Showroom Grigory's 6DOF #2 -- compact and no linear actuators, 220V servos

This will be my second project-description thread, but actually dedicated to my third motion sim project.

The first one was this 6dof with DIY linear acuators: https://www.xsimulator.net/community/threads/grigory’s-6dof-24v-dc.8908/. It worked fine, but due to its size was only suitable for a weekend house garage, not the city appartment.

The second attempt was this seat mover with a traction loss:

I actually liked it very much. To me, the pressure on the hands and legs that is felt when the seat moves against stationary wheel and pedals is maybe even more immersive than full 6dof movement (especially with the four buttkickers). This sim was also compact enough to put it in the urban man cave

The drawbacks were only the lack of heave and not so good compatibility with flight sims. I almost exclusively drive, but there is some flying once in a while.

So, this time I decided to build a third simulator version that would be 6dof, but considerably smaller than the initial one.

My general thoughts on this new project are as follows:

1) I was not able to figure out how to make a linear acuator-based design materially smaller than my first rig, so I would be giving up linear actuators and going to lever+rod solution

2) the reason I used linear acutators in the initial 6dof was the extreme backlash in my wormdrive gearboxes, so using the same motors is not an option

3) the motors that I used for the seat mover were NPC-64T, which I had available after some other earlier project. These motors are very powerful and have gearboxes with very little backlash.



They are widely used for robot combat, so they should be quite reliable.

For my new 6dof, I plan to use NPC-74T, the sligtly more powerful version of this motor with the same gearbox: http://www.robotmarketplace.com/products/0-t74.html

4) the cockpit of the first 6dof was made of steel and it was definitely way too heavy. Since I now have a TIG welder, the new cockpit will be made of aluminum tubing.


Here's the design that I have in mind:







The moving platform together with three monitors, accuforce wheel, fanatec pedals, shifter and a seat with a 70 kg driver will weigh about 130 kg.

Levers will be 150mm. If my calculations are correct, each motor will be delivering around 30 kg of force at the end of the lever when using the maximum continious current of Sabertooth 60x2 and about 70 kg of force when taking Sabertooths peak maximum of 120A. The smaller texture will be handled by the buttckickers, so these forces should be enough, I hope.

I also plan to use my own 6DOF software between Simtools and the simulator itself -- to allow adjustment of rotation centers for each DOF.

I already made the prototype in VB.Net, using this logic: Grigory’s 6DOF 24V DC

This prototype accepts the required pose, calculates the lever angles and shows positions of the attachment points on the moving platform in relation to the motors. Here is an example showing effects of traction loss angle.

Neutral position:


10° traction loss -- it is seen that rotation is happening around the axis that is somehwere close to the feet, not around the center of the moving frame.


Some mechanical works are already completed as well. I'll put that in the next post

Nice work . For your situation, did you ever think of building a G-Seat? I must sound like a broken record , but they are very good and can be super compact and quiet.

Really nice!Looking forward to see it moving!
Is somewhere a description of your seat mover with the traction loss?I am trying to build one in a couple of months and I collect information...

Actually yes

I’m thinking of adding it to this sim after the rig itself is made. The problem is that my seat has very thin padding and I have not yet figured out how to fit the flaps without sacrificing comfortable seating position.

And I really like your idea of taking the motors away from the moving platform and using cables.

By the way, do you think the “heave” gseat flaps are needed for the 6dof?

Thanks!

No, I did not describe it anywhere, since it was prettty straightforward: steel welded frame,
unijoint from a Russian 4x4 steering,
industrial cart wheels attached to the back of the traction loss platform and rolling on the aluminum plates
the front traction loss joint is again an industrial cart caster wheel with the wheel itself removed.
The monitors are stationary, not mounted to any of the moving parts.

All three motors are NPC-64T, but that is probably way excessive. I just happened to have these.

I am using sabertooth 60x2 with kangaroos, 1024ppr encoders and contact limit switches.

The power comes from two SDR-960-24 units and two small 12v batteries in series.

There are four buttkickers with Emotiva A500 amplifier and the Accuforce steering wheel — running with simcommander/simvibe

I can share the dimensions as well, but I do not have them at hand right now.

When you are already sitting in a 6DOF, I do not think it is needed for cars, but an extended G-system for heave with Aircraft, spacecraft and roller coasters etc would be useful . Just finished watching 'Ready Player One' - were does it end.. .

Thank you very much for the information.If you find sometime the dimensions and share them, would be great!

Remember, I'm not using 'flaps' in my G-System, just the vest, harness and head bowden driven systems.

Oh, I missed that. What was the reason to replace flaps with a vest? Does the vest give a more natural feel?

It was all to test a few ideas that I had and my current rig made it easy to implement.

Here is some further progress.





Mechanical part is mostly ready, all painted and relocated to the city appartment for final assembly.

I hope to finish wiring in the next couple of days.

I am using Sabertooth 2x60, which require baterries to absorb regen currents. The problem is that Sabertooths can still be getting power from these batteries when the power supplies are off, which I do not like. It would certainly be better to be able to totally cut the power.

My three Sabertooths can deliver 360A continuosly and twice as much for short periods of time, so it was quite challenging to find a convenient switch that could handle that. Here is what I am going to use:

https://www.gruner.de/ablage/med_00000234_1502371891_en750.pdf

This is an automotive relay, which is heavily used in Russia by 4x4 enthusiasts for their winches. Its rated for continous 16V 250A and bursts up to 3000A, and people have succesfully operated it at 24V. I hope that it will be strong enough for my application.

Sabertooth manufacturer also advises against totally disconnecting the batteries, as Sabertooths may get damaged if the motors are moved and there is nowhere to dump regenerative current. I will be using a diode connected in parallel with my relay and allowing feeding current back to the batteries even when the relay is off. That should also allow using that relay as an emergency switch.

Change of plans: I had several problems with my initial idea of using NPC-74T, so I decided to make a radical move and experiment with using 220V servo motors and planetary gearboxes. On paper, this would allow a speed of about 1600 mm/s, nominal 50 kg force at the end of 150mm lever with short peaks of 150 kg of force. Should be plenty for my platform having 35 kg without the driver.

The motors arrived today, the gearboxes should be here by the first week of September, so I hope to start testing in 2-3 weeks from now.

I also decided to follow SeatTimes' advice and add a cable-actuated G-seat. I created a separate thread for this: https://www.xsimulator.net/community/threads/cable-driven-gseat-to-be-mounted-on-6dof.12314/

Really love this compact design @Grigory. It reminds me of the Full Motion Dynamics rig. Can you outline the difficulties with the 24volt setup? Was it confined to the sabertooth regen or was there other power issues? The lever arms do look longer than anything I’ve seen on a DC voltage setup.

Thanks! Indeed, I was inspired by Full Motion Dynamics design.

Actually, I think that the initial 24V design could work and the problems were not that unsolvable.

The first problem was with the temperature. These motors are really powerful, they had absolutely no problem moving the rig even with these 150mm levers. The other side of this power is, however, the huge current requirement and, hence, tendency to quickly get very hot. Also, the wires initially installed on these motors are not thick enough. Making the levers shorter helped to decrease the temperature and I believe that the issue could have been totally removed if I replaced the motor wires and added some fans to cool the motors.

The second problem was with control tuning. I was using Kangaroo x2's as PID controllers and for some reason they just refused to properly auto-tune. The tuning process completed without issues, but then bad ocscilations occured at some parts of the lever travel. I tried setting the coefficients manually, which helped a little, but could not fully solve the problem. I spent couple of weeks trying to fix that -- without success. Probably, I would have eventually found the cause of the problem or maybe I could have switched to SMC3.


The reason why I decided to switch to 220V servos is simply that I have recently re-fitted my lathe's spindle with such 220V servo motor and I really liked its power and convenience of control. In addition, I've found nice planetary gearboxes with less than 0.1° backlash.

Hello. You've mentioned that you're building the rig smaller to fit in apartment room - how does it sticks to current requirement for 6x AC servo motors? You can't put them all on a single outlet/breaker, or do you ?

Sure, that is one of the issues that would need to be handled. I have six motors with nominal current of 4A each. The power outlets in my appartment are rated 16A and each room is connected to a separate 16A breaker. Based on that the plan is to divide the motors in two groups and use cable extenders to connect these groups to the outlets in two different rooms. If needed, I can split them in three groups as well.

My servos do not fit in the base that I previously made for the 24v motors, so I need to build a new one:

Now my rig is disassembled and I am starting to put together the new base for 220v servos: