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 Dschadu's 6 DOF build

Discussion in 'DIY Motion Simulator Projects' started by Dschadu, May 23, 2020.

  1. Dschadu

    Dschadu Member

    Joined:
    Jan 2, 2017
    Messages:
    62
    Location:
    Germany
    Balance:
    1,584Coins
    Ratings:
    +89 / 1 / -0
    D-Hi everyone,

    long in the planing, finally making it! I started to research on motion platforms 5 years ago and started to plan my own actuators and electronics.
    But these tasks turned out to be enormous and I ditched the DIY-electronics (which already eaten up about 300h development - with approximately 200 more to go).
    Recently @SilentChill showed his BLDC Actuators - the design is simple, easy to build and powerful!

    So I started to buy stuff...
    3x oDrive 3.6 56V Version without connectors - 183,46€ each
    6 x Turnigy Aerodrive SK3 - 6374-149KV Brushless Outrunner Motor - 89,18€ each
    12x IGUS GSM121425 bearing - 1,20€ each
    6x IGUS GFM252625 bearing - 1,33€ each
    6x Encoder AMT102-V - 20,55€ each
    IMG_0007.JPG IMG_0008.JPG IMG_0010.JPG
    IMG_0011.JPG IMG_0013.JPG IMG_0014.JPG

    The encoders do come with a lot of different mounting hardware. The biggest plus is that they are wear-free as there are no moving or touching parts.
    IMG_0020.JPG IMG_0021.JPG


    To get things right, the I also bought the correct connector:
    6x MOLEX 50-57-9405 - 3,29€ each
    100x MOLEX 16-02-0069 - 7,92€
    IMG_0019.JPG

    And a nice SubD9 connector, too:
    ENCITECH 6355-0009-01 - 1,78€ each
    IMG_0015.JPG

    And the other SubD stuff:
    50m LIYCY 5 x 0.25 - 35,20€
    6x D-Sub Female - 0,30€ each
    6x D-Sub Male - 0,96€ each (I have no idea why the are 3x the price...)

    To keep the odrive cool, I got these 40mm fans:
    12x 12VDC 40x40x10m 15,7dBA; SUNON HA40101V4-1000U-A99 - 2,55€ each
    12x 40 x 40mm fan grill - 0,43€ each
    50x Screw STP5X10 4.2mm - 4,55€
    IMG_0023.JPG

    And some connectors I already had:
    IMG_0022.JPG


    oDrive Case and assembly:
    Its not fully assembled on the last picutre. Still needs the leads for the second motor and the subd9 connector.
    Airflow is designed to go over the cooling fins. They get controlled from a central spot to keep them as quiet as possible. At 7 V they are very silent and they do push a lot of air ofter the odrive board.
    IMG_0031.JPG IMG_0032.JPG IMG_0038.JPG IMG_0039.JPG IMG_0042.JPG

    The encoder cable is fully shielded.
    IMG_0041.JPG IMG_0043.JPG

    As power source I bought a WS-CAC-6000W - 38€. The scrap price would score more... absolutely incredible.
    IMG_0035.JPG IMG_0036.JPG IMG_0037.JPG

    Only downside: The fan is really loud. So I removed it and put a more silent one in, only to discover that the integrated fan controller is not happy with the low RPM and shuts down the output. Solution was a simple NE555 circuit to emulate a faster spinning fan. And it worked :)
    Heat-shrink around it to avoid bad thinks from happening and put it all back together.
    IMG_0044.JPG


    More pictures and infos to come!

    Attached Files:

    • Like Like x 7
    • Winner Winner x 1
  2. Ads Master

    Ads Master

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

    hannibal Active Member

    Joined:
    Sep 29, 2018
    Messages:
    555
    Balance:
    3,735Coins
    Ratings:
    +235 / 2 / -0
    My Motion Simulator:
    3DOF, 6DOF
    nice work!!
  4. Dschadu

    Dschadu Member

    Joined:
    Jan 2, 2017
    Messages:
    62
    Location:
    Germany
    Balance:
    1,584Coins
    Ratings:
    +89 / 1 / -0
    A custom connector for the power supply. Its not perfect and also looks not that good, but it should do the job!
    IMG_20200508_171928.jpg IMG_20200508_171950.jpg

    The actuator itself - as mentioned above - is based on SilentChills design. But I had to do some changes to fit the parts I'm able to supply. In particular I had to change the bottom mounting, as I couldn't find the huge U-Bolts that are used in the original design. While I where on it, I removed some material in some place that is not structural critical, mainly to reduce printing time. I also added the possibility to add a 40mm fan below the motor, just in case.
    Some prints turned out great with the first print, some parts needed adjustment, usually because I didn't get the tolerances right.
    After everything arrived, I could start to build the first actuator.

    6x Bearing FK12 - 15,90€ each
    6x Ballscrew 1605 600mm - 36,90€ each
    4x Aluminium pipe 12x1mm, 2000mm, 2,64€ each
    4x Aluminium pipe 25x2mm, 2000mm, 7,04€ each
    Different bolts and nuts in M4 - M6, too much to list here.
    IMG_20200510_121348.jpg IMG_20200510_121404.jpg

    Now with everything in place... time to test the odrive. The documentation is not optimal and the usage is a pain, as there is no UI. But I got the hang of it pretty quickly and it turned out that there are not many commands and settings. I limited the motor to 10 A @ 24 V and 150 rpm for testing and it worked so far. Final version will go with 70 A @ 42 V (at least that's the plan).

    Then the outer pipes arrived:
    6x Aluminium pipe 90x3mm, 800m, 18,40€ each
    IMG_20200515_130442.jpg

    The first actuator is nearly completed:
    IMG_20200515_154145.jpg IMG_20200516_164916.jpg

    And a short video on how it is assembled:


    Now my 3d printer needs to print all the other parts (~160h), but this will take some time as I don't let it run unattended. So only a few parts on weekends.
    In the meantime I can prepare the pipes and start building the frame.
    • Like Like x 7
    • Winner Winner x 1
    • Friendly Friendly x 1
  5. Dschadu

    Dschadu Member

    Joined:
    Jan 2, 2017
    Messages:
    62
    Location:
    Germany
    Balance:
    1,584Coins
    Ratings:
    +89 / 1 / -0
    Its big update time :)
    I can't believe that it has been 3 months... incredible. The amount of work is just insane. I had two weeks off and was working 4-5h every day to get things forwards.

    I started doing a bit electronics: Prepared the oDrives and cabling.
    IMG_20200714_092623.jpg IMG_20200715_160552.jpg IMG_20200715_172404.jpg


    And then i started the task of creating the lower linkage. I need four aluminum block for each actuator. Times 6 makes for 24 identical parts.

    First, cutting:
    IMG_20200810_111217.jpg

    Then marking:
    IMG_20200810_115745.jpg

    Started to drill some holes and tapping the threads:
    IMG_20200810_140434.jpg

    And more drilling (this is a 27 mm drill. It produces insane chips)
    IMG_20200808_091740.jpg IMG_20200808_092513.jpg IMG_20200810_141819.jpg IMG_20200810_162911.jpg

    After I got all 24 pieces drilled and cleaned, I started on the lower frame:
    Cutting, drilling, welding... a lot of work again! I tried my best to get this as accurate as possible. I ended up with 3-4 mm difference between CAD and reality (the long side edge-to-edge). I would say that's pretty good.
    The frame is 40 x 80 x 3 mm. The plates are 5 and 6 mm (had some stock still laying around). Welding is done with a AC/DC TIG welder. The head gets really hot really fast, so I had to make a lot of pauses to let it cool down. Annoying...
    After welding was finished, I only primed the frame to prevent it from rusting. I'm going to paint it black if everything is done.
    IMG_20200811_161056.jpg IMG_20200817_145532.jpg IMG_20200817_162426.jpg IMG_20200819_145032.jpg IMG_20200819_145525.jpg IMG_20200820_094947.jpg

    Then it was time to assemble the actuators... A huge pile of material!
    I drilled all the tubes beforehand and also assembled a few things as I was waiting on the 3d printer to finish. The aluminium pipes are also sanded to prepare them for painting. But this will happen after everything works to prevent scratches. This means I have to disassemble everything again...
    IMG_20200820_131030.jpg

    Some detail-views and part-groups ready to be put together.
    IMG_20200820_133102.jpg IMG_20200822_110914.jpg

    All 6 actuators partly assembled:
    IMG_20200823_110435.jpg


    This is the latest addition: Some laser-cut parts for the lower end of the actuator. They will replace a more crude version made with an angle grinder ;)
    IMG_20200826_175147.jpg


    And besides all that... I also renovated my gaming-room: Put in new outlets, a second circuit just for the Sim, some Network ports and painted everything.
    IMG_20200718_093545.jpg IMG_20200718_174825.jpg

    Before and after:
    IMG_20200718_134143.jpg IMG_20200830_114045.jpg


    Next step is to get the actuators completely assembled. Then put them on the lower frame and build a temporary wooden upper frame. Get the electronics done and test the motion... I think this will take another month or so, as I only have the weekends to work on it.
    • Like Like x 4
    • Winner Winner x 1
  6. Dschadu

    Dschadu Member

    Joined:
    Jan 2, 2017
    Messages:
    62
    Location:
    Germany
    Balance:
    1,584Coins
    Ratings:
    +89 / 1 / -0
    Its again time for a big update!

    The bottom frame found its home (remember I will paint it black when everything is done)
    IMG_20200902_193047.jpg

    The actuators are assembled and ready to go onto the frame. I suggest to get a second person to help you do this. My actuators cannot be laid on the ground, as the lower joint does not allow this much movement.
    IMG_20200902_193054.jpg IMG_20200903_181306.jpg IMG_20200903_190220.jpg

    Build a wooden frame. If something goes wrong, the wood can break and shatter (hopefully)
    IMG_20200905_124245.jpg

    Now it is time to get all the ODrive into their cases
    IMG_20200905_200340.jpg IMG_20200905_200442.jpg IMG_20200905_200501.jpg

    I designed two circuits to fight two issues: Back current into the PSU and overvoltage. The left one is an ideal diode. Instead of a diode with 0.7 V voltage drop it uses a Mosfet with only 8mOhm resistance. Lower losses, lower temperature, more voltage to the motors :)
    The right one is a brake-chopper circuit. It enables a resistor when the voltage rises above a defined threshold.
    IMG_20200930_184900.jpg IMG_20200930_193248.jpg

    The E-Stop stuff and the PSU needs a place to go. I ordered some laser-cut sheet metal and build my custom cabinet. Walls are 1.5mm, bottom is 2mm.
    IMG_20201002_170909.jpg

    All this stuff has to go into there... (and of course the PSU)
    IMG_20201004_085209.jpg

    Some build pictures:
    IMG_20201004_085602.jpg IMG_20201004_092250.jpg IMG_20201004_093030.jpg IMG_20201004_100338.jpg IMG_20201004_151450.jpg

    I wanted to use the original PSU 42V connector, but opted to connect the wires internally.
    IMG_20201006_183200.jpg IMG_20201006_183207.jpg

    Two detail pictures:
    IMG_20201006_183213.jpg IMG_20201009_172442.jpg

    The copper bars distribute the power. Maybe a bit overkill ;) The other picture is the ideal diode. It uses the case as heatsink and is electrically isolated to it.
    IMG_20201009_172446.jpg IMG_20201009_172453.jpg

    ODrive mounting position:
    IMG_20201010_132033.jpg

    Do you remember where I said that I hope the wooden frame would take the beating and not... something else? Well... I was wrong. Had issues with the Index-Signal search. The actuator went from "very slow" to "full speed" and "full extend" within .75sec. The 3d-Printed part got obliterated.
    Luckily only this part got damaged. The spindle puked out all of its ball bearings, but after 45min of fiddling around I got all of them back in.
    IMG_20201010_151949.jpg IMG_20201010_151959.jpg IMG_20201010_160430.jpg IMG_20201016_144108.jpg IMG_20201016_144118.jpg

    The Index signal looked like this. Its just... bad.
    IMG_20201010_152637.jpg IMG_20201011_101901.jpg

    I set-up my oscilloscope to measure the signals on my bench. The first picture is the signal when turned by hand. The 2nd is when the motor is running. The frequency matches the motor PWM frequency.
    IMG_20201011_103656.jpg IMG_20201011_103837.jpg IMG_20201011_101705.jpg

    After hours and hours of trouble shooting and trying different settings.... I found the issue: I got the wrong firmware version. If you are looking into ODrives, download THIS version and only THIS version!
    https://drive.google.com/drive/folders/135nR860-Wdb2QgAaqxzzxanJe04ZAtfT
    Also use the config you will find there. Friendly provided by @SilentChill :)

    Only thing you need to do:
    change "odrv0.config.brake_resistance" to your resistor value.
    Run "odrv0.axis0.requested_state = AXIS_STATE_FULL_CALIBRATION_SEQUENCE"
    Set "odrv0.axis0.controller.config.trav_upper_lim" to your values:
    cpr * (stroke / screw pitch)
    example: 8192 * (460 mm / 5 mm) = 753664
    Don't forget to save! "odrv0.save_configuration()"

    Even with this firmware, I noticed some spikes. Fix was easy. Just added some ferrit to the motor leads. This reduced the EMI by a significant amount.
    IMG_20201013_190634.jpg IMG_20201013_190801.jpg IMG_20201013_190920.jpg

    This is now the current build state:
    IMG_20201017_141441.jpg

    Everything looked good. Index search went good, too: Actuator does 2 turns outwards, 1 turn back, 1 turn outwards. The initial two turn are a bit jerky, but that's normal. I connected Mover and... the actuators went the wrong way. After two or three hours of searching and looking into things, I decided to change the A/B encoder cables. And this was the issue.
    I think the software ignores the A/B signal direction when looking for the index signal. It just goes into one direction. Only after the index is found the software looks at the A/B signal. And if they are switched around, the actuator goes the wrong direction.

    Got it moving yesterday:



    But now I got EMI issues again. The USB connections drop out when all three ODrives are active. So I have to order some good USB cables and a good powered USB Hub.

    Big thanks at this point to SilentChill for his help and patience :)
    • Like Like x 5
    • Winner Winner x 2