1. For downloading SimTools plugins you need a Download Package. Get it with virtual coins that you receive for forum activity or Buy Download Package - We have a zero Spam tolerance so read our forum rules first.

    Buy Now a Download Plan!
  2. Do not try to cheat our system and do not post an unnecessary amount of useless posts only to earn credits here. We have a zero spam tolerance policy and this will cause a ban of your user account. Otherwise we wish you a pleasant stay here! Read the forum rules
  3. We have a few rules which you need to read and accept before posting anything here! Following these rules will keep the forum clean and your stay pleasant. Do not follow these rules can lead to permanent exclusion from this website: Read the forum rules.
    Are you a company? Read our company rules

Showroom Budget 2DOF VR Flight Simulator rig - MSFS, SMC3, FlyPT Mover, Transducers

Discussion in 'DIY Motion Simulator Projects' started by Ronan Design, Jun 11, 2021.

  1. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Update 3:
    All design files, Blender design files, 3D-print STL files, build photos and part lists for my Motion VR Cockpit and all controllers are now available for FREE DOWNLOAD (optional donation) on my website. If you are looking into building a similar rig - feel free to use my designs as a starting point.

    Update 2:

    Here's a video of the latest version of my rig, complete with a custom-built hall-sensor pendular yoke convertible to a joystick, encoder box, switch/gear lever box, HOTAS throttle convertible to heli collective control and its switch box.



    [​IMG]

    Update:

    All right! I have completed my project (for now, at least) and it works great so far!
    Here is a detailed video - demo flight, hardware and software walkthrough etc.




    [​IMG]

    ---
    Original post:

    I'm in a planning stage for building my budget 2DOF VR Flight Simulator (MSFS) rig. I appreciate your advice regarding my design, engineering solution and a parts list.

    Did I forget anything? Any bad decisions I'm making? Is there anything wrong with my design?

    I'm starting this thread and will refresh it as I build, with pictures and videos.

    Here is my design. Low budget is essential, and I don't have metalworking equipment, so I will be building it out of wood: 2x4" and 1x4" lumber mainly. I don't know how to get and set up a suitable Universal Joint without welding and for cheap, so I'm planning to build it into the rig.

    Here is my general design. For VR without motion compensation I need fast movement but small angles. As I'm not into combat flying and more into MSFS, this is fine. If motion compensation for Reverb G2 in WMR mode (not SteamVR) will become possible I would like to keep my option for larger angles, but I may limit the angles at first of they are too much for VR. DIY Center-mounted joystick will be optional, when I need it for a specific aircraft. Yoke can be swapped for a wheel for racing games. Please tell me what you think about the design and what am I doing wrong, if anything. Advise on improving the design is appreciated!
    upload_2021-6-11_13-25-15.png
    upload_2021-6-11_13-32-40.png

    Here's a close-up of the "universal joint" arrangement with 4 pillow-block ball bearings (see parts list below):
    The pitch: 2 long beams (going left to right) pivot on the 2 bearings on Green Y axis. Roll: the square "u-joint" wooden block pivots on Red X axis on its 2 pillow block bearings. All bearings are on the bottom side, so the upward pressure works for the stronger attachment.
    upload_2021-6-11_13-31-29.png

    I tried calculating with SimCalc but I'm not sure it's making sense. Am I doing it right?

    If I set "motor position" Y to 20cm as I measure it in my model, it says "invalid geometry", but 23cm works. Why is that? I'm definitely doing something wrong. Here is the working result with 23cm:
    upload_2021-6-11_13-41-59.png
    These are my distances:
    upload_2021-6-11_16-7-55.png

    PART LIST

    Here is what I'm planning to order. Feel free to point out problems / suggest alternatives. I appreciate if you confirm that those are sound and proper choices for my rig. I list my questions for respective parts, and your advise is much appreciated:

    Motors: PGSAW 12V 75-80RPM 50:1 x 2
    https://www.ebay.ca/itm/371596157992
    $128.50 x 2
    [​IMG]

    - Are those the right motors for my rig?
    - Why is there a calculation problem with my numbers in SimCalc?
    - How to securely attach the arm to this shaft, so it doesn't slip under load?
    - I can get hoverboard motors for free but from what I understand brushless motors are not suitable for SMC3 control, and motor drivers have to be different. If anyone has done this successfully, so I can use the free motors instead of expensive ones. Can it be done?


    Motor Driver: IBT-2
    https://www.aliexpress.com/item/1005001631113020.html
    $6.83
    [​IMG]

    - Do I need just one board for 2 motors?

    Power Supply: HP Server 12V 750W power supply, possibly x 2
    $0 (I can get them used for free)

    Arduino R3 clone
    https://www.aliexpress.com/item/1005001621923182.html
    $8.11
    [​IMG]

    - Should I provide 5V from my Power Supply or USB?

    DIY Hall effect frictionless position sensors: AH3503 hall sensors + 5x1.5mm Magnets (possibly stacked) attached to motor shafts.
    https://www.aliexpress.com/item/32835617607.html
    $2.17
    [​IMG]
    https://www.aliexpress.com/item/1005001832523652.html
    $6.73
    I'm building a joystick now that uses those sensors and magnets. I think I should be able to make it work for 2DOF rig as well: epoxy the magnets to motor shafts and 3D-print sensor holder so it's fixed a few mm from the magnets to register rotation. This would have an advantage over (expensive) off-the-shelf angle sensors because there is no friction, no gears or belts are involved, less things to misalign and break. Will need to test this.

    DIY Universal Joint: UCP202 Pillow Block Bearing x 4

    $20
    [​IMG]

    - Would it work? Is my geometry feasible?


    Rod Ends M12 x 4
    https://www.aliexpress.com/item/32878395360.html
    $14.90
    [​IMG]

    - Is the m12 size a good option?


    I will design and 3D-print conical washers for the fill range of the movement.

    Rods M12 x 25cm x 2
    https://www.aliexpress.com/item/1005002205826524.html
    $19.04
    [​IMG]


    M12 Bolts x 8
    https://www.homedepot.ca/product/pa...-hex-cap-screw-din-931-zinc-plated/1000132169
    $23
    For the u-joint I would have to 3D-print some filler jackets to make sure bolts and pillow bearings are snug. Smaller pillow bearings seem too flimsy to carry the load, and may slip sideways and break when the rig is angled. These look solid enough, but the bolt size is too wide for what I can get here. We'll see.

    M12 Nuts x 12
    https://www.homedepot.ca/product/pa...metric-hex-nut-din-934-zinc-plated/1000128525
    $5.52

    M12 Washers x 12
    https://www.homedepot.ca/product/paulin-m12-metric-lock-washers-zinc-plated/1000128577
    $2.40

    Lumber:
    2x4 x2 = $22
    1x4 x2 = $8

    - Is there anything I forgot, or listed incorrectly?

    Thanks for all advice and suggestions!
    • Winner Winner x 1
    Last edited: Sep 10, 2022
  2. Ads Master

    Ads Master

    Balance:
    Coins
    Ratings:
    +0 / 0 / -0
  3. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    Even with pgsaw 50:1s I would suggest moving the motors further back, as far as you can which is basically the back board of the base, as compact designs require as much design torque as possible.
  4. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Thanks @noorbeast . What does moving them back do? I can even extend the frame if needed. Should the rod mount points be moved as high as possible, or as low as possible if I move the motors back? Is the goal to have zero position with arms looking up to have rods as close as possible to a 45 degree angle?
    If I move motors back, I guess the arms should be looking outside then, so the full rotation is possible (in emergency case when something goes wrong). I n that case, should I extend the top rod mount board sideways, or keep it as is and have the rods at an angle in a sideways direction too?
  5. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Come to think of it, if arms are angled and motors moved backwards, but the rod mount points are lower than main seat rotation point (u-joint), then pitch up movement (tugging seat back down) would be at a disadvantage... Should the rod attachment point be moved higher than the u-joint, if motors are moved back? Kind of closer to the SimCalc illustration? I just can't picture the geometry on my rig, with motors on the back board, pulling seat down. I think it would lock...
  6. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    You have SimCalc, so increasing the Y distance between the pivot and motors will demonstrate an increase in the torque.

    The angle of the lever to the rods should be 90 degrees (max torque efficiency) when the rig is level and at rest. The torque efficiency of a lever drops off rapidly over about 40 degrees up and down. Think of it like a door, if you are at 90 degrees you can push it with your little finger, but the closer you are to the door edge the less efficient the force.
    • Like Like x 1
  7. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Somehow it doesn't change at all:
    If I increase Motor Position Y from 0.23 to 0.43 for example, I get exactly the same torque:
    upload_2021-6-11_20-12-30.png upload_2021-6-11_20-12-52.png
    My pitch and roll travel and rotation all decrease, and torque stays the same? I tried other values - same result. Looking at my 3D image now I see that the maybe arms should be slightly farther than 90 degrees to the rods, to counteract loss of torque at extension? But you say they should be 90 degrees...
  8. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    You need to move the rod mount position on the Y axis back as well.

    50-60mm is likely in the ballpark for a compact design lever CTC distance.

    Keep in mind that in a compact design the actual perceived linear speed is multiplied by the seat back height, where you head is, rather than the measure at the base.
  9. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    I tried moving both motors and rod positions back, and shortening CTC to 60mm and got results that have more torque and much less travel because of shorter CTC. Also twice less rotation. @noorbeast is this configuration what you meant?
    upload_2021-6-12_13-25-54.png
    The rig now looks like this:
    upload_2021-6-12_13-26-27.png
    Or did you mean I needed to angle the rods? Would 60mm CTC give me enough travel? 15 degrees seems like too little...
  10. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    In VR less is more, even for flight you want fast precise movement, rather than large wallowing axis movements, and on a compact rig you need plenty of torque to move it. You have to keep real physics in mind and what it actually takes to drive a loaded rig.

    Yes I do mean extend Y rod and motor position as much as possible within the design constraints.
  11. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    So does my updated sketch correspond to what yo usuggest? Motors and rod attachment points moved back, rods perpendicular to the ground and motor arms, CTC 60mm. Does it make sense?
  12. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    It makes sense for the current motor placement design, with the motors recessed in the frame.

    But if the motors were mounted a bit higher on the frame they could be moved to the very back edge of the frame with the levers being placed on the other side and extended out to clear the frame on the outside with enough room to rotate, similar to my compact rig, but motors mounted to the back instead of the front: https://www.xsimulator.net/community/threads/dx-compact-simulator.5866/
  13. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Great rig! Looks like a perfect production quality, it looks almost like an industrially manufactured device. You seem to be using similar motors to what I'm planning to use. Are the arms attached by a bolt straight through the shaft and a couple of nuts? Is that enough to hold the arms and prevent them from rotating around the shaft? What bolts did you use?
  14. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    I use solid shaft 60:1s, so have keyway attached levers, but a bolt is used on hollow shaft wormdrives, see @Nick Moxley's example in the FAQs: https://www.xsimulator.net/community/faq/how-to-attach-lever-crank-arm-to-motor-shaft.182/
  15. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    Thanks. I don't have access to a proper drilling press and welding equipment, which may complicate things. But I have an idea now about what people do, so will try to figure this out.
    • Like Like x 1
  16. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    I have another question:

    Is it very important for pitch and roll axes to be at the same plane (i.e. a u-joint)? Or is it OK if X and Y rotation points under the seat will be one above another at about 5-10cm apart on a Z axis? I don't have a suitable universal joint, but I got 4 x pillow-block bearings. I can build what I designed in my sketch (essentially a u-joint) with 4 of them, but it will be too massive and heavy (blocks are larger than I anticipated). It would be best if I use just 2 of them, but then they won't be in the same plane, one would have to be above the other at 90 degrees. Is it a bad idea, or geometry should still be OK?
    [​IMG]
    Last edited: Jun 22, 2021
  17. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    While it would work introducing geometry quirks can result in motion profile tuning challenges, it sort of depends how far apart the planes are, the greater the distance the more possible complications.
  18. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    OK, thanks. Will think it over. It's an extra kilo of weight plus a wider and heavier frame VS separate planes.
  19. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    OK, finally I have (very carefully) tested a first rough version of my rig. Generally, it works, including my DUY Hall Sensor + Magnet 3D-printed contactless rotation sensors. Motor mounts need reinforcement, and pitch travel feels much less than roll. So maybe I need to move the rod attachment points forward (closer to COG)?



    Also, without me in the seat, the rig is very wobbly when moving, often moving in juddered steps. It gets better when I'm in the seat though. But I wonder what can cause that - is it the mechanical problem, or PID tuning? You can see my settings in the video, and see and hear the wobble effect:



    I'm also having problem with end switches: I decided to put limit switches to the frame, so motor arms couldn't crash into a frame if soft limits failed or hall sensors detached or anything else. I use 2 x micro limit switches. They are quite small and cheap (Ali) - see image. They are rated 5A on 125/250V, but I wasn't sure I can put the main motor current through 2 of these (for each motor), so I interrupted the PWM leads to Arduino pins 9/10, thinking "no PWM - no movement". Wrong. After I cut any of the switches, the motor does 2-3 random small movements in one direction. So what line can I put these on, what should I interrupt to immediately stop the motor upon hitting the end switch?

    [​IMG]
    My connection is like this:
    [​IMG]
    Maybe I should cut +5V to IBT2's instead? Or something else? I can only cut one line per IBT2, so can't cut 2 or 3 of them together.

    My first version of the rig works, but I would like to have a real protection from crashing it in case of any failure.
    • Like Like x 1
  20. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    18,635
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    133,282Coins
    Ratings:
    +10,263 / 48 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    What CTC are you using on the levers?

    You need to refine your SMC3 settings further, in particular limit the lever range and smooth out the response.

    A lever operating over about 38-40 degrees up and down is very inefficient in terms of effective use of torque, and with a compact design efficient use of torque is paramount.

    OverRotate.png
  21. Ronan Design

    Ronan Design Roman Design - Custom MSFS Scenery

    Joined:
    Jun 4, 2021
    Messages:
    90
    Occupation:
    Web Designer, MSFS Scenery Designer
    Location:
    Toronto, Canada
    Balance:
    748Coins
    Ratings:
    +21 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino
    What's your advice on SMC3 settings? Do you think the range looks too large? How would I smooth out the response more?

    I'm using CTC of 6cm, and I have reserved holes at 8cm and 10cm. So far I think 6cm is efficient enough, considering I will have VR without motion compensation (MSFS on Reverb G2 WMR)

    Looks like there is much more roll than pitch available, so I'm thinking about moving motors forward a few cm, to allow to move rod attachment points forward too - closer to CG, so there is more pitch effect. Also, I want to lower the attachment points just a little bit, because angle looks not 90 degrees but a bit more, so there is less forward pitch than backward.

    Is anything wrong with the rig in your opinion, or things that can be improved? Note, I tested in FlyPTMover by directly slaving joystick to roll and pitch (with some smoothing), haven't tested on any sims yet.