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 Download Package Now!
  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 here. Do not following these rules will lead to permanent exclusion from this website: Read the forum rules.

Showroom 2DoF Seat Mover

Discussion in 'DIY Motion Simulator Projects' started by Bord-Ing., Jan 29, 2018.

  1. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    Hi there,

    just after 1.5 years of thinking (and doing nothing) I startet to build my 2 dof seat mover sim rig. The design is influenced - er, mainly copied from ;) - the APX series from Ricmotech. I increased the length to have space for the wheel chair motors, 24 V 350 W each:

    pic01.jpg pic02.jpg pic03.jpg

    Machining the wooden parts:
    pic04.jpg pic05.jpg

    I used an industrial cardan which I got at a cnc forum meeting for a few euros:
    pic07.jpg

    "Final" (provisional) assembly:
    pic08.jpg pic09.jpg pic10.jpg

    Next step is the paint work, I've got black left over from a garden project...

    ... And then: try to get it move :D I plan to use a chinese dual channel 100A driver.

    Options: I read about the G-seat and figured out, that my seat can be modified for that purpose. If I manage to keep the parts thin, they could easily be placed under the cushions (which are only fixed with velcro strips).
    • Like Like x 6
  2. Ads Master

    Ads Master

    Balance:
    Coins
    Ratings:
    +0 / 0 / -0
  3. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    It's too cold to paint, so I went on with some mechanics:

    I shortened the axle of the worm gear to fit a lever 20 mm thick. And made some threads to fit the position sensors. I plan to go with DIY hall sensors. I did some tests here.
    pic04.jpg pic05.jpg

    And it's time for the power supply. From an old CNC project I had a toroidal transformer left over, which outputs 2 x 15V /26A. I connected the 2 coils in parallel resulting in 15V 52A AC. After rectification there should be 24+ V DC available.
    pic01.jpg pic02.jpg

    For voltage damping I milled a PCB, which gets equipped with 20 pcs 4700 µF capacitors and fuses for the drivers.
    pic03.jpg
  4. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    Ready to use: Capacitor board...
    pic1.jpg pic2.jpg

    ...and sensor pcb :)
    pic3.jpg

    I need a smaller soldering iron :confused:
    • Like Like x 2
  5. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    Not so much progress this weekend :(

    I turned the housings and clamping rings for the sensor pcb:
    pic1.jpg pic2.jpg

    Just fitted to the motors:
    pic3.jpg pic4.jpg

    To adjust the sensors I loosen the screws, turn the sensor housing and fasten the screws again. Not much torque needed, 6 screws are totally oversized :cool:
    • Like Like x 2
    Last edited: Feb 11, 2018
  6. kanuk

    kanuk If it ain't moving, it ain't simulating...

    Joined:
    Jan 24, 2015
    Messages:
    214
    Occupation:
    Eng
    Location:
    Melbourne Australia
    Balance:
    2,269Coins
    Ratings:
    +202 / 1 / -0
    My Motion Simulator:
    2DOF, 3DOF, DC motor, Arduino, JRK
    That dear sir is a nice piece of engineering right there! Very very nice... Well done! :thumbs
  7. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
  8. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    So, first problem arose...

    Tests with Arduino Uno + SMC3 + hall sensors gave way to low working angle of the lever. Only perhaps +/- 30 degrees. A deeper look on the curves from SMC3Util showed a very steep slope around the center (0 - 30 degree) and a wide, flat area over some 40 degree. Dimensions of the magnet holder were the same a at my testings, but I used other magnets. For the tests I used some 3 x 3 x 1 mm N42 magnets, but I had only these two pieces, so I bougth some new. For easier machining I purchased round 5 x 2 mm, but in N52 quality. This lead to the extreme early saturation of the SS49E hall sensor.
    So I did some more tests with different distances of the magnets:
    1. Holder #1: 3 x 3 x 1 mm N42, distance 7 mm
    2. Holder #2: D5 x 2 mm N52, distance 7 mm
    3. Holder #3: D5 x 2 mm N52, distance 13 mm
    4. Holder #4: D5 x 2 mm N52, distance 10 mm
    Result: Holder #4 ist most suitable with most usable angle, nearly like holder #1
    Magnet Holder.jpg

    So I switched to holder #4 (had to build 2 of them) and ran some more tests:
    SMC3-Capture.jpg

    I have little overshoot in the square step response, and can't get rid of them :confused:
    My settings are Kp = 400, Ki = 0, Kd = 1000, Ks = 3. Even with the smallest Ki I get oscillations...

    I also observed some "stepwise" motion of my levers. As if the position is faster reached as the SMC3Util can send next position. Does anybody know the update frequency of new position data from SMC3Util?
  9. Bord-Ing.

    Bord-Ing. New Member

    Joined:
    Sep 17, 2016
    Messages:
    25
    Location:
    South Germany
    Balance:
    349Coins
    Ratings:
    +17 / 0 / -0
    Nonetheless I continued to build my rig. Milled the levers and turned the pushrods.
    pic1.jpg pic2.jpg

    Until now I use this chinese dual H-bridge:
    pic3.jpg

    The used components are not so bad, FETs are IRFB 3307 (75 V, 130 A from datasheet), and it has it's own step down converter on board. But thermal dissipation design is - er - ok, suboptimal... That nice, little, green, hollow (!) heatsink is way to small, and is glued with some thermal silicone on the plastic side of the FETs. For SMT FET there should be a larger area of copper for heat dissipation. But there is - none... At my tests with too high Ki and some oscillations the driver got very hot and burnt a part of the pcb -> smoke. But surprisingly nothing died...
    It's really a shame. On the electronic side this H-bridge could be a winner, but thermal design - along with the underdesigned screw terminals - no comment. Just buy a sabertooth...
    Last edited: Feb 20, 2018
  10. RandomCoder

    RandomCoder Active Member Gold Contributor

    Joined:
    Feb 19, 2017
    Messages:
    185
    Occupation:
    Control Engineer
    Location:
    UK
    Balance:
    1,626Coins
    Ratings:
    +138 / 0 / -0
    Here's my understanding of three term controllers (PID's).

    The Proportional term (P) is the gain, it's how hard the controller tries to force the output to the required value. Too much gain will cause overshoot, too little will result in a very damped response.

    The Derivative term (D) affects how the controller reacts to rate of change. This can be the trickiest term to set because any noise on the feedback signal can cause spurious results. But to get the system critically damped its important to get the balance between P and D correct.

    The Integral term (I) removes steady state error. This is applied when the output never actually reaches the required value given a fixed output.

    There's several methods of tuning PID's some of which are more suited to heating control and others to positional feedback etc. Rather than copy someone else's post I'll link you to a PID seeing guide... https://robotics.stackexchange.com/questions/167/what-are-good-strategies-for-tuning-pid-loops
    • Informative Informative x 2
    • Like Like x 1