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

JADs 6dof with ballscrews

Discussion in 'DIY Motion Simulator Projects' started by JAD, Feb 25, 2018.

  1. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    I'll will come back to you but still need more time. I havent had a chance to get out to the shed for a while.
    Im just catching up on overdue work lately.
  2. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0

    Hi @thefost
    Been a while, I still havent spent much time in the shed this last month.
    We did talk about 1) query in Discord but I'll repeat it here in the build log with a bit more info

    1)
    I havent used single ballnuts before, this was my first actuator with doubles. The only reason I chose doubles is due to adopting the unframed push rod type of actuator.
    I thought the double nut being longer and offering 2 ball races at a distance apart that it would survive better any bending forces imparted by the the tube extension.
    This is important if you have a flexible or plastic antiwobble ring at the top, then the double ball nut is taking all the bending forces.
    But in hindsight, I had a solid ball bearing as the anti wobble that was very close fitting to the inside of the tube.
    This means it reduces the bending load on the ball nut as the load is shared over two contact points at a much longer distance apart.
    I would be happy to do it again with a single nut but only if the anti wobble was close fitting and rigid.
    I also dont think we are talking about much in the way of bending forces. It would only be self weight of the actuator given its inclined over from the vertical. And a little from inertia with rapid movements.

    You asked earlier if they are noisy but I cant say if they are or not since I dont have a comparison. But Im guessing it is would be.
    I wouldnt say its too noisy though. The fans on the 6 power supplies make more noise allthough they are a constant drone so you can tune out to it.

    The overall noise is bothering me and Im thinking of buying decent noise cancelling wireless ear phones.
    Im currently using Pimax 5k+ with the deluxe headband from the original Vive.
    I dont want over ear headphones as I find Im taking the headset on and off frequently so Im looking around for a wireless ear phones.
    I probably would still be wanting these even if its single ball nut.


    2)I found the cross bar between the front pairs of actuators to be the porblem.
    Initially, I had the seat back and higher, but the cross bar was in the way of my lower legs.
    I shifted the seat lower, but the bulky pedal set was getting too close to the front actuators at extreme positions.
    I really wanted to enjoy the full travel of all the actuaotors so I didnt want to limit the movement.
    So I moved the seat forward to get the pedal set clear out the front of the actuators. Now I only have the narrow single beam passing between the front pair and the pedals and single beam all clear the actuators in extreme positions.
    But moving forward meant my knees were hitting the cross beam, so I had to lower the seat again.
    In the end it was fortunate it worked out, but only just by luck.
    I wouldnt recommend having a straight beam across the front pairs. It is too limiting. If the ballscrews were shorter, I would have been in trouble.

    3)
    I always knew that I could not anticipate the best geometry and configuration of all the componenets so it was very advantagous to use the T slot.
    I do recommend it for your first rig.
    I have a feeling it is quite heavy but now after some time feeling how responsive the actuators are performing Im not concerned about the weight any more.
    I will eventually have a second go and redo the whole upper frame. I think on the next one it will be half and half T slot and plain box sections.

    4)
    When I was thinking of the shorter ballscrews, I was thinking that my next upper frame would be more of a flat deck on top of the actuators.
    In this way you would have shorter ranges but by making the connection points of the upper and lower frame closer together(in plan), you can still achieve adequate movements.
    At the time, I was also wanting to move the rig into the spare room in the house when fininshed so I would have liked the footprint to be smaller.
    But Ive given up on moving it in the house because the rig is never done and we are always working on it and adding to it(no surprise there!)
    Im happy now to stick with 600 or 650 ballscrews and live with the space the whole thing tales up.

    I dont think you need 20mm ballscrews if you are doing a conventional framed actuator.
    As for ballscrew pitch , as you know Im using 5m pitch but I am curious about how 10mm pitch would perform. Im not sure myself.
    But Im inclined to think 5mm pitch certainly gives you enough speed when you want it.

    There have been times Ive crashed in game and not had adequate filters running and Im just clenching through the whole experience hoping nothing is going to break. So Im not so sure you would want any more speed with 10mm ballscrews.
    Im just running nominal 3000rpm motors here. (Im not sure what rpm they are really running at in service)
    The other factor is with 5mm you get better vibration feel and quicker acceleration which I think give you a far better feeling than top speed.
    Its just like what we are seeing with most electric cars being quicker to accelerate but high performance ICE cars can eventually overtake at the higher speeds.
    For us, I think the tradeoff is better to have a short and sharp feeling. The top speed of 5mm screws is just enough and not worth chasing higher speeds.
    If still chasing a higher top speed, you might be better choosing motors with more than 3000rpm than swapping to a 10mm pitch. But motor torque needs to go up to keep the same short and sharp feelings(with 10mm)

    Sorry my reply is late.
    Just been smashed lately at work and been working most days.

    Cheers
    • Informative Informative x 3
    Last edited: May 2, 2020
  3. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    @dureiken
    Attached are some dimensions added to photos.
    Let me know if you like some other dimensions missing from these.

    Attached Files:

    • Like Like x 1
  4. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    I was thinking a bit more about what I said earlier on the the effect of going from 5mm to 10mm pitch .
    And when I was referring to the advantage of 5mm being faster to accelerate for short and sharp feelings versus 10mm giving top speed, this makes sense to be when referring to the same 16mm diameter ballscrew.
    We can think of it in terms of the incline angle of the grooves as you roll around the shaft. The 1610 angle would be double the 1605.

    But then I got thinking about the effect of a 1605 ballscrew and a 2010 ballscrew.
    The incline angle is not double but only an extra 60% steeper due to the greater circumferance.

    So using a 2010 ballscrew might be a good middle ground between the aceleration and top speed.

    Then again, there is more rotor inertia for the 20mm ballscrew that reduces the acceleration.
    I'll leave it at that. That is something for a mech eng uni student to delve deeper into the weeds
    Getting a bit too acedemic for me.


    @dureiken
    Are you building with 2010? Im looking forward to see how it performs.
  5. dureiken

    dureiken Active Member

    Joined:
    Sep 28, 2016
    Messages:
    151
    Location:
    France
    Balance:
    720Coins
    Ratings:
    +98 / 1 / -0
    @dureiken
    Are you building with 2010? Im looking forward to see how it performs.[/QUOTE]

    thanks for all your answers

    Yes I'am using 2010 noisy ballscrew ! But rig is still not alive :)
  6. thefost

    thefost Member

    Joined:
    Sep 20, 2019
    Messages:
    31
    Balance:
    224Coins
    Ratings:
    +13 / 0 / -0
    Thank you so so much JAD!!! You have no idea how helpful all this is for me. I'll respond more in depth a bit later but I've beenplaying around with your rig dimensions in fusion for the last few hours :grin
  7. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    @thefost

    I was talking earlier about my difficulty with the front cross beam being very limiting for positioning the seat.
    I found on Discord chat this example that shows the positions of the bottom joints further apart which would solve the problem I had.
    So in essence, the lengths of the 3 sides of the triangle gets larger with the same actuator range.
    So your platform is sitting higher up in the front V of the front actuators.
    You just have to picture the bottom triangle below the floor (project along the actuator lines)
    You could try this if your wanting to use a straight cross beam and avoid sitting up too high.
    But the effect of the bigger triangle means that for the same length actuator throw, you get less range on some of the dof's like roll and pitch

    You got me thinking about what I would do on my next platform.
    Lots of tradeoffs to consider!

    upload_2020-5-3_15-20-15.png
    • Like Like x 1
  8. thefost

    thefost Member

    Joined:
    Sep 20, 2019
    Messages:
    31
    Balance:
    224Coins
    Ratings:
    +13 / 0 / -0
    @JAD
    Yeah I've been struggling with the same questions. I've been trying to design my rig with as few compromises as possible which has been tricky. I've spent way too many hours playing with Flypt mover trying to get the numbers to work.

    Here's where I'm at so far building off of your dimensions. This is all really rough.

    Annotation 2020-05-03 084029.jpg


    My solution to having a lot of roll/pitch is to make the top platform as small as possible. I'm making foldback actuators very similar to @dureiken because the small top platform would make the rig too large/tall otherwise.

    Similar to the picture of Coops rig you posted, I'm spreading out the bottom actuators in a semi-hexagon pattern. The actuators are also going pretty much straight up, perpendicular to the ground. Should reduce the issue of collisions considerably. Plugging these dimensions into FlyPT with 650mm ballscrews , absolute max numbers in each DOF are +/-405mm sway, +/-420mm surge, +/- 275mm heave, +/-43° yaw, +/- 31° roll , +36 /-29° pitch. I can get a combined 10° yaw, 20° pitch, and 20° roll simultaneously. Agian, all ignoring collisions which I need to work on next. But that's probably more than I would ever need in terms of motion capabilities.

    Attached Files:

    Last edited: May 3, 2020
  9. thefost

    thefost Member

    Joined:
    Sep 20, 2019
    Messages:
    31
    Balance:
    224Coins
    Ratings:
    +13 / 0 / -0
    So, I was planning on 10mm pitch more for noise than for speed. From what I've read, ballscrew RPM is the biggest factor in noise. If I use 10mm pitch I can reduce the rotations in half and still get the same speed.

    I can do this two ways, the first of which is to just reduce the speed in software. Obviously I'd end up with half the toque this way, but I'm not sure if this is even an issue. The 80ST servo motors I'm using are 2.4Nm, 7 peak, 750W. So far peacemaker is pulling just 500W total for all 6 of his motors combined in use. Seems like a whole lot of headroom.

    The second way I can adjust the speed is to just change the gearing ratio on the foldback actuator pulleys. This would get me my torque back if I need it. So it seems like 10mm pitch is the way to go for me.

    The only issue for me is whether to go 1610 or 2010. I've been concerned for a while now that rotor inertia might be a factor. That's really the only thing holding me back from buying 2010 ballscrews right this second.


  10. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    I just jumped on board with the 80ST M02430 servo motors also. Ordered everything last week.
    Im going to swap over just the moter end and keep the exisitng top end. But I do have plans to swap the top end also but I might wait to see how you guys go with 10mm pitch screws.
    My main motivation for changing motors was to get the smoother slow speed feeling. The brushed DC motors were feeling a bit notchy when just floating around in flight sims and all 6 are heaving slowly.

    I have a gut feeling there is enough torque headroom to not worry much about the tradeoffs decisisions we are talking about.
    I think its a good idea you are going foldback with gears. Its not much extra cost and like you said, you can adjust these tradeoffs afterwards.

    With the choice of 1610 and 2010,
    Im guessing with the torque we have available, the extra rotor inertia is a small factor in reducing the acceleration
    If ballnut noise is a factor, need to consider that the ball race length in a 2010 is 25% longer per turn so for the same rpm = faster ball rolling speed and extra noise.
    Same again, Im guessing its a small difference and shouldnt be the main factor for the choice.
    Im inclined to think the 2010 as the better choice as you get the not so steep ball race and get back some of the short and sharp feeling when going to 10mm pitch.

    I like your rig design.
    I see why you have spread the bottom points to make space but I think you can afford to bring them in a bit closer.
    Since you have a smaller triangle, and you have your legs and pedals well out clear past the front V this has avoided the problem I had.

    With a 650 ballscrew, Im guessing you would have a mech range of 450-500
    and maybe a in service range of 400.
    You definitely have plenty of DOF range!
    I too was wanting heaps of range but found I have been dialling it back.
    The only time the extra range is great is when in VR with motion compensation, otherwise you will get sick with all this huge movement.

    With a smaller triangle and actuators standing up a bit taller, you are going to want tight connections and rigid frame.
    I noticed this rig and watch for the hard stops after movements, you can see a little wobble.
    This looks like a very well built frame with tight connections but it still wobbles a little. And even more so when the platform is loaded up with weight.

    Im thinking you would want to lean over the actuaotrs as far as you ca to make for a more rigid and stable platform.
    Particularly in the surge, sway and yaw movements.

    Perhaps if you make the botom frame a hexagon with equal 6 sides, and you tek screw the bottom joint to the frame, you could later on just relocate the positions after some trial and error.

    Its great discussing this stuff, its good for thinking through the various effects of different designs.

    Cheers
    • Like Like x 1
  11. adgun

    adgun Active Member

    Joined:
    Jan 28, 2008
    Messages:
    422
    Occupation:
    mechanic
    Location:
    Netherlands
    Balance:
    5,001Coins
    Ratings:
    +100 / 3 / -0
    a 750 watt motor with 1610/2010 ballscrew wont be a problem, it still delivers 135kgf at rated torque
  12. JAD

    JAD Active Member

    Joined:
    Jan 17, 2016
    Messages:
    193
    Location:
    Australia
    Balance:
    1,294Coins
    Ratings:
    +140 / 3 / -0
    Time for long overdue update.
    The AC Servo upgrade is coming along.
    IMG_20200816_152010.jpg
    IMG_20200816_174825.jpg
    IMG_20200816_174832.jpg
    IMG_20200816_213822.jpg
    • Like Like x 6
  13. Gadget999

    Gadget999 Well-Known Member

    Joined:
    Dec 27, 2015
    Messages:
    1,036
    Location:
    London
    Balance:
    7,489Coins
    Ratings:
    +237 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino, 6DOF
    What servo drivers are they ?
  14. Thanos

    Thanos Building the Future one AC Servo at a time... or 6

    Joined:
    Jul 6, 2017
    Messages:
    941
    Occupation:
    Electronics Engineer
    Location:
    United States
    Balance:
    4,530Coins
    Ratings:
    +792 / 8 / -0
    My Motion Simulator:
    AC motor, Motion platform, 4DOF, 6DOF
    AASD-15A...
    • Like Like x 1