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

Long time lurker, First time poster

Expand Messages
  • johntyson@bellsouth.net
    Hello everyone! I recently saw a presentation on rapid prototyping and was instantly hooked! After some research I quickly decided that since I didn t have
    Message 1 of 7 , Jul 1, 2009
    • 0 Attachment
      Hello everyone!

      I recently saw a presentation on rapid prototyping and was instantly
      hooked! After some research I quickly decided that since I didn't
      have 15G's burning a hole in my pocket that I would have to DIY it.
      This forum has been an invaluable resource.

      I've gathered/scavenged all my parts and begun building (pics in the
      photo section).

      I started with a dusty old Lexmark 5700, stripped it and deduced its
      workings. I was going to use drawer slides but lucked into some very
      nice scavenged rails. I haven't mounted the printer yet as I'm still
      working on the electronics. For the powder roller I'm using the photoreceptor drum from an HP toner cart. It's about 1 inch diameter,
      smooth, lightweight (it's hollow), and free.

      The original paper feed stepper didn't have nearly enough torque to
      move a gantry so I decided to couple the original feed motion to a
      more powerful stepper with an encoder. I didn't have one and was too
      cheap to buy one so I made one from an opto-interrupter and a piece of film. Since the number of ticks on the wheel was custom, I was
      able to (in theory) compensate for the difference in the diameter of
      the feed roller and the new timing belt cog. Actually I'm off by ¼ tep per revolution but I'm hoping that I can fix that with the
      printer calibration.

      The paper path sensor was a little lever operated flag that went into
      an opto-interrupter. To fool it I placed another interrupter on top
      of it. Now I can just switch the power on and off to it and the
      printer thinks the lever has been tripped. I'm going to control the
      recharge cycle with g-code. I have a very old and mostly useless
      computer that can run TurboCNC.

      The print cycle should work like this:
      The printer hardware takes care of the printing of a layer. When the
      gantry reaches the end of the page it trips a limit switch which
      signals the computer to take over. The power to the encoder is cut
      and the power to the paper path sensor is cut. The printer then tries
      to feed the next sheet of paper and eventually decides it is out of
      paper. Meanwhile control of the gantry stepper has been given to
      TurboCNC which is merrily running the recharge cycle (running the
      roller motor, adjusting bin heights, and returning the gantry to home
      position). When it's done, TurboCNC repowers the encoder and paper
      path sensor and the printer has control and thinks more paper has
      been loaded. TurboCNC then waits for the next handshake and does it
      all again.

      I've tested the electronics on the bench and everything seems to be
      in order. My next step is to mount the printer on the rails and build my bins.

      I'll post more pics as I progress. Any questions or comments are welcome.

      John
    • Boman33
      Welcome John, It sounds like you are making great progress. I am in favor of divide and conquer to simplify things. From your description of the progress so
      Message 2 of 7 , Jul 1, 2009
      • 0 Attachment

        Welcome John,

        It sounds like you are making great progress.

         

        I am in favor of divide and conquer to simplify things.  From your description of the progress so far, a very good test would be to place a paper at the correct height and print a page of text.  Then print a photo.  If there are some problems hiding it is much easier to find them now then later with many more variables to worry about.

        Good Luck,

        Bertho

         

        From: johntyson@...  Sent: Wednesday, July 01, 2009 12:37
        Hello everyone!

        I recently saw a presentation on rapid prototyping and was instantly
        hooked! After some research I quickly decided that since I didn't
        have 15G's burning a hole in my pocket that I would have to DIY it.
        This forum has been an invaluable resource.

        I've gathered/scavenged all my parts and begun building (pics in the
        photo section).

        I started with a dusty old Lexmark 5700, stripped it and deduced its
        workings. I was going to use drawer slides but lucked into some very
        nice scavenged rails. I haven't mounted the printer yet as I'm still
        working on the electronics. For the powder roller I'm using the photoreceptor drum from an HP toner cart. It's about 1 inch diameter,
        smooth, lightweight (it's hollow), and free.

        The original paper feed stepper didn't have nearly enough torque to
        move a gantry so I decided to couple the original feed motion to a
        more powerful stepper with an encoder. I didn't have one and was too
        cheap to buy one so I made one from an opto-interrupter and a piece of film. Since the number of ticks on the wheel was custom, I was
        able to (in theory) compensate for the difference in the diameter of
        the feed roller and the new timing belt cog. Actually I'm off by ¼ tep per revolution but I'm hoping that I can fix that with the
        printer calibration.

        The paper path sensor was a little lever operated flag that went into
        an opto-interrupter. To fool it I placed another interrupter on top
        of it. Now I can just switch the power on and off to it and the
        printer thinks the lever has been tripped. I'm going to control the
        recharge cycle with g-code. I have a very old and mostly useless
        computer that can run TurboCNC.

        The print cycle should work like this:
        The printer hardware takes care of the printing of a layer. When the
        gantry reaches the end of the page it trips a limit switch which
        signals the computer to take over. The power to the encoder is cut
        and the power to the paper path sensor is cut. The printer then tries
        to feed the next sheet of paper and eventually decides it is out of
        paper. Meanwhile control of the gantry stepper has been given to
        TurboCNC which is merrily running the recharge cycle (running the
        roller motor, adjusting bin heights, and returning the gantry to home
        position). When it's done, TurboCNC repowers the encoder and paper
        path sensor and the printer has control and thinks more paper has
        been loaded. TurboCNC then waits for the next handshake and does it
        all again.

        I've tested the electronics on the bench and everything seems to be
        in order. My next step is to mount the printer on the rails and build my bins.

        I'll post more pics as I progress. Any questions or comments are welcome.

        John

      • johntyson@bellsouth.net
        Bertho, I was already thinking along those lines. This weekend I hope to have the printer mounted on the rails so I can test print some pages on paper. I m not
        Message 3 of 7 , Jul 1, 2009
        • 0 Attachment
          Bertho,

          I was already thinking along those lines.
          This weekend I hope to have the printer mounted on the rails so I
          can test print some pages on paper. I'm not as concerned with
          the image being stretched, I can adjust for that. What I'm hoping
          for is tight registration from one layer to the next. I'll print
          some register marks for a few cycles on the same piece of paper
          and see how they line up. I'll let you know how it turns out.

          John

          --- In diy_3d_printing_and_fabrication@yahoogroups.com, "Boman33" <boman33@...> wrote:
          >
          > Welcome John,
          >
          > It sounds like you are making great progress.
          >
          >
          >
          > I am in favor of divide and conquer to simplify things. From your
          > description of the progress so far, a very good test would be to place a
          > paper at the correct height and print a page of text. Then print a photo.
          > If there are some problems hiding it is much easier to find them now then
          > later with many more variables to worry about.
          >
          > Good Luck,
          >
          > Bertho
          >
        • M. Ganter
          John, it is worth spending a good amount of time on layer-to-layer registration. We ve had two teams of engineering students work on the personal 3D printer
          Message 4 of 7 , Jul 1, 2009
          • 0 Attachment
            John, it is worth spending a good amount of time on layer-to-layer
            registration. We've had two teams of engineering students work on
            the personal 3D printer (P3P) concept. The BIG problem has always
            been layer-to-layer registration.

            Also, I've got to ask the general members of this group a quick
            question, "Do anyone have a P3P/DIY 3DP system working yet?"

            Cheers ....M

            johntyson@... wrote:
            > Bertho,
            >
            > I was already thinking along those lines.
            > This weekend I hope to have the printer mounted on the rails so I
            > can test print some pages on paper. I'm not as concerned with
            > the image being stretched, I can adjust for that. What I'm hoping
            > for is tight registration from one layer to the next. I'll print
            > some register marks for a few cycles on the same piece of paper
            > and see how they line up. I'll let you know how it turns out.
            >
            > John
            >
            > --- In diy_3d_printing_and_fabrication@yahoogroups.com, "Boman33" <boman33@...> wrote:
            >> Welcome John,
            >>
            >> It sounds like you are making great progress.
            >>
            >>
            >>
            >> I am in favor of divide and conquer to simplify things. From your
            >> description of the progress so far, a very good test would be to place a
            >> paper at the correct height and print a page of text. Then print a photo.
            >> If there are some problems hiding it is much easier to find them now then
            >> later with many more variables to worry about.
            >>
            >> Good Luck,
            >>
            >> Bertho
            >>
            >
            >
            >

            --
            ========================================
            Organically Blown Glass (tm)
            http://425glass.com

            http://feeds2.feedburner.com/425glasscom
          • buzz
            ... I have recently completed what I believe is the first Rep(St)rap in Australia, and am merrily printing imperfect shot-glasses and calibration test/s,
            Message 5 of 7 , Jul 1, 2009
            • 0 Attachment
              >Also, I've got to ask the general members of this group a quick
              >question, "Do anyone have a P3P/DIY 3DP system working yet?"
              >Cheers ....M

              I have recently "completed" what I believe is the first Rep(St)rap in Australia, and am merrily printing imperfect shot-glasses and calibration test/s, etc.    Seems to be working well so-far.   Does that count as a 3D Printer in your books?

              I'm still calibrating/tweaking it, but I am currently extruding about 2000mm /minute of linear ABS output.  (ie printing a shot-glass takes approx 15 minutes )

              Buzz.
            • vrsculptor
              Curious, the registration problems that you saw... we they side to side or length wise. Roger
              Message 6 of 7 , Jul 2, 2009
              • 0 Attachment
                Curious, the registration problems that you saw... we they side to side or length wise.

                Roger

                > John, it is worth spending a good amount of time on layer-to-layer
                > registration. We've had two teams of engineering students work on
                > the personal 3D printer (P3P) concept. The BIG problem has always
                > been layer-to-layer registration.
              • M. Ganter
                The registration problem is lengthwise. The side-to-side registration is handled by the printer. ...M ... -- ========================================
                Message 7 of 7 , Jul 3, 2009
                • 0 Attachment
                  The registration problem is lengthwise. The side-to-side registration
                  is handled by the printer. ...M

                  vrsculptor wrote:
                  > Curious, the registration problems that you saw... we they side to side or length wise.
                  >
                  > Roger
                  >
                  >> John, it is worth spending a good amount of time on layer-to-layer
                  >> registration. We've had two teams of engineering students work on
                  >> the personal 3D printer (P3P) concept. The BIG problem has always
                  >> been layer-to-layer registration.
                  >
                  >
                  >

                  --
                  ========================================
                  Organically Blown Glass (tm)
                  http://425glass.com

                  http://feeds2.feedburner.com/425glasscom
                Your message has been successfully submitted and would be delivered to recipients shortly.