Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Thursday June 11 2015, @01:52AM   Printer-friendly
from the big-boys-toys dept.

MakerBot's 3-D printers will soon be able to produce items that look like bronze, limestone, and wood, thanks to a new line of plastic-based composite materials shipping later this year. But the launch may be too little, too late: Entrepreneurs and artists interested in working with metal and wood are already embracing desktop milling machines that can handle the real deal.

The calculation is simple: Buy a MakerBot Replicator, the leading desktop 3-D printer, for $2,889, and you can produce plastic prototypes or the kind of trinkets that you might find in a Happy Meal. Buy a small-scale milling machine like the Othermill, which retails for $2,199, and you can make jewelry and mechanical parts out of everything from aluminum to walnut.

"Once you can cut metal, you can make things that last," says Danielle Applestone, chief executive of Other Machine Co. "For the first couple of months that I was working here, I was scared of cutting with metal. It was louder, I was worried I was going to break the tool. But as soon as I jumped in, it quickly became like wax to me."

"Metal is power, it really is," she says. "You don't go back."

It should be noted that MakerBot's base model also went from $400 to almost $3K when Stratasys acquired them.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 4, Interesting) by anubi on Thursday June 11 2015, @03:44AM

    by anubi (2828) on Thursday June 11 2015, @03:44AM (#194827) Journal

    I think most of the money is in your number 6.

    I have been toying with building a CNC machine myself, but I find myself with a chicken and egg situation. Same situation as when I got into Arduinos. I had to get a professionally made one first, but once I saw how that one worked, I leveraged off of it to spawn off other Arduino-compatible designs which were much more what I had in mind.

    For now, I find trying to make all the mechanical pieces of a CAD system to be a huge time taking effort. I would have to justify someone else's mass produced design. The Other Machine CNC mill looks really neat, but it appears to be only Apple driven. Something like that I kinda expect to be a USB plug-in to a PC, and run the existing open source CAM software already out there.

    Some more CNC machines and their software [cnczone.com] are discussed here.

    The Shapeoko [shapeoko.com] CNC milling kit ( around 1K$ ) looks pretty good to me... but I still like screw drives on the positioners - this one looks like it uses belts.

    --
    "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
    Starting Score:    1  point
    Moderation   +3  
       Interesting=3, Total=3
    Extra 'Interesting' Modifier   0  

    Total Score:   4  
  • (Score: 2) by tibman on Thursday June 11 2015, @04:26AM

    by tibman (134) Subscriber Badge on Thursday June 11 2015, @04:26AM (#194836)

    I have heard that belts are faster than threaded rods. I'm guessing accuracy is the tradeoff.

    --
    SN won't survive on lurkers alone. Write comments.
    • (Score: 3, Informative) by anubi on Thursday June 11 2015, @04:54AM

      by anubi (2828) on Thursday June 11 2015, @04:54AM (#194842) Journal

      Yes, the belts faster, simpler, and less expensive. As you noted, the tradeoff is accuracy.

      Its a function of the step size on the stepper motor as well as any stretch on the belt.

      ( One can get around the step size problem using DC servomotors and quadrature optical feedback disks ).

      A relatively fine-pitched screw may take several steps of the motor to travel .001" ( the exact ratio is given to the computer; from that the computer knows how many steps to index for any given position ).

      My planned modus operandi is to get the machine started and go to bed. I can provide means of telling if something is overheating or jammed and shut the whole shebang down if it encounters an anomaly. It would not surprise me if it takes the machine several hours to chew out some of the things I would like it to make for me.

      --
      "Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
      • (Score: 2) by c0lo on Thursday June 11 2015, @09:39AM

        by c0lo (156) Subscriber Badge on Thursday June 11 2015, @09:39AM (#194898) Journal

        A relatively fine-pitched screw may take several steps of the motor to travel .001"

        Huh? What's your desired precision? Because if you go that much down, you'll spend fortunes on ball-screws as actuators (acme screws will still have a backlash with a magnitude in this range [wikipedia.org], and recirculating ball screws are in general about 10 times more expensive than an acme thread).

        --
        https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
        • (Score: 2) by mhajicek on Thursday June 11 2015, @11:26AM

          by mhajicek (51) on Thursday June 11 2015, @11:26AM (#194915)

          There are anti-backlash nuts, you don't need ball screws.

          --
          The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
          • (Score: 2) by LoRdTAW on Thursday June 11 2015, @11:54AM

            by LoRdTAW (3755) on Thursday June 11 2015, @11:54AM (#194919) Journal

            Horses for courses. Small desktop mills can get away with inefficient acme screws as their loads are light and feed rates are very slow. Any faster or larger system is going to employ ball screws.

            • (Score: 2) by mhajicek on Thursday June 11 2015, @04:11PM

              by mhajicek (51) on Thursday June 11 2015, @04:11PM (#195016)

              Depends on what you mean by "small". Also depends on the anticipated cutting forces, duty cycle, and lifetime of the assembly. That said, you could start with lead screws and later upgrade to ballscrews once the machine is making money.

              --
              The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
              • (Score: 2) by c0lo on Thursday June 11 2015, @07:38PM

                by c0lo (156) Subscriber Badge on Thursday June 11 2015, @07:38PM (#195116) Journal

                and later upgrade to ballscrews once the machine is making money.

                Whoa there... we aren't talking about minting coins on CNCes, are we?
                (grin)

                --
                https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
                • (Score: 2) by mhajicek on Thursday June 11 2015, @10:42PM

                  by mhajicek (51) on Thursday June 11 2015, @10:42PM (#195175)

                  I have used my MaxNC to make coining dies. Not for counterfeiting mind you, just medieval style coins for reenactors.

                  --
                  The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
      • (Score: 5, Informative) by LoRdTAW on Thursday June 11 2015, @12:14PM

        by LoRdTAW (3755) on Thursday June 11 2015, @12:14PM (#194922) Journal

        Its a function of the step size on the stepper motor as well as any stretch on the belt.

        You would be surprised how accurate timing belts can be. I have talked with Aerotech, makers of high end motion control systems, about compact stages that use a brushless servo motor mounted 180°. The motor drives the ball screw through a timing belt and can still maintain 0.0001" inch or better accuracy. I also have a short acme screw Z stage here on a laser welder that has a brushless servo motor with a timing belt. Has no problems keeping 0.001" accuracy.

        Simple step motor drivers from an arduino are stuck at 200 steps/rev or what ever the motor count is, 200 is almost a standard. Good drivers can microstep which employs a current control method to hold the rotor between the poles. Some drivers can easily divide a single step into a hundred or more steps. So a step motor can have thousands of steps/rev. I worked with an X-Y table from newmark systems that was driven by a Galil motion controller and IMS all-in-one stepper drivers (IMS = Intelligent motion systems, now owned by Schneider electric). The stages screw pitch was 10 revs an inch and the micro stepper was set up for 20,000 steps/rev! I had to move the system 200,000 counts to go one inch. So you can go really high resolution with steppers if you need.

        • (Score: 1) by WillAdams on Thursday June 11 2015, @01:58PM

          by WillAdams (1424) on Thursday June 11 2015, @01:58PM (#194950)

          I've got 400 step motors on my Shapeoko 1 --- I've found mention of 800 step motors, and have been trying to find a supplier w/ a reasonable price --- anyone got one?

          • (Score: 2) by LoRdTAW on Thursday June 11 2015, @03:32PM

            by LoRdTAW (3755) on Thursday June 11 2015, @03:32PM (#194996) Journal

            LIN engineering has 0.45 degree/step which is 800 steps/rev: http://www.linengineering.com/stepper-motors/5704.aspx [linengineering.com]. They also have 400 count motors too.

            Oriental motor also carries 500 and 1000 step/rev motors from but has a weird 5 phase bipolar setup which needs its own drive: http://www.orientalmotor.com/products/stepper-motors/stepper-motor-only.html [orientalmotor.com]. But they also carry 200 and 400 count motors as well.

            • (Score: 1) by WillAdams on Thursday June 11 2015, @04:29PM

              by WillAdams (1424) on Thursday June 11 2015, @04:29PM (#195020)

              Thanks, I think that's where I learned of them --- the problem is they seem to only sell in large quantities.

              I'd like to find a storefront which sells direct to the public in small quantities. Even better would be a U.S. manufacturer (are there any left for stepper motors?).

              • (Score: 2) by LoRdTAW on Thursday June 11 2015, @05:29PM

                by LoRdTAW (3755) on Thursday June 11 2015, @05:29PM (#195049) Journal

                Dumb question but what CNC controller are you using? Something arduino based?

                I ask because instead of hunting for motors you could simply stick with cheap 200 step/rev motors and use microstepper drives to run the motors. Then set the microstepping to whatever you need and roll from there. The only signals you need are a step clock and direction signal, two wires. Maybe your CNC controller has a step/direction mode. An arduino could do it with the right code: http://www.airspayce.com/mikem/arduino/AccelStepper/ [airspayce.com]

                Then use a driver based on the Allegro Micro A4988 which can andle 35V @ 2A. Enough to run a 61 oz-in stepper which is just above what your shapeoko calls for.

                Also, I forgot to mention Automationdirect who I use regularly for Automation parts. They have pretty cheap step motors which I have used on a number of projects here. Cheapest motor is a 61 oz-in NEMA 17 for $18. Though all of their motors are 200 step/rev. Their drives are easy to use as well though not exactly cheap. The intelligent drive is easily programmable though you need windows and a serial port to do it (no problem for a VM.) It is also quite expensive. I built a few little motion systems using those parts and 10,000 step/rev using micro stepping.

                And one last tip: Steppers work best on higher voltage buses. So use a power supply with the highest voltage you can for your driver. That allows for the current to rapidly build in the motor windings allowing for more torque at higher revs. A good drive operates using a closed current loop to run the motor. The larger step systems I have built which run at 100+IPM are ran by Aerotech CP drives which pretty much rectify the 120V coming in and feed that strait to the bridge driver.

                • (Score: 1) by WillAdams on Thursday June 11 2015, @05:51PM

                  by WillAdams (1424) on Thursday June 11 2015, @05:51PM (#195067)

                  I want actual, discrete steps, plus further microsteps --- I don't want to fake precision w/ finer microsteps.

                  • (Score: 3, Informative) by LoRdTAW on Thursday June 11 2015, @07:04PM

                    by LoRdTAW (3755) on Thursday June 11 2015, @07:04PM (#195100) Journal

                    May I ask why you want 800 full steps and microstepping? How much precision do you really need? I say this because you might want to take a step back and re-evaluate your requirements. Using hard to find 800 step motors makes things difficult in the short and long run. 200 step motors are very common and cheap.

                    Microstepping certainly isn't fake precision. As long as the controller is able to hold the current, the precision is there. As I am typing this I have a stand alone stepper controller I am building right next to me. It is uses an Aerotech Soloist MP drive which is pretty much all-in-one. I have the microstepping set to 4000 steps/rev or 20 microsteps per full step. Using the full 2.7 Amps of holding current I can twist the shaft pretty hard and the shaft will turn a bit. But once I let go, the rotor snaps right back to position as it aligns with the magnetic field.

                    Our requirements here are pretty strict so we wouldn't use "fake" precision if it could not do what we need. I have two steppers running in microstepping mode in a fixture for 4 axis electron beam welding of fuel nozzles that go into GE jet engines (sitting on a brushless driven XY table). Trust me, if we can do that and satisfy GE engineers, then microstepping will get you to where you want to be.

                    And BTW, that newmark system was designed for repeatability at 0.0001 inches or better with 20,000 steps/rev and a 10 turn/inch screw. If you need precision beyond that, you may probably need different hardware.

                    • (Score: 1) by WillAdams on Thursday June 11 2015, @07:35PM

                      by WillAdams (1424) on Thursday June 11 2015, @07:35PM (#195113)

                      Yeah, that was a bit curt.

                      Let me rephrase:

                      I'm curious about 800 step motors, what their prices are in small quantities and what advantages, if any, they afford over 400 or 200 step motors.

                      • (Score: 2) by LoRdTAW on Thursday June 11 2015, @08:31PM

                        by LoRdTAW (3755) on Thursday June 11 2015, @08:31PM (#195133) Journal

                        No worries.

                        In my opinion, going for more discreet steps has no real benefit unless you are looking for crazy high resolution per rev using microstepping. That or more steps when not using microstepping (e.g. when using a dumb step driver without current control.)

                        And crazy high resolutions are not necessary when you are driving lead screws as they have imperfections in them such as backlash and winding. Winding? Hold a straw on one end and twist the other. The straw winds up. Happens on screw stages when the loads are high and the commanded move very small. The motor moves, the encoder sees the move but the table has not moved as the shaft and/or coupler twists slightly. Even with a stepper without feedback the same thing happens.

                        Crazy high precision usually means linear or piezo stages with laser interferometer feedback. Then you can enjoy submicron precision.

  • (Score: 2, Informative) by WillAdams on Thursday June 11 2015, @01:02PM

    by WillAdams (1424) on Thursday June 11 2015, @01:02PM (#194933)

    Wait until you hit #7:

      - endmills, fixtures, jigs and other tooling

    The machine is just a downpayment on all that (thought to be fair, CNC does ameliorate the need for special shapes of endmills and jigs, and one can use a 3D printer to make some nice fixtures)