8 Replies Latest reply on Jul 11, 2021 10:00 AM by dougw

    Would You Like To RoadTest a Servo Driver for a Robotic Gripper?

    rscasny

       

      I was approached by a supplier who would like to have its servo driver for a robotic gripper roadtested.

       

      Here's an image and the basic description:

       

      The TMCM-1617-GRIP-REF is an open-source hardware reference design for the TMCM-1617 BLDC servo driver. To be used in robotic gripper applications, the board is designed in a standard gripper electronics form-factor. It is able to control a BLDC motor via EtherCAT®, IO-Link®, or RS485 using Trinamic's TMCL protocol. In addition, the board features 1 configurable analog output and 1 configurable analog input, using Maxim’s MAX22000 industrial configurable analog I0 device and 4 configurable digital input, digital outputs, using Maxim’s MAX14906 device.

       

       

      My question is how would you choose to roadtest it? Would you build a robotic gripper? Do you need a bldc motor? encoder? hall sensor?

       

      It sounds like a great product, but depending on what you would like to do with it the roadtest could be simple to complex.

       

      What would you need beside the servo driver to roadtest this product?

       

      This page has the documentation: https://www.trinamic.com/support/eval-kits/details/tmcm-1617-grip-ref/

       

      Sincerely,

       

      Randall Scasny

      RoadTest Program Manager

        • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
          dougw

          I might be interested if I had a robot of this caliber, but I am a long way from having an appropriate platform.

          • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
            michaelkellett

            This poll seems to exist twice on my E14 - once with the description of the board and Doug's comment, and  and once without.

             

            Apart from that, the problem is that this board seems to be aimed at a very specialised market.

             

            There is an article about it here:

            https://www.automation-mag.com/news/41217-trinamic%E2%80%99s-open-source-reference-design-shrinks-and-speeds-development…

             

            Interesting, but - I think you need a £100k robot to test it on, not to mention the robot test lab etc.

             

            This just isn't viable for the typical RoadTester - and I doubt very much if the economics of RoadTesting would work out either - a free £500 board isn't going to count for much in a £1k+ per day test lab.

             

            MK

            2 of 2 people found this helpful
            • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
              14rhb

              I don't know much about such control mechanisms but would like to learn more from a hobbyist viewpoint. As michaelkellett mentions, the higher costs of this industrial board may mean I am not the ideal candidate. That said I would still apply with the possibility I am chosen and can (1) help Trinamic identify any product documentation and software issues from a newcomers perspective and (2)  to learn something new.

               

              A motor and load sensor would help make a better roadtest package. I'd have a look and suggest something suitable in my application for my intended project. The proximity sensor isn't such and issue (if needed for the roadtest project) as it would be made from IR LED etc. "Anything else?" My existing PSU is 15v 3A and therefore a item such as this 30v 5A PSU from CPC Farnell would also be very helpful .

               

              In terms of a roadtest project I would plan to design a generic bench-mounted gripper using OpenSCAD and print using my 3D printer, which mounts the supplied motor, load cell and proximity sensor. This design would also include a simple servo control to lift the gripper. Using a RS485 driver board I'd utilise the Trinamic's TMCL libraries to control the board - probably settling on the RPi and C. I may decide however to use Labview from a PC as I see that is also supported. The gripper would be tested against a range of sample items of various composition to see how the software can utilise the proximity and motor control allowing faster approaches to the test items before using torque control before adjusting the final grip (with feedback from the load cell). A latter stage once gripped would be to power the servo and undertake a controlled lift of that item, pause whilst in air, lower and then release grip. The ultimate test I feel would be to lift an egg without any breakage !

               

              As mentioned, I'm just intrigued about such technology and my approach is more akin to learning robotics in a University practical session. I don't currently have a huge yearning for build anything useful - but quickly can think of some cool projects such as a gripper in an automatic wire stripper.

               

              I hope that helps answer some of your questions

              • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
                robogary

                Artie Johnson would say....verrrrrry interrresting......

                A flyby of the spec sheet sounds like a stepper motor can be used, or a DC brushless with an encoder for vector control of torque (hopefully self tunes )

                 

                Could do some testing without a gripper, but having a gripper makes more sense. A load cell feedback (on gripper touch point) to the controller makes sense too.

                Does the vendor have a test GUI to send RS485, EtherCAT®, IO-Link® using vendors protocol on a PC to this controller ?  Configuration software with self tuning & calibration ?

                Like other commenters have mentioned, the material list and knowledge required for RoadTesting seems fairly significant for a layman, but E14 has some pretty knowledgeable folks available.

                • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
                  rsc

                  Looks interesting, but you'd need a motor, sensor(s)/encoder, and some cables just to start it up.  Is the vendor looking for feedback on hardware/software or interfacing?

                  Scott

                  2 of 2 people found this helpful
                  • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
                    kmikemoo

                    It's very cool - but I don't think that I could build a suitable gripper.

                    • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
                      Fred27

                      I've got some experience with using robotic grippers - specifically this UR5e robot and Robotiq gripper. Unfortunately I no longer have access to them as development has finished and they're now being used in their final installation.

                       

                      This strikes me as a very difficult product to effectively road test. The hardest parts about using this would be the physical build of the gripper and interfacing with the robot in a way that proves useful. Unless you just happened to have a robot already and be trying to build your own gripper this would be difficult. Did the manufacturer give any suggestions as to what they were expecting?

                      3 of 3 people found this helpful
                      • Re: Would You Like To RoadTest a Servo Driver for a Robotic Gripper?
                        dougw

                        I actually have a gripper and arm still unassembled (waiting for a suitable application project) but that system is designed to use standard hobby PWM "servo" motors.

                        The company I work for makes big grippers to handle explosives remotely (no hobby parts involved), but those systems are already in production.

                        1 of 1 people found this helpful