6dof motion platform ‘lite’

Hello everyone,

First I'd like to say that I have very similar thread on another sim forum.
I hope that this is OK, as I would like to get feedback and suggestions from everyone who is willing to help.

I am looking into building a 6dof motion platform for use with VR headset (D2 on the way) for racing and flying simulations, (and hopefully Star Citizen), and would like to hear your thoughts on 6dof simulator with levers.

The simulator needs to be able to lift about 100Kg of weight, and I’d like to have 7-10 cm of movement.
I have been looking for existing 6dof concepts, and so far merchan-e concept with 6 wipers seems like the most cost effective way to go

http://www.xsimulator.net/community/threads/prototype-6dof.5045/



I have seen a lot of simulators with motor arm connected directly to the push rods and the platform.
According to ebay:
“it is NOT recommended to put weight directly on a gear motor's shaft! An excessive radial bearing load may damage a motor's gear box quickly! It is always safe to isolate the load by introducing a pulley or sprocket system."
http://www.ebay.com/gds/How-to-Buy-the- ... 002/g.html

What do you guys make of it?

The way I see it,
Wiper motor without lever:

Wiper motor with lever


Similar concept has been used a lot with 2dof wiper simulators (seat balancing on a pivot point).

I plan to add 6 levers (seesaws), one for each motor, with 1 to 3 ratio so motors only have to lift one third of the platform weight.




This is what the levers and wipers might look like;


Would it be better for motors if the weight of the platform is balanced by a counterweight on the longer side of the seesaw?
How would the change of momentum affect motors?



Here is a rough model of the entire simulator:
(Dimensions and scales are not accurate)



I am thinking about symmetrical design, and positioning wipers and levers in line with pushrod angle.


however I am not sure how great would be the benefit.
Levers and pushrods should be symmetrical (I'm not so worried about motor positions)

Engines:
I really like characteristics of these worm drive motors:
http://www.motiondynamics.com.au/worm-d ... orque.html
Except shipping, 6 motors cost ~$650, shipping cost $360- 430 for UK
I’d be very happy to hear experiences from someone who has these motors.
With 1/3 sesaw ratio, 6 of these motors would have to move about 33.3Kg (100Kg/3)
Motors would lose 2/3 od their effective speed to about 50-60 RPM, which I thing should work great.
Has anyone tried running these at 24V?
How much power would they use, and 24V and how hot do they get? how long can these motors run in one go?

To power these I’d be getting one of IBM server PSU for 12V (or two for 24V)
These PSU’s are built to last and particular model outputs 147A from 12V.
http://www.ebay.co.uk/itm/IBM-BladeCent ... 5afdd59381

Looking forward to your thoughts.

Thanks,
Tino.

These low powered DC 6DOFs are all new territory at the moment, so there is not allot of real data to go on - apart from it may not work. I'm soon to load test the same motors but with steel geared 60:1 gearboxes (allot stronger than the stock nylon gears) which will basically give me the same output torque as your levers, but likely more movement. I'm hoping this will be able to lift with six motors myself (I'm not that heavy) and a lightweight frame (myself/frame should come in well under 100Kg). The counterweights should work, but will add mass which could slow down the response of the sim. I'm currently running these motors around 13VDC and have not had a problem (The full specs for the motors are on the motion dynamics site)

Hi SeatTime,

Sounds great! I'd love to hear how these motors perform.
I also though about replacing gearboxes, but that is additional cost.
I like levers because they don't add much friction (much less than gearboxes), and they are cheap to make.

Reduced motion is downside of levers, but this motor does 180 rpm, so even if it slows down to one third, 180/3 is 60 rpm which should be pretty good.

@Tino , the idea of see saw is a good idea and certainly worth a try. Give it a run and show us something new.
As for ebay saying its wrong to put the load directly on a motor shaft, well, obviously they are reading from a textbook or have simply no idea. I have to disagree.
Frame design is crucial and you could use the most powerful motors around, but if the setup is poor, its a waste.
Apart from yourself looking at the idea, a zillion DIY sim builders have not done this. Probably because a lever on a motor shaft just works.
Using a seesaw effect to increase load carrying but at the cost of a overall speed, well you might as well just use a higher ratio gearbox on it. Say a 50:1 or 60:1 , giving you about 70-90 rpm or so. Still good speed with a lever on it still works out near 200mm per second or more. Then you have more actual lifting torque.
With the motors , where are you in the world? There may be an equivalent somewhere nearby you. Motioncontrol in the UK has these motors , ebay has a crabpot motor on ebay from a supplier in the US , I think.
And if you plan to use 24 v, use 24v motors, dont overdrive a 12v with 24v, and also the opposite, dont use a 24v motor on 12v.
@merchan-e has a good idea going and has proved that 240v motors are not needed for a 6DOF, as well as @Historiker , as he is building a 6DOF with 12v boat winches.
Another idea for the seat frame, would be to make it like a bath tub, so the motors connect higher up above centre of gravity. And you sit down in it. It has been done like this and if I find the picture later, I will post it up. Having motors underneath is fine, but if you can go higher, it will help the sim react a lot better. Even moving motor connections from right underneath the seat to even 180mm up the side does wonders to improve performance. (desk racer design proved this)
The motion dynamics motors are excellent, as the supplier is 5 mins drive from me, so I use them, however the ebay crabpot and motion control in the UK are very similar motors. Use whats local to you. @Roadster2 , can you advise on motors?
Otherwise its a plan you have to start with.
Cheers, David.

@bsft Thanks, I'm glad you like the idea.
I live in London, UK and it would be very helpful if you or someone recommend worm drive motors around 150-200w and 20 Nm torque available here.
I agree that good frame design is crucial. I quite like the frame design from Full Motion Dynamics, I'm thinking of building something similar.
I'm hoping that adding 6 pivot points will increase the stability and safety of the platform. It won't cost much and will make me happy

To me using seesaw in 6dof makes sense just like using pivot point in 2dof simulators.
In theory you could make 2dof by mounting seat directly to a motor shaft and 2 motors would do the job, but pivot point makes it possible to use much smaller engines, adding levers seems like a logical step to make 6dof with smaller motors.

If my math is correct on the other side of the lever I'll have 6 200w 20Nm motors to move about 33Kg of weight.
Motors will have to cover greater distance, so I'll have to see what will be the final length of the motor lever.

I like the idea of connecting motors to the seat above center of gravity .
Do you see any problems with rotating the motors 180 deg so back of the seat is connected one tip of the 'triangle' like in FMD design?
If this is OK then for final design I could have pushrods connected to the seat a bit higher, like:




Regards,
Tino.

you already found the design I was looking for

I havent built a 6DOF, so @Historiker and @merchan-e are best to advise on that.
As for 6 big worm gears moving 33kg, I take it thats each motor as 6 wipers move 90 kg fairly well.

Hi
Can I ask what you intend to use to solve the problem of a cheap 6 dof controller and 6 dof inverse kinematics.
regards Terry

Yeah 6DOF using 200W DC motors maybe way cheaper than a big AC system but it is still not cheap. Six times: motors/power supplies/controllers/driver boards/USB cables/heatsinks/fans/enclosures + 12 rod ends/bearings for levers + steel/aluminium/carbon fibre + welding wire + race seat + race harness + data cable + power cable + computer + steering wheel/pedals + screens/Rift + bass shakers etc - all this adds up and is still a big investment for a first sim were things will likely go wrong.

True, maybe try a small scale model first with RC servos to see if your theory works

Great! I call it 'standing on shoulders of giants' there is a lot of knowledge out there and these forums are a great place to bring best designs.

With lever of 1 to 3 ratio, the weight on the wiper side of lever should be one third, if my understanding of Archimedes is correct , so 100 Kg weight would be reduced to 33Kg and six wipers should easily move that.
I think of levers as less complicated and cheaper version of liner actuators

Levers should work with both faster and less powerful motors (reducing the weight and speed) and with more powerful slower motors, for those you would reverse the lever (motors on the shorter side) workload and speed would be increased but hopefully pivot point on the lever would take a lot of weight.

I thought about going via diy route, however I don't have a lot of experience with micro controllers so to save time and effort I'll be doing a lazy man thing and buy Thanos 6dof electronics.
I do intend to run Xsimulator software and other simulator software on my rig!

I agree simulator hardware is not cheap, however the big reason I decided to build 6dof platform is that I believe it should cost less than a high end PC, about $ 1000-2000

@merchan-e proved that this can be done cheaply, the only thing I'm adding to his design are 6 levers (to reduce workload and speed of 180 rpm wipers) I can make levers myself and I plant to use a lot of cheap used car parts.

Lots of people who are looking into motion simulators already have all the hardware, lots of people even buy or build their own racing or flying cockpits http://www.gtomegaracing.com/gt-omega-art-racing-simulator-cockpit

The only thing extra on top of cockpit like the one above would be:

• Additional metal (cheap if you have mechanical engineer dad with couple of welding machines only problem is that his workshop is in another country).
• These 6 180rpm 200W motors cost (without shipping) $650, and if they only have to move 33Kg then maybe I could even get cheaper motors.
• 1800W server psu is about $30 on ebay
• electronics and wiring might be another 500$ (correct me if I'm wrong)

I think this is not too bad for 6dof motion platform!

Yes, due to space constraints I will probably build a model first.

Please DO NOT run a thanos card, save yourself the hassle and either run 6 x JRKS or an ard and h-bridges like @Historiker .
There is a zillion DIY actuators out there, but yes, they cost more to run and setup than DC motors on their own,
http://www.xsimulator.net/community/threads/linear-actuator-designes.5019/
http://www.xsimulator.net/community/threads/diy-actuators-plans-by-ian-bff-shortened-by-me.5279/

If I had a greater knowledge of micro controllers then that's what I'd be doing, and I will probably do at a later date.
For now there is enough unknown variables in the design so I'd like to get something that should work.

Does anyone here have any news about Star Citizen 6dof simulator support?

I too would recommend a different micro controller purely on price. Three Arduino UNO boards would be 30.00. And if something did happen to one then you would only need to pay 10.00 to replace it. There are three very good sketches for the Arduino right here on XSimulator.net, I have used all three. Mercan-e also used one for his build. They are not at all difficult to learn to use.

Hi Historiker,
I promise I will build my own electronics at a later date (it might be possible to use even smaller engines, perhaps 6 x 100w wiper, so Arduino UNO boards would work great), for my first build (custom 6dof setup where motor movement is inverted because of the lever) this is not something I'm comfortable doing.

Hi @Tino
The only 6 dof diy software that I know of is the Thanos conversion of Robins original inverse kinematic software, which I believe only works with his 1280 card ($450.00), or Ian's BFF software for FSX or X plane.

With regard to the rift, how are you going to compensate for the sim moving and the head tracking?

not trying to be a kill joy, but I have spent a long time head scratching trying to find the answers.

regards Terry

tino...

just for the record this are the pics and the treat about linking two wiper... the thing about wipers is that you can get them almost for free in a car junk yard, the only thing about using two motor per axis insted of one is that you need extra power supply... but again... i had some tests using a nice cheap personal computer power supply that worked nice... so just consider it... check this pics (that are comming from this nice forum, where the fresh ideas are really wellcommed)





you can see the beauty of the two motors been linked using a nice cheap trick.... this guy says that he has not a single problem with his wipers...

and again... used wipers are far cheaper than other options.... but again this is just a sugestion, i have not a real idea about how this works....

you can see the treat here

http://www.xsimulator.net/community/threads/one-motor-or-two.5390/#post-55053

best regards...

cheap is nice.



fer from mexico

Hi Terry,
I'm not worried much about sim movement and head tracking,

DK2 has 6dof IR tracking camera which I will hook up to the simulator platform, where you would normally place monitors.
so head tracking will be already in line with some of the platform movement, there might be a small issue with gyros from the HMD but I can't tell how big the effect will be.
If tracking is done well, and they say it is, then it should be just like when the car is moving, you can still move your head and keep eyes on the road.

Also I don't plan to have a huge motion range.
If it starts bothering me then I'm sure it won't be difficult to extract the motion data from sim software, so when platform is moving you can deduct the platform motion data from head movement data.
Here is another more hardware approach
https://developer.oculusvr.com/forums/viewtopic.php?f=29&t=7933

Hi Fer
Thanks for the suggestion but adding extra motor is one of the things I'm trying to run away from

Also I don't have access to cheap or free wiper motors.
I don't want to start a flame war (as I said, I already have a thread on other sim forum), I respect both communities and I am really happy there are different sites committed to the same goal.
I think there is lot to be learned from everyone and would like to encourage the cross pollination of ideas whenever possible.
I want to be able to use both motion sim solutions and please let me know if you know of any other!
Regards,
Tino.

Is this necessary to use inverse kinematic algorithm for 6-dof sim? May be we can create special setup for SimTools and it will enough?

Hi @prilad

I look forward to that, its the main drawback to getting 6 dof for diy builders,
On my platform I am going to try 3 off single pivot points under the top platform (two rose joints on one bolt), so each motor pair connect to one pivot, that way I reckon I can run Ian's 6dof for flight sims, and 4dof for racing games (R, P, H, Y) those four seem straight forward, its sway and surge I think are the problem without Inverse K.

regards Terry

Ok, Now I work on 6-dof firmware for arduino. And I'll try to use inversion kinematic. If you know easy algorithms for this - it will be useful...

Regards,
Ale