Loading ...
Sorry, an error occurred while loading the content.

Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

Expand Messages
  • cob674@yahoo.com
    I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the
    Message 1 of 11 , Aug 1, 2011
    • 0 Attachment
      I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
      Regards,
      Doyle

      Sent via DROID X


      -----Original message-----
      From: david_scorbot <david.scorbot@...>
      To:
      ScorbotUserGroup@yahoogroups.com
      Sent:
      Sun, Jul 31, 2011 09:15:43 GMT+00:00
      Subject:
      [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

      March 2011

      Notes from testing Scorbot motors with an aftermarket servo motor drive,
      using both Scorbot motor encoder and aftermarket encoder.

      ------------------------------------------------------------------------

      Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                to determine if a higher resolution aftermarket encoder could be used with the motors.


      Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                Scorbot 3-slot encoder was tested and did not work.
                The 6-slot encoder was not tested, but is unlikely to work either.


      ------------------------------------------------------------------------

      Hardware used:


      - G320X DC servo motor drive, by Geckodrive

      - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

      - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


      *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

      *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
          It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Export/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


      ------------------------------------------------------------------------

      Photos:


      Breakout Boards for DB50 robot arm cable
      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


      Scorbot Motor Encoders
      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


      CUI Motor Encoder
      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


      Geckodrive
      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


      ------------------------------------------------------------------------

      Scorbot motor voltage:

      The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

      Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
      24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


      ------------------------------------------------------------------------

      Power supply:


      Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

      2 x 12v AC transformers I had laying around
      - connected in series
      - bridge rectifier
      - that made about 24v DC
      - then a big cap, which now shows 32v DC

      Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


      ------------------------------------------------------------------------

      Wiring the motors:

      One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
      A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

      The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
      Wiring pinout was in the Scorbot manual.

      Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

      D50 interface module products from Phoenix Contact
      http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

      BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
      http://www.winfordeng.com/products/brkdd50.php 

      Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


      ------------------------------------------------------------------------

      Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



      On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
      This also disables the 'fault detection' feature but it won't work otherwise.

      Wired the motor to the Geckodrive, with
      Encoder LED0 --> channel A
      Encoder LED1 --> channel B
      but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

      The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

      Connected up the motor. Can step the shoulder up/down.
      With gain on Gecko set to mid-range, there is noticeable dithering noise.
      Set the gain to lowest, using DIP switches, and dithering noise is gone.

      Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

      Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

      The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
      The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


      ------------------------------------------------------------------------

      Scorbot spare motor (3-slot encoder) connected to Geckodrive:


      Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

      With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
      Stepping the motor doesn't make it move.


      ------------------------------------------------------------------------

      Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


      Remove existing encoder from the scorbot motor:

      To remove the outer plastic casing, remove the two hex screws. Size might be 7/64" or 3/32"
      The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

      Measured rear shaft with vernier, measured approx
      0.155 inches or 4mm


      Fitting the new encoder:

      Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
      DIP has plastic cover. Using a pocketknife can push switch through cover.
      Set it to 0001 = 1000 PPR
      note the DIP is reversed, with the Ones on the left.

      Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
      Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
      The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

      To tune the Gecko, HEDs option was turned off.
      Moved Gain up one notch to "Medium gain 2"
      Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
      With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
      Because of the gear box, the movement was still barely noticeable at the output shaft.
      By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

      One might think that the CUI Encoder's lowest count setting of 48 might be enough,
      but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

      When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

      So the CUI encoder does work, at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

      ------------------------------------------------------------------------


    • Alan
      So you want the commands issued from the Scorbase program? I just got my ER-3 running; I could set up another computer and “eavesdrop” on the
      Message 2 of 11 , Aug 1, 2011
      • 0 Attachment

        So you want the commands issued from the Scorbase program? 

         

        I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

         

        Alan  KM6VV

         

        From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
        Sent: Monday, August 01, 2011 10:55 AM
        To: ScorbotUserGroup@yahoogroups.com
        Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

         




        I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
        Regards,
        Doyle

        Sent via DROID X



        -----Original message-----

        From: david_scorbot <david.scorbot@...>
        To:
        ScorbotUserGroup@yahoogroups.com
        Sent:
        Sun, Jul 31, 2011 09:15:43 GMT+00:00
        Subject:
        [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

        March 2011

        Notes from testing Scorbot motors with an aftermarket servo motor drive,
        using both Scorbot motor encoder and aftermarket encoder.

        ------------------------------------------------------------------------

        Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                  to determine if a higher resolution aftermarket encoder could be used with the motors.


        Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                  The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                  Scorbot 3-slot encoder was tested and did not work.
                  The 6-slot encoder was not tested, but is unlikely to work either.


        ------------------------------------------------------------------------

        Hardware used:


        - G320X DC servo motor drive, by Geckodrive

        - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

        - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


        *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

        *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
            It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Export/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


        ------------------------------------------------------------------------

        Photos:


        Breakout Boards for DB50 robot arm cable
        http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


        Scorbot Motor Encoders
        http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


        CUI Motor Encoder
        http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


        Geckodrive
        http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


        ------------------------------------------------------------------------

        Scorbot motor voltage:

        The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

        Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
        24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


        ------------------------------------------------------------------------

        Power supply:


        Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

        2 x 12v AC transformers I had laying around
        - connected in series
        - bridge rectifier
        - that made about 24v DC- then a big cap, which now shows 32v DC

        Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


        ------------------------------------------------------------------------

        Wiring the motors:

        One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
        A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

        The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
        Wiring pinout was in the Scorbot manual.

        Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

        D50 interface module products from Phoenix Contact
        http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

        BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
        http://www.winfordeng.com/products/brkdd50.php 

        Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


        ------------------------------------------------------------------------

        Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



        On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
        This also disables the 'fault detection' feature but it! won't work otherwise.

        Wired the motor to the Geckodrive, with
        Encoder LED0 --> channel A
        Encoder LED1 --> channel B
        but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

        The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

        Connected up the motor. Can step the shoulder up/down.
        With gain on Gecko set to mid-range, there is noticeable dithering noise.
        Set the gain to lowest, using DIP switches, and dithering noise is gone.

        Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

        Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

        The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
        The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


        ------------------------------------------------------------------------

        Scorbot spare motor (3-slot encoder) connected to Geckodrive:


        Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

        With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
        Stepping the motor doesn't make it move.


        ------------------------------------------------------------------------

        Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


        Remove existing encoder from the scorbot motor:

        To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
        The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

        Measured rear shaft with vernier, measured approx
        0.155 inches or 4mm


        Fitting the new encoder:

        Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
        DIP has plastic cover. Using a pocketknife can push switch through cover.
        Set it to 0001 = 1000 PPR
        note the DIP is reversed, with the Ones on the left.

        Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
        Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
        The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

        To tune the Gecko, HEDs option was turned off.
        Moved Gain up one notch to "Medium gain 2"
        Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
        With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
        Because of the gear box, the movement was still barely noticeable at the output shaft.
        By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

        One might think that the CUI Encoder's lowest count setting of 48 might be enough,
        but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

        When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

        So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

        ------------------------------------------------------------------------




      • Eran Gal-Or
        Hi Alan, Have you managed to run Scorbot controller/arm using arduino? if so, how? Thanks Eran ________________________________ From: Alan
        Message 3 of 11 , Aug 1, 2011
        • 0 Attachment

          Hi Alan,

          Have you managed to run Scorbot controller/arm using arduino? if so, how?

          Thanks
          Eran

          From: Alan <KM6VV@...>
          To: ScorbotUserGroup@yahoogroups.com
          Sent: Mon, August 1, 2011 9:14:59 PM
          Subject: RE: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

           

          So you want the commands issued from the Scorbase program? 

           

          I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

           

          Alan  KM6VV

           

          From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
          Sent: Monday, August 01, 2011 10:55 AM
          To: ScorbotUserGroup@yahoogroups.com
          Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

           




          I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
          Regards,
          Doyle

          Sent via DROID X



          -----Original message-----

          From: david_scorbot <david.scorbot@...>
          To:
          ScorbotUserGroup@yahoogroups.com
          Sent:
          Sun, Jul 31, 2011 09:15:43 GMT+00:00
          Subject:
          [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

          March 2011

          Notes from testing Scorbot motors with an aftermarket servo motor drive,
          using both Scorbot motor encoder and aftermarket encoder.

          ------------------------------------------------------------------------

          Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                    to determine if a higher resolution aftermarket encoder could be used with the motors.


          Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                    The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                    Scorbot 3-slot encoder was tested and did not work.
                    The 6-slot encoder was not tested, but is unlikely to work either.


          ------------------------------------------------------------------------

          Hardware used:


          - G320X DC servo motor drive, by Geckodrive

          - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

          - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


          *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

          *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
              It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Exp!%0d%0a%20ort/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


          ------------------------------------------------------------------------

          Photos:


          Breakout Boards for DB50 robot arm cable
          http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


          Scorbot Motor Encoders
          http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


          CUI Motor Encoder
          http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


          Geckodrive
          http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


          ------------------------------------------------------------------------

          Scorbot motor voltage:

          The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

          Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
          24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


          ------------------------------------------------------------------------

          Power supply:


          Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

          2 x 12v AC transformers I had laying around
          - connected in series
          - bridge rectifier
          - that made about 24v DC- then a big cap, which now shows 32v DC

          Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


          ------------------------------------------------------------------------

          Wiring the motors:

          One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
          A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

          The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
          Wiring pinout was in the Scorbot manual.

          Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

          D50 interface module products from Phoenix Contact
          http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

          BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
          http://www.winfordeng.com/products/brkdd50.php 

          Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


          ------------------------------------------------------------------------

          Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



          On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
          This also disables the 'fault detection' feature but it! won't work otherwise.

          Wired the motor to the Geckodrive, with
          Encoder LED0 --> channel A
          Encoder LED1 --> channel B
          but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

          The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

          Connected up the motor. Can step the shoulder up/down.
          With gain on Gecko set to mid-range, there is noticeable dithering noise.
          Set the gain to lowest, using DIP switches, and dithering noise is gone.

          Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

          Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

          The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
          The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


          ------------------------------------------------------------------------

          Scorbot spare motor (3-slot encoder) connected to Geckodrive:


          Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

          With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
          Stepping the motor doesn't make it move.


          ------------------------------------------------------------------------

          Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


          Remove existing encoder from the scorbot motor:

          To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
          The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

          Measured rear shaft with vernier, measured approx
          0.155 inches or 4mm


          Fitting the new encoder:

          Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
          DIP has plastic cover. Using a pocketknife can push switch through cover.
          Set it to 0001 = 1000 PPR
          note the DIP is reversed, with the Ones on the left.

          Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
          Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
          The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

          To tune the Gecko, HEDs option was turned off.
          Moved Gain up one notch to "Medium gain 2"
          Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
          With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
          Because of the gear box, the movement was still barely noticeable at the output shaft.
          By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

          One might think that the CUI Encoder's lowest count setting of 48 might be enough,
          but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

          When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

          So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

          ------------------------------------------------------------------------




        • DmitriA
          ... commands to ... unable to ... 4r to ... appreciated. ... I m not sure why would you want to send commands to the Scorbot controller if the
          Message 4 of 11 , Aug 1, 2011
          • 0 Attachment
            > I built a Sorbot 3 arm controller using an Arduino and L298 dual motor 
            > controllers (posted on Youtube). I would like to find the serial commands to 
            > send to the Scorbot controller. I have the DOS based software but unable to 
            > move past the Com1 problem. I can send "4S" to turn on an output an "4r" to 
            > turn number 4 output off. Any assistance in this is greatly appreciated.
            > Regards,
            > Doyle
            >

            I'm not sure why would you want to send commands to the Scorbot controller if the motors/encoders are hooked up to the Arduino setup? Sounds like you'd need to send commands to the Arduino instead? Or did you mean you wanted the Arduino-based controller to respond to the same commands that Scorbase software issues to the stock Scorbot controller? That would make sense...
            Anyway, can you post a link to the Youtube video you've mentioned? Maybe things will get more clear then.
            Cheers!

             
            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
            Building a robotics lab and an online robotics community keeps me busy :)
          • Doyle Maleche
            DmitriA, Thanks for the responce. Sorry for not being clear. My Scorbot/L298/Arduino interface is simple. The GUI sends joystick commands to the Arduino which
            Message 5 of 11 , Aug 1, 2011
            • 0 Attachment
              DmitriA,
               
              Thanks for the responce. Sorry for not being clear. My Scorbot/L298/Arduino interface is simple. The GUI sends joystick commands to the Arduino which then sends position and velocity controls the L298 for each motor. The encoders are NOT used in this setup.
              Yes, I would like to send serial commands to the Scorbot controller to mimic the Scorbot software. I have not 'sniffed' the serial commands (yet) and solicited for them as a quick answer.
              The link to the Youtube video is here;
              Thanks!
              Doyle

              From: DmitriA <rcdd@...>
              To: ScorbotUserGroup@yahoogroups.com
              Sent: Mon, August 1, 2011 11:27:08 AM
              Subject: [ScorbotUserGroup] Re: Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders



              > I built a Sorbot 3 arm controller using an Arduino and L298 dual motor 
              > controllers (posted on Youtube). I would like to find the serial commands to 
              > send to the Scorbot controller. I have the DOS based software but unable to 
              > move past the Com1 problem. I can send "4S" to turn on an output an "4r" to 
              > turn number 4 output off. Any assistance in this is greatly appreciated.
              > Regards,
              > Doyle
              >

              I'm not sure why would you want to send commands to the Scorbot controller if the motors/encoders are hooked up to the Arduino setup? Sounds like you'd need to send commands to the Arduino instead? Or did you mean you wanted the Arduino-based controller to respond to the same commands that Scorbase software issues to the stock Scorbot controller? That would make sense...
              Anyway, can you post a link to the Youtube video you've mentioned? Maybe things will get more clear then.
              Cheers!

               
              ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
              Building a robotics lab and an online robotics community keeps me busy :)

            • Doyle Maleche
              Thank Alan, Yes, I would like to send commands from an Arduino or PC to emulate the Scorbot program. I could also develop a GUI to run on current windows OS
              Message 6 of 11 , Aug 1, 2011
              • 0 Attachment
                Thank Alan,
                 
                Yes, I would like to send commands from an Arduino or PC to emulate the Scorbot program. I could also develop a GUI to run on current windows OS and distribute it for free.
                 
                What did you do to run your ER-3 and are you using an Arduino to control your ER-3
                 
                Warm regards,
                Doyle


                From: Alan <KM6VV@...>
                To: ScorbotUserGroup@yahoogroups.com
                Sent: Mon, August 1, 2011 11:14:59 AM
                Subject: RE: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders



                So you want the commands issued from the Scorbase program? 

                 

                I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

                 

                Alan  KM6VV

                 

                From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
                Sent: Monday, August 01, 2011 10:55 AM
                To: ScorbotUserGroup@yahoogroups.com
                Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                 




                I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
                Regards,
                Doyle

                Sent via DROID X



                -----Original message-----

                From: david_scorbot <david.scorbot@...>
                To:
                ScorbotUserGroup@yahoogroups.com
                Sent:
                Sun, Jul 31, 2011 09:15:43 GMT+00:00
                Subject:
                [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                March 2011

                Notes from testing Scorbot motors with an aftermarket servo motor drive,
                using both Scorbot motor encoder and aftermarket encoder.

                ------------------------------------------------------------------------

                Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                          to determine if a higher resolution aftermarket encoder could be used with the motors.


                Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                          The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                          Scorbot 3-slot encoder was tested and did not work.
                          The 6-slot encoder was not tested, but is unlikely to work either.


                ------------------------------------------------------------------------

                Hardware used:


                - G320X DC servo motor drive, by Geckodrive

                - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

                - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


                *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

                *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
                    It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Exp!%0d%0a%20ort/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


                ------------------------------------------------------------------------

                Photos:


                Breakout Boards for DB50 robot arm cable
                http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


                Scorbot Motor Encoders
                http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


                CUI Motor Encoder
                http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


                Geckodrive
                http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


                ------------------------------------------------------------------------

                Scorbot motor voltage:

                The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

                Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
                24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


                ------------------------------------------------------------------------

                Power supply:


                Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

                2 x 12v AC transformers I had laying around
                - connected in series
                - bridge rectifier
                - that made about 24v DC- then a big cap, which now shows 32v DC

                Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


                ------------------------------------------------------------------------

                Wiring the motors:

                One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
                A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

                The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
                Wiring pinout was in the Scorbot manual.

                Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

                D50 interface module products from Phoenix Contact
                http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

                BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
                http://www.winfordeng.com/products/brkdd50.php 

                Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


                ------------------------------------------------------------------------

                Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



                On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
                This also disables the 'fault detection' feature but it! won't work otherwise.

                Wired the motor to the Geckodrive, with
                Encoder LED0 --> channel A
                Encoder LED1 --> channel B
                but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

                The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

                Connected up the motor. Can step the shoulder up/down.
                With gain on Gecko set to mid-range, there is noticeable dithering noise.
                Set the gain to lowest, using DIP switches, and dithering noise is gone.

                Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

                Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

                The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
                The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


                ------------------------------------------------------------------------

                Scorbot spare motor (3-slot encoder) connected to Geckodrive:


                Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

                With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
                Stepping the motor doesn't make it move.


                ------------------------------------------------------------------------

                Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


                Remove existing encoder from the scorbot motor:

                To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
                The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

                Measured rear shaft with vernier, measured approx
                0.155 inches or 4mm


                Fitting the new encoder:

                Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
                DIP has plastic cover. Using a pocketknife can push switch through cover.
                Set it to 0001 = 1000 PPR
                note the DIP is reversed, with the Ones on the left.

                Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
                Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
                The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

                To tune the Gecko, HEDs option was turned off.
                Moved Gain up one notch to "Medium gain 2"
                Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
                With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
                Because of the gear box, the movement was still barely noticeable at the output shaft.
                By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

                One might think that the CUI Encoder's lowest count setting of 48 might be enough,
                but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

                When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

                So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

                ------------------------------------------------------------------------






              • Doyle Maleche
                Eran, I m using an Arduino Mega and three L298 motor controllers from Solarbotics.com to control my Scorbot Arm. You can view my newly uploaded video at
                Message 7 of 11 , Aug 1, 2011
                • 0 Attachment
                  Eran,
                   
                  I'm using an Arduino Mega and three L298 motor controllers from Solarbotics.com to control my Scorbot Arm. You can view my newly uploaded video at http://www.youtube.com/watch?v=qjfd1MgDbTw
                   
                  The encoders are not used, just position and velocity commands.
                   
                  Software: MS Visual Basic 6.0
                  Arduino Mega
                   
                  Enjoy!
                  Doyle


                  From: Eran Gal-Or <erangalor@...>
                  To: ScorbotUserGroup@yahoogroups.com
                  Sent: Mon, August 1, 2011 11:24:40 AM
                  Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders




                  Hi Alan,

                  Have you managed to run Scorbot controller/arm using arduino? if so, how?

                  Thanks
                  Eran

                  From: Alan <KM6VV@...>
                  To: ScorbotUserGroup@yahoogroups.com
                  Sent: Mon, August 1, 2011 9:14:59 PM
                  Subject: RE: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                   

                  So you want the commands issued from the Scorbase program? 

                   

                  I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

                   

                  Alan  KM6VV

                   

                  From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
                  Sent: Monday, August 01, 2011 10:55 AM
                  To: ScorbotUserGroup@yahoogroups.com
                  Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                   




                  I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
                  Regards,
                  Doyle

                  Sent via DROID X



                  -----Original message-----

                  From: david_scorbot <david.scorbot@...>
                  To:
                  ScorbotUserGroup@yahoogroups.com
                  Sent:
                  Sun, Jul 31, 2011 09:15:43 GMT+00:00
                  Subject:
                  [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                  March 2011

                  Notes from testing Scorbot motors with an aftermarket servo motor drive,
                  using both Scorbot motor encoder and aftermarket encoder.

                  ------------------------------------------------------------------------

                  Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                            to determine if a higher resolution aftermarket encoder could be used with the motors.


                  Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                            The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                            Scorbot 3-slot encoder was tested and did not work.
                            The 6-slot encoder was not tested, but is unlikely to work either.


                  ------------------------------------------------------------------------

                  Hardware used:


                  - G320X DC servo motor drive, by Geckodrive

                  - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

                  - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


                  *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

                  *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
                      It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Exp!%0d%0a%20ort/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


                  ------------------------------------------------------------------------

                  Photos:


                  Breakout Boards for DB50 robot arm cable
                  http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


                  Scorbot Motor Encoders
                  http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


                  CUI Motor Encoder
                  http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


                  Geckodrive
                  http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


                  ------------------------------------------------------------------------

                  Scorbot motor voltage:

                  The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

                  Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
                  24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


                  ------------------------------------------------------------------------

                  Power supply:


                  Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

                  2 x 12v AC transformers I had laying around
                  - connected in series
                  - bridge rectifier
                  - that made about 24v DC- then a big cap, which now shows 32v DC

                  Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


                  ------------------------------------------------------------------------

                  Wiring the motors:

                  One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
                  A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

                  The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
                  Wiring pinout was in the Scorbot manual.

                  Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

                  D50 interface module products from Phoenix Contact
                  http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

                  BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
                  http://www.winfordeng.com/products/brkdd50.php 

                  Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


                  ------------------------------------------------------------------------

                  Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



                  On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
                  This also disables the 'fault detection' feature but it! won't work otherwise.

                  Wired the motor to the Geckodrive, with
                  Encoder LED0 --> channel A
                  Encoder LED1 --> channel B
                  but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

                  The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

                  Connected up the motor. Can step the shoulder up/down.
                  With gain on Gecko set to mid-range, there is noticeable dithering noise.
                  Set the gain to lowest, using DIP switches, and dithering noise is gone.

                  Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

                  Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

                  The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
                  The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


                  ------------------------------------------------------------------------

                  Scorbot spare motor (3-slot encoder) connected to Geckodrive:


                  Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

                  With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
                  Stepping the motor doesn't make it move.


                  ------------------------------------------------------------------------

                  Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


                  Remove existing encoder from the scorbot motor:

                  To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
                  The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

                  Measured rear shaft with vernier, measured approx
                  0.155 inches or 4mm


                  Fitting the new encoder:

                  Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
                  DIP has plastic cover. Using a pocketknife can push switch through cover.
                  Set it to 0001 = 1000 PPR
                  note the DIP is reversed, with the Ones on the left.

                  Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
                  Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
                  The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

                  To tune the Gecko, HEDs option was turned off.
                  Moved Gain up one notch to "Medium gain 2"
                  Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
                  With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
                  Because of the gear box, the movement was still barely noticeable at the output shaft.
                  By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

                  One might think that the CUI Encoder's lowest count setting of 48 might be enough,
                  but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

                  When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

                  So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

                  ------------------------------------------------------------------------






                • Alan
                  Hi Eran, I haven’t. So far, I’ve just moved the servos a little with the controller box. I have more checkout to do. I’d like to throw out some ideas.
                  Message 8 of 11 , Aug 1, 2011
                  • 0 Attachment

                    Hi Eran,

                     

                    I haven’t.  So far, I’ve just moved the servos a little with the controller box.  I have more checkout to do. 

                     

                    I’d like to throw out some ideas.

                     

                    I’m more inclined to run the servos with a RoboClaw or Gecko drivers, and then my CNC controller program would run the arm. 

                     

                    A “driver box” built with L298 driver chips is certainly possible.  There are also “stock” CNC driver boards like from Geckodrive

                     

                    http://www.geckodrive.com/

                     

                    Geckodrive modules (a little pricy) are available for both stepper motors and Servo motors.  Both take step/direction commands. 

                     

                    RoboClaw or Gecko drivers do a “closed loop” control over the servo motors.  RoboClaw takes serial commands (or PWM), the Geckos take step/direction commands. 

                     

                    These may seem a little expensive, and indeed they are, being designed for CNC work.  I have geckos on both my CNC’d Sherline mill and lathe, and I drive them (step/direction) from my CNC controller program. 

                     

                    Mach1 and TurboCNC also issue step/ direction commands.

                     

                    But a “modular” approach could make things more useful.  The “driver box” could be driven from either an Arduino, or a PC.  Just thinking at this point! 

                     

                    Certainly simple enough to start out with PWM and some L298 drivers for the Arduino.  And others have been able to write code for the Arduino to generate step/ direction commands, and also parse G-code (CNC “block” instructions).

                     

                    OK, I’m getting a little ahead of myself!  Just being able to run Scorbase code would be useful. 

                     

                    Comments?

                     

                    Alan  KM6VV

                     

                    From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of Eran Gal-Or
                    Sent: Monday, August 01, 2011 11:25 AM
                    To: ScorbotUserGroup@yahoogroups.com
                    Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                     




                     

                    Hi Alan,

                     

                    Have you managed to run Scorbot controller/arm using arduino? if so, how?

                     

                    Thanks

                    Eran


                    From: Alan <KM6VV@...>
                    To: ScorbotUserGroup@yahoogroups.com
                    Sent: Mon, August 1, 2011 9:14:59 PM
                    Subject: RE: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                     

                    So you want the commands issued from the Scorbase program? 

                     

                    I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

                     

                    Alan  KM6VV

                     

                    From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
                    Sent: Monday, August 01, 2011 10:55 AM
                    To: ScorbotUserGroup@yahoogroups.com
                    Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                     



                    I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
                    Regards,
                    Doyle

                    Sent via DROID X



                    -----Original message-----

                    From: david_scorbot <david.scorbot@...>
                    To:
                    ScorbotUserGroup@yahoogroups.com
                    Sent:
                    Sun, Jul 31, 2011 09:15:43 GMT+00:00
                    Subject:
                    [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                    March 2011

                    Notes from testing Scorbot motors with an aftermarket servo motor drive,
                    using both Scorbot motor encoder and aftermarket encoder.

                    ------------------------------------------------------------------------

                    Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                              to determine if a higher resolution aftermarket encoder could be used with the motors.


                    Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                              The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                              Scorbot 3-slot encoder was tested and did not work.
                              The 6-slot encoder was not tested, but is unlikely to work either.


                    ------------------------------------------------------------------------

                    Hardware used:


                    - G320X DC servo motor drive, by Geckodrive

                    - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

                    - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


                    *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

                    *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
                        It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Exp!%0d%0a%20ort/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


                    ------------------------------------------------------------------------

                    Photos:


                    Breakout Boards for DB50 robot arm cable
                    http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


                    Scorbot Motor Encoders
                    http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


                    CUI Motor Encoder
                    http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


                    Geckodrive
                    http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


                    ------------------------------------------------------------------------

                    Scorbot motor voltage:

                    The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

                    Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
                    24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


                    ------------------------------------------------------------------------

                    Power supply:


                    Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

                    2 x 12v AC transformers I had laying around
                    - connected in series
                    - bridge rectifier
                    - that made about 24v DC- then a big cap, which now shows 32v DC

                    Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


                    ------------------------------------------------------------------------

                    Wiring the motors:

                    One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
                    A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

                    The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
                    Wiring pinout was in the Scorbot manual.

                    Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

                    D50 interface module products from Phoenix Contact
                    http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

                    BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
                    http://www.winfordeng.com/products/brkdd50.php 

                    Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


                    ------------------------------------------------------------------------

                    Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



                    On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
                    This also disables the 'fault detection' feature but it! won't work otherwise.

                    Wired the motor to the Geckodrive, with
                    Encoder LED0 --> channel A
                    Encoder LED1 --> channel B
                    but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

                    The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

                    Connected up the motor. Can step the shoulder up/down.
                    With gain on Gecko set to mid-range, there is noticeable dithering noise.
                    Set the gain to lowest, using DIP switches, and dithering noise is gone.

                    Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

                    Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

                    The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
                    The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


                    ------------------------------------------------------------------------

                    Scorbot spare motor (3-slot encoder) connected to Geckodrive:


                    Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

                    With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
                    Stepping the motor doesn't make it move.


                    ------------------------------------------------------------------------

                    Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


                    Remove existing encoder from the scorbot motor:

                    To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
                    The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

                    Measured rear shaft with vernier, measured approx
                    0.155 inches or 4mm


                    Fitting the new encoder:

                    Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
                    DIP has plastic cover. Using a pocketknife can push switch through cover.
                    Set it to 0001 = 1000 PPR
                    note the DIP is reversed, with the Ones on the left.

                    Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
                    Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
                    The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

                    To tune the Gecko, HEDs option was turned off.
                    Moved Gain up one notch to "Medium gain 2"
                    Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
                    With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
                    Because of the gear box, the movement was still barely noticeable at the output shaft.
                    By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

                    One might think that the CUI Encoder's lowest count setting of 48 might be enough,
                    but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

                    When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

                    So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

                    ------------------------------------------------------------------------






                  • Alan
                    I’ve done nothing so far! Just ran the control box in checking out the servos. A GUI would be nice! I have Arduinos running in other projects, and I’m
                    Message 9 of 11 , Aug 1, 2011
                    • 0 Attachment

                      I’ve done nothing so far!  Just ran the control box in checking out the servos. 

                       

                      A GUI would be nice! 

                       

                      I have Arduinos running in other projects, and I’m good with C.

                       

                      The “CNC approach” is of interest to me, and also doing (at some point) IK for the arm. 

                       

                      A “Servo driver box” with 12V – 48V power supply to run 6 or 8 DC servos would be of interest to me, and fit in with my CNC plans as well. 

                       

                      Alan  KM6VV

                       

                      From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of Doyle Maleche
                      Sent: Monday, August 01, 2011 12:18 PM
                      To: ScorbotUserGroup@yahoogroups.com
                      Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                       




                      Thank Alan,

                       

                      Yes, I would like to send commands from an Arduino or PC to emulate the Scorbot program. I could also develop a GUI to run on current windows OS and distribute it for free.

                       

                      What did you do to run your ER-3 and are you using an Arduino to control your ER-3

                       

                      Warm regards,

                      Doyle

                       


                      From: Alan <KM6VV@...>
                      To: ScorbotUserGroup@yahoogroups.com
                      Sent: Mon, August 1, 2011 11:14:59 AM
                      Subject: RE: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders


                      So you want the commands issued from the Scorbase program? 

                       

                      I just got my ER-3 running; I could set up another computer and “eavesdrop” on the communications.  I also have Arduino UNO boards available.

                       

                      Alan  KM6VV

                       

                      From: ScorbotUserGroup@yahoogroups.com [mailto:ScorbotUserGroup@yahoogroups.com] On Behalf Of cob674@...
                      Sent: Monday, August 01, 2011 10:55 AM
                      To: ScorbotUserGroup@yahoogroups.com
                      Subject: Re: [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                       



                      I built a Sorbot 3 arm controller using an Arduino and L298 dual motor controllers (posted on Youtube). I would like to find the serial commands to send to the Scorbot controller. I have the DOS based software but unable to move past the Com1 problem. I can send "4S" to turn on an output an "4r" to turn number 4 output off. Any assistance in this is greatly appreciated.
                      Regards,
                      Doyle

                      Sent via DROID X



                      -----Original message-----

                      From: david_scorbot <david.scorbot@...>
                      To:
                      ScorbotUserGroup@yahoogroups.com
                      Sent:
                      Sun, Jul 31, 2011 09:15:43 GMT+00:00
                      Subject:
                      [ScorbotUserGroup] Testing Geckodrive with Scorbot motors, using factory & aftermarket encoders

                      March 2011

                      Notes from testing Scorbot motors with an aftermarket servo motor drive,
                      using both Scorbot motor encoder and aftermarket encoder.

                      ------------------------------------------------------------------------

                      Purpose:   To test if a Scorbot robot arm could be controlled by an off-the-shelf servo motor drive, and
                                to determine if a higher resolution aftermarket encoder could be used with the motors.


                      Results:   The Geckodrive G320X can control a Scorbot motor with 20-hole encoder or aftermarket CUI brand encoder.
                                The 20-hole encoder in robots like the ER-4 appears to create sufficient resolution as-is, but using aftermarket encoders could be worthwhile for an older robot like the ER-3.
                                Scorbot 3-slot encoder was tested and did not work.
                                The 6-slot encoder was not tested, but is unlikely to work either.


                      ------------------------------------------------------------------------

                      Hardware used:


                      - G320X DC servo motor drive, by Geckodrive

                      - Pittman motor in shoulder of Scorbot ER4PC, with attached 20-step encoder  *1

                      - Pittman motor (formerly from conveyor) with aftermarket encoder AMT102-V & 5-pin latching cable assy by CUI Inc.  *2


                      *1  This particular 20-step encoder has 20 oval-shaped holes. There is also a newer 20-step encoder with 20 rectangular slots and this would likely produce a similar electrical output.

                      *2  The CUI encoder kit and cables are sold separetely:  http://www.digikey.com/us/en/ph/cui/amt.html 
                          It is a capacitive motor encoder, with adjustable resolution.  http://www.digikey.com/Web%20Exp!%0d%0a%20ort/Supplier%20Content/CUI_102/PDF/CUI_WP_Capacative_vs_Optical_Encoders.pdf?redirected=1 


                      ------------------------------------------------------------------------

                      Photos:


                      Breakout Boards for DB50 robot arm cable
                      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/105751163/pic/list 


                      Scorbot Motor Encoders
                      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/2128576455/pic/list 


                      CUI Motor Encoder
                      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1267894192/pic/list 


                      Geckodrive
                      http://groups.yahoo.com/group/ScorbotUserGroup/photos/album/1537466657/pic/list 


                      ------------------------------------------------------------------------

                      Scorbot motor voltage:

                      The manual says motors are 24 volt. But a voltage meter was connected to the shoulder motor, powered by the Scorpower ER4PC, which showed a running voltage of around 12 volts.

                      Connected spare Scorbot motor to 12v and it ran at a nice steady pace.
                      24v would be too fast, but the Geckodrive needs minimum 18v, so this will provide plenty of headroom for speed control.


                      ------------------------------------------------------------------------

                      Power supply:


                      Needed minimum 18 volts for Gecko motor drives, so built a 24v supply.

                      2 x 12v AC transformers I had laying around
                      - connected in series
                      - bridge rectifier
                      - that made about 24v DC- then a big cap, which now shows 32v DC

                      Power supply stores some punch after its turned off, so I use a spare motor to drain the cap.


                      ------------------------------------------------------------------------

                      Wiring the motors:

                      One was an auxiliary motor from a Scorbot conveyor. This came with a DB9 plug.
                      A matching DB9 connector was used, with wires soldered to the appropriate pins as per the Scorbot ER-3 manual ("Appendix F, The Motor Kit")

                      The other motor used was in the shoulder of a Scorbot ER4PC. The robot arm has a flying lead with male DB50 connector. A matching, female DB50 breakout board was sourced, to allow easy screw connection of wires.
                      Wiring pinout was in the Scorbot manual.

                      Only two such breakout boards could be found:  the FLK-D50/DFLK-D50 by Phoenix Contact, and the BRKDD50F by Winford Engineering. Both are rated at 2.5 Amps per connection, but Winford's was significantly less expensive.

                      D50 interface module products from Phoenix Contact
                      http://www.phoenixcontact.com.au/company/search.jsp?q=DB50&x=0&y=0 

                      BRKDD50F (Female DB50 socket breakout to screw terminals) by Winford Engineering
                      http://www.winfordeng.com/products/brkdd50.php 

                      Note, while rated at 2.5A per signal, the PCB track widths on the breakout board are very thin (1mm) because there's so many signals to route. The current draw of the Scorbot motors will need to be tested because, by comparison, the PCB track widths from the motor drivers inside the Scorpower box are much wider.


                      ------------------------------------------------------------------------

                      Scorbot shoulder motor (with 20-hole encoder) connected to Geckodrive:



                      On Gecko G320X, change DIP switch SW6 to ON for optical encoders.
                      This also disables the 'fault detection' feature but it! won't work otherwise.

                      Wired the motor to the Geckodrive, with
                      Encoder LED0 --> channel A
                      Encoder LED1 --> channel B
                      but reversing that doesnt seem to make any noticable difference while testing the Geckodrive.

                      The Geckodrive was tested as per the manual, with the motor terminals disconnected. When the motor shaft is moved out of position the status lights on the Gecko correctly changed from InPosition --> Error --> Fault.

                      Connected up the motor. Can step the shoulder up/down.
                      With gain on Gecko set to mid-range, there is noticeable dithering noise.
                      Set the gain to lowest, using DIP switches, and dithering noise is gone.

                      Be careful, as the manual states, if you adjust the P or D too high/low, the motor will start convulsing!

                      Also, if the motor sometimes "steps" by itself, you may have the step/direction control wires running parallel with the power wires, introducing electrical interference.

                      The control and precision created by the Geckodrive is impressive. When stationary, it's practically impossible to force the Scorbot arm out of position.
                      The setup was left powered-on for over 24 hours and no heat could be felt from the motor or Geckodrive. There was some heating when the motor was dithering more.


                      ------------------------------------------------------------------------

                      Scorbot spare motor (3-slot encoder) connected to Geckodrive:


                      Encoder line-count is too low. With the encoder connected to the Geckodrive, any forced movement of the output shaft puts the drive into a 'fault' state.

                      With the motor connected, the Geckodrive either faults straight away, or flickers between InPosition/Fault.
                      Stepping the motor doesn't make it move.


                      ------------------------------------------------------------------------

                      Scorbot spare motor, aftermarket AMT102-V encoder, connected to Geckodrive:


                      Remove existing encoder from the scorbot motor:

                      To remove the outer plastic c! asing, remove the two hex screws. Size might be 7/64" or 3/32"
                      The 3 slot wheel has 2 tiny grub screws. It's a 1.27mm hex/allen key. Only a good set will have that size, many only go down to a 1.5mm size.

                      Measured rear shaft with vernier, measured approx
                      0.155 inches or 4mm


                      Fitting the new encoder:

                      Need to set the resolution. Default is all DIPS off which is 0000 = 48 PPR
                      DIP has plastic cover. Using a pocketknife can push switch through cover.
                      Set it to 0001 = 1000 PPR
                      note the DIP is reversed, with the Ones on the left.

                      Following directions, fitted the 4mm shaft sleeve, using the tool to space it away from the back of the motor. Then fitted Shaft adapter using plastic tool.
                      Then the plastic base. The normal base has 4 holes in the right spot, but they're a bit smaller. You could enlarge them, but the hex screws from the Scorbot encoder will hit the new encoder. Might need to source some countersunk screws.
                      The encoder base is suprisingly a press fit over the end of the motor, so that was sufficient for testing.

                      To tune the Gecko, HEDs option was turned off.
                      Moved Gain up one notch to "Medium gain 2"
                      Tweaked the "P" and "D" pots back a bit to around 9-10 o'clock until dithering noise went quiet.
                      With the 1000 count encoder x 4 (quadrature encoding) the step movement was tiny. By enabling the x10 Step Multiplier on the Geckodrive a more noticeable movement of the motor shaft could be seen.
                      Because of the gear box, the movement was still barely noticeable at the output shaft.
                      By lowering the Gain to lowest via DIP switches I could raise the P and D a bit to 10-11 o-clock.

                      One might think that the CUI Encoder's lowest count setting of 48 might be enough,
                      but when tested it didn't work. The Geckodrive fault light was on and the motor was advancing by itself.

                      When the motor was run for a duration, the CUI encoder sounded a bit noisy, possibly because it was not properly attached/aligned.

                      So the CUI encoder does work! , at least on the 1000 line count setting. The resolution is overkill, and it would be best to firstly test how the CUI encoder could be permanently attached, but this is definately a viable and cheaper alternative to finding new Scorbot encoders to upgrade your old robot.

                      ------------------------------------------------------------------------



                       




                    Your message has been successfully submitted and would be delivered to recipients shortly.