully today

regards
Peer

@noorbeast : Yes you are extremely right...
But when it runs again, it´s great

The new PSU was delivered today and modified in 20 minutes
The new harness needed a bit more time.
But now i can switch on the caps to see if there is an improvement or not

And? YES it´s a massive improvement.
It fells like 10-15% faster, but most importantly, it no longer stutters.

Here you can see the different just in SMC3:

Without caps:

WITH caps:


As you can see at the green line (feedback) it is realy much faster and also there is no stuttering anymore.

I´m really happy with it.

But you have to be careful: The CAPS need a lot of current to fill up.
it takes about 10-30 seconds. I can see it at the fan of the PSU.
If its running with normal speed, the cap is fully loaded.

Next step is to improve my mover settings.

c u soon
Peer

If I may, I recommend for your filters in Mover to use the Logistic and the Emalp in cascade.
The Logistic clamp the max value.
The Emalp filter the high noise.

Knowing that would have saved me a lot of time.

Hi Pierre
Thank´s for your help
I hope you can give me some more information and help, how to use these filters.

In the past i saw, that my PC was extremely on its limit.
So i decide to upgrade a little bit

From Xeon 2650 with 8Gb RAM, older graphic card and normal HDD´s
to Ryzen 5 3600, 32GBRam, ASUS TUF Gaming Plus, GTX 1060 and WD_Black SN750 M.2
This makes a lot of fun

Now my PSVR runs without any problems on this PC
Today i made the first "test ride" with Dirt Rally 2.0 in VR !!!
AWESOME

But there is still a lot of work in fine tuning (at the moment it´s more "heavy testing")

I only get movement on accidents, litte jumps or acceleration or breaking brings no movement.
But a rollover is much to much movement.

So @all and especially @Pierre Lalancette please help me

Here are my actual settings in FlyPT-Mover:

You can see: No filters at the moment



Thanks in advance
Peer

First, make sure you have the right dimension of your rig entered into the mover diagrams.

Now, you can go here for a way to calibrate the option Scale in mover. Notice that in Simtools, I do not put my actuator limitor to 70% anymore. I put 100% and reduce it in mover now. You can now use Mover instead of Simtools to test your maximum heave, by grabbing the blue scare in the In/Out value of the Pose window.

6 DOF from scratch

Once you are at the right scale. You decide how much distance each forces can travel. For me, it was trial and error. Sometimes, if you give too much distance to too many forces, you will reach your max distance rapidly, because they add together.

For safety, start with a low Gain (in the Hexapod window) and remove Auto gain. I'm not sure if it can add gain or not. Put it back after to limit big inputs. You add Gain slowly while testing until you get your max.


Filters: There is a manual in the download. Look at it. It gives many good advises.

What you get from Mover is a Value. The filters will influence that value. So, you put it inside a filter and change parameters to influence it. Example with Assetto Corsa:



I have the force sway that stay at 0 for 99 samples and goes to 100 at this sample. My VALUE is inside a EMALP with the parameter 75. It looks the 75 last samples values (0) and this new one (100). It makes a average and you get around 2 or 3. That slows down rapid changes. If you put only 10 as a parameter, it will change fast; 500, it will go slow. And that is the value that is send to your motors. But before, I put it inside another filters (LOGISTIC) that will smoothly clamp the value to 50 with a curve, if the value goes that high. The clipping value depends of the forces send by the game. The parameter 1 affect the shape of the curve.

That's my way of doing it. There must be many others ways, with so many types of filters.

Right away, I would like to say that I never found a satisfactory way to have a good acceleration feeling, and a limited force in crashes. I have to compromise my acceleration so I don't destroy my rig on crashes.

For the suspension, it is a bit weird. I use a Remap (-88;12;0;50). It put my suspension to the middle, but my front and back are not at the same height, so that was my compromise. When the input is -88, the ouput is 0; when the input is 12, the output is 50. At first, the last value was 100, but I reduce it to 50 because the suspensions was making my rig move to strongly.

You can record a run with Mover by pressing the red dot here:



You can then use a Player Source to load the file and drive your Hexapod in Mover with it. Nice way to test. Then use the Graphics Viewer to compare acceleration with crashes, and the results of different filters.



There you go. I hope it will help you.

Pierre.

Hello Pierre
Many thanks for your help
I´m playing arround with the filters and get it even better and better.
But is still try and try... (or error )

Hello all...
I also tried a lot of fans for the wind simulation.
Some 12cm som 14cm case fan, but nothing makes me happy.
Then i found these car-fan on amazon:
They cost only 13€ and lock like this:


Now they look like this and are working perfectly:


I still have some problems with the IBT_2
The overcurrent-protection is very often activatet.
Sometimes it helps to press the stop button, but sometimes i have to reboot the complete rig (all arduino...)

So i´m not really happy with my construction and think about changing from rotation to linear rig.
And here i need your help.
I don´t want to make too much mistakes and so i will ask a lot.

From now on i have to read a lot about ballnut
Has anybody released something like a shopping-list and the needed 3d-printer parts?

kind regards
Peer

@wannabeaflyer2 provided a full plan of his linear actuator. Look out for it. It could give you many ideas.
I did not follow any plan for mine, as there is not much out there. I went with it all along. Asking myself questions. Oh, man, I remember all the questions, all the time, it was nerve breaking.
I think I provided my parts 3D print files in my thread, but I could not find them back. I could look for them if you are interrested.

@noorbeast : Sorry i forgot he link to the fans (i guess they did not work so i added some -- )
--https://www.amazon.de/gp/product/B087FDJ39R
i will make a video soon. But how can i show you the wind in a video?
Maybe witch some cotton strings?


I´m reading a lot about linear actors.
but the most projects use 24V or step motors
My hope is to reuse a lot of the electronic i have.
So the idea is to use some 12v motors again.
i found these here:
--https://www.amazon.de/Elektromotor-Speed-Permanent-Magnet-Generator/dp/B07CH9JZLC
Do you think this will work with arduino->IBT_2 and 12v PSU?
Or do you think it will be better to collect more money and use Thanos AMC-AASD15A with 80ST-M02430 with AASD15A servo driver? (Just linke Peacemaker105?)

I found FlyPT´s and @Peacemaker105 projects really interesting.
And both share their 3D files witch will help me a lot.

kind regards
Peer