SCARA Arm 3D Printer


[Quentin Harley] must really have wanted to test his snuff when it comes to mechanical engineering. He’s been hard at work for a couple of years now designing his own SCARA arm 3D printer. That link leads to a recent summary article in which he shows off the build as seen above. It’s not fully functional yet, but he’s at the point where it’s time to develop the driver circuitry and firmware so he’s close. His blog is dedicated to this single project so click around and see what he went through along the journey.

The SCARA arm is seen in blue, using a couple of stepper motors to move the extruder mount along the x and y axes. The bed itself moves along the Z axis via two precision rods with a threaded rod in the center. As you can see, some of the parts are made of wood, and he used PVC for the cross supports between the upper and lower base platforms. But the majority of the build uses 3D printed parts, including the arms, drive gears, and mounting brackets.

[Thanks Peter]

29 thoughts on “SCARA Arm 3D Printer

        1. Yes, that is always possible. You have to design your system in such a way to make step skips unlikely.
          Morgan is no heavyweight tool swinging cnc or pnp. The motors are also driven in microstep mode, making skips highly unlikely (given that the forcve required for movement does not exceed the motor torque at the selected driver current)
          All of the RepRAP fdm machines currently use this control system, and is working just fine at the moment. Even the deltabot, Rosstock-by Johann Rocholl.

  1. Soon… he will discover the math needed to run this thing accurately is impractical to implement with bad tolerances. He is not the first to fail at this design, but it is a nice build attempt.

    1. @lol

      Thanks for your concers, but you will excuse me if I remain optimistic of my current design. You will know soon enough if I prove myself right (or wrong as it were)


      1. @quentinharley
        You will need to get some real industrial servo control loops to get this type of arm even remotely accurate. These were common for 1980s pick-and-place DIP insertion because they could wobble the chip alignment to inaccurate PCB holes.

        There are several designs over the years that failed to achieve even comparative results to primitive gantry based systems. Even $1700 worth of professional industrial servos won’t fix the problems.

        Maybe you guys are more qualified than the other engineers that failed to get it to work properly.

        Best of luck…

        1. Hi lol

          Thanks. I will need all the luck I can get.

          No, not more qualified at all, but living in an era where microcontollers run circles around any of the computers of the 80’s. This is hihgly unfair to those poor engineers that had to pioneer using the tools of the day, but I am not competing with them at all. Morgan is my pet project. I don’t care if it takes me 10 years to complete, but I have a feeling I am getting close. You should have seen some of the previous prototypes…

          See you at the finish line ;-)

  2. Hey Hackaday

    This is unexpected. Thanks.

    You will excuse me if I remain optimistic about the design… I did a couple of things differently than fellow SCARA hackers, and I would love to be the one proving myself right (or wrong…)


      1. Perhaps consider “inverting” the mechanism so arms point inwards from driven concentric drums. Then little need for rhombus, same effect achievable with a “V” [from hinges on the drums, to the head]. accessible table size = circle that fits inside the drums. accuracy would suffer in the middle with a two arm design. add a 3rd drum&arm if that needs dealt with.

          1. I like the concept of the link you posted, and one of my earlier prototypes followed that arrangent, but the weak link in this case is the weight, and the flexibility of the plastic – not a great combination for reprap. To bring cost down, I also needed to strip away some belt, and the toothbelt pulleys.

            I currently only need two pieces of open ended belt, and two pulleys. The drive wheels are printable, and not toothed.

            It looks like you are very experienced in the field of robots. Thanks for your inputs

      1. 1 1 ms 1 ms 1 ms []
        2 3 ms 3 ms 3 ms
        3 3 ms 3 ms 3 ms
        4 3 ms 4 ms 2 ms []
        5 30 ms 29 ms 36 ms []
        6 134 ms 120 ms 119 ms []
        7 117 ms 117 ms 117 ms []
        8 123 ms 130 ms 121 ms [
        9 * * * Request timed out.
        10 * * * Request timed out.
        11 * * * Request timed out.
        12 * * * Request timed out.
        13 * * * Request timed out.
        14 * * * Request timed out.

  3. I managed to get SCARA inverse kinematics running on RAMPS with Marlin. Check out my new post…

    Still needs some tuning, but it is doing exactly what I wanted for this point in the game!


Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.