6DOF parts back from paint!

Have had a new build in the "planning" stage for quite a while, the main structural parts got back from paint today so now starting the build.
It's a bit of a different approach so it will be interesting to see how it all pans out.

Disclosure: This is just a tech demonstrator for now, still quite a way's from a commercial product.



Base plate after tapping

Picked up from powder-coaters



Motors ready for electronics



Under construction.....

Well the motors to move the rig certainly look innovative, good luck .

Linear motors?
Something I looked before making my rig.
But going that way, I would still be in the actuator build fase making coils.

Good luck with the project! Those are really fast. Are they strong enough?

Thanks! @pmvcda RE force output, that's the hope, We are hoping to be able to move a ~250lbs load for the demonstrator. It's using a combination of linear motors and pneumatics (to offset "empty" weight.) One of the drawbacks of linear motors is that they burn power holding a load at rest, so the concept is to use the pnumatics to offset the weight of the platform (and the rider) to a neutral point and then use the linear motors for actually moving things about.

Now of course it remains to be seen just how well this will work!


A few more pics from this week:


Never build something without RGB.....



Base assembled without legs.


Legs going on


We built a thing...



Now.... Does it work.....?

Video!

Edit:
What kind of resolution do you expect?
How many bits for positioning on the firmware/controller?

We still have to get all the PIDs dialed in but positional accuracy after filtering should be somewhere in the neighborhood of ±7μm.

Hoping to get a decent video out sometime next week

Still lots of tuning to go, but she moves!

Hi Walker YYZ

I am quite impressed by your work!! Are you doing this private? or maybe in a university or commercially in a company?

Second question: What you are doing: is it more rocket science or is it doable DIY?

Will you share the details of your work?

Wish you good luck for your project!
Wickie

Hey Peter, Thanks!

Were doing this project basically as an experiment of a possible application of linear motors that our company designs/builds. Most of our business is custom OEM stuff, so something cool to demonstrate their capability was the idea here.

As for the DIY"ness" of this, there are some pretty cleaver people out there, but I think for a single person trying to do this from scratch would take quite a bit.

Happy to share on the rest of it however!

As mentioned the motors are ours, they are a "sized up" version of a sort of standard design we have, however in the long run I suspect something a bit larger would be ideal. Our motors are designed with fully integrated drivers (they live on the motors themselves), so there are no external/3rd party motor controllers being used here. The motors are running over RS485 back to our controller.
The motors are 24V and are powered directly off of a LiFePO4 power pack. In this case we are actually running each individual motor through a pretty high accuracy bi-directional current shunt/power distribution rail. In theory we should be able to do some power recapture during certain braking movements, so we wanted to be able to have a way to capture some data on just how much juice we may be able to put back into the battery bank.

The mechanicals are all pretty simple COTS parts from McMaster and a bunch of waterjet plate. Not quite as elegant as some other designs, but again at this point its just a development platform/tech demonstrator.

It should be pretty fast when its done, both with respects to motor acceleration, but also system latency. This will allow it to also run some pretty high frequency vibrations (Khz range) layered on-top of any gross movements that may be commanded, so should be no need for a but kicker. Its also pretty quiet.

I expect a single 'upscaled' linear motor would have already been load/speed tested before building more and moving the project on. Exciting.. Look forward to see it in action.

Was a busy Christmas season, so finally just getting back to it. Still waiting on parts for the chair.

HA!
Yes that's probably what we should have done, on this. The original project that kicked it off wasn't necessarily intended to move a person around however, more of a scaled down 6DOF for stabilizing other equipment onboard a moving vehicle. That being said after starting to put the first one together we had a bit of "feature creep" and the "scale model" ended up being a bit larger than originally intended.

Yea I figured that and have seen 'feature creep' destroy many a project. Experience can be a tough master, especially in the real commercial world were your employment often depends on the Projects Success and keeping to budget and schedule.

Maybe you could try Mover, and play with high and low frequencies.
Really curious to see you pushing for it.
Great work!

Just caught your video - nice work . Will be Interesting to see how much operational load they can support. Edit - Oh there is another video with someone siting on the platform - nice. What is the advantage of this tech over a normal servo motor setup?

Those neon strips look cool! Good job.

Sorry for the delay in response. There are a few differences:

-They are very fast both with respects to max speed but also with respects to latency. A lot of that speed comes from these being our own motor drivers, and our own controllers, but loosing all the mechanical problems of gears/screws etc speeds things up tremendously. Our current "motion to photon" time "would be" in the 200µs range if there was a headset that was fast enough. So currently we can command a movement, send it to the FPGA, have the FPGA calculate the IKM, send commands to the actuators, have the actuators interpret those commands, power the coils, have the coils come up to full field strength and start moving, + resolve the new platform position and pass it back to the game environment at >5Khz. Most of where this is useful is in providing very smooth motion, IE no bucking/shaking/etc (unless you want there to be!)

-Having very little mechanical tomfoolery means no backlash issues and "in theory" fewer mechanical issues over the service life.

-The motors can generate some pretty aggressive vibrations that you would have a hard time doing safely with more traditional motors.

-Not necessarily unique to us, but being low voltage DC means you can offload surge power draws to a battery bank, meaning you can run this thing on a standard wall outlet. Also we can regenerate some spent power and send it back to the batteries depending on what movements are going on.

-They are pretty quiet

-Being force controlled motors means we can detect user weight and user weight distribution, so for a bespoke application shifting the users weight side to side can be interpreted as a user input. IE say you have an experience were you are flying a dragon or riding a dolphin or something, the guest doesn't need a joystick, just their butt in a saddle/etc.

Also an updated Video:

Very impressive!

So... I want to see a test with Mover: https://www.xsimulator.net/community/threads/flypt-mover.13464/

I want one of those!

This could be good for a moto simulations, where we need to tilt the moto...

@WalkerYYJ - Extremely impressive - especially where it launches that guy. Looks like he gets air? That would imply a 2G acceleration? And those vibrations are nuts! It looks like you could feel every rock in a dirt road with that!

The 260W number (IIRC) - that’s per actuator or the full machine? Was that a resting value? For instance, what is the power draw when launching folks into the air?

It’s air preloaded isn’t it? You said before you were just going to float the empty frame/seat/etc so pilot weight is additional and supported by the linear motors or do you up the preload to compensate somewhat for different weight pilots? Do you have a max weight the platform can support or curves of platform weight vs. acceleration?

One last question is it looked like maybe there is some bounce to the linear motor and pneumatics combination? Am I seeing that right? It would seem like with the vibration demo those linear motors can move fast. Are they actively correcting for bounce in the pneumatics?

You mention selling these as a product. Are you going to post details where we could build our own possibly sourcing parts from you or others, would you be selling subassemblies/components, or just complete rigs?

Very impressive! It also seems pretty compact.