Hi All,

So I’ve reached an Impasse.  I really want my kickstarter ready before Maker faire, but I also have a couple of machine tweaks I want to make before Maker Faire.  I don’t think that I can accomplish both.  Here’s the design changes I’d like to finish before Maker Faire:

  1. Extruder change.  I’m currently installing a new hot-end, to get  a reliable .35mm extrusion.  Right now, I have a mildly broken .50 hot end that has worn to .69.   With this change, I think I can hit Replicator or better print quality.  The only thing stopping me before was my hot end.  I’ve also designed a new cabling system for the extruder/hot end/thermsistor, that I’m testing at the same time as this change.  It allows use of CAT-6e 550MHz ethernet cable.  Why Cat-6e 550MHz cable?  It’s usually 23 AWG solid core.  Cat-5 is 24AWG, usually.  I’m hunting an easy cable that has 6 or more conductors, can be bought cheaply pre-made, has a positive-lock and keyed connector,  has some sort of screw-terminal protoboard, and is at least 23 AWG stranded wire for each conductor.  Ethernet has almost all of  these, except the stranded wire.  I’m willing to test it and see how it goes.  I f*ng hate building cables.  The BLS connectors and pin header used on RAMPS really pisses me off.  I’ve lost 3-4 days of time on those connectors, sleeves, etc…
  2. Belt path change.  This will increase the usable Y space during printing in fixed gantry mode.
  3. Adding a second Y motor.  This will allow milling of tougher materials.  I don’t think I can pull this off — I have the motor and parts, but I don’t have the time needed to re-wire for this change.  This likely won’t make it to Maker Faire.
  4. Getting some samples made.  I’ve only got today and tomorrow left, and then only the evenings after work.
  5. Electronics.  This is bothering me to no end.  I love the 5 stepper configuration of RAMPS.  I HATE the pin header connectors.  I love and hate the Allegro stepper drives.  ( Love them — good current.  Hate them — overheat. )  I want to allow dual extruders on the final version, as well as a bowden system.  That can only be done with RAMPS.  But I don’t want people building cables.  RAMPS is also very expensive!  To fix all the problems I see, I may have to design my own board around a stepper logic driver.  But I want to ship this year.  So, I’ll probably live with the cost, and design a case for RAMPS with an integrated fan, as well as fixed-position connectors on both the case and on the machine.  This way, people only plug-in the “moving” part of the cable.
As it stands, right now, I may not even have samples in time for Maker Faire.  Milled samples are the hardest.  I have milled some parts for EasyMaker, but they’re used on EasyMaker.  Hard to show off!  I want to mill something pretty, ornamental, and just plain cool.  But my GCode generator is error prone, and only works well on simple plates with holes.  I use a Mac, so I don’t have a lot of choices when it comes to GCode generators.  This really limits how cool I can make the samples.  I think I’ll design something as a derivative from a thingiverse thing.  Something that I’ve tried before, but removing the error-causing details.  Printed samples are also hard, since I haven’t even installed the new hot end or finished the new wiring!  I might be printing my samples at Maker Faire — if they let me.  I might even be milling my samples there — if they let me.  I really hope to get one or two really cool samples ready before hand.
If all the stars align and I get some divine intervention, I may be able to complete these all and make the kickstarter video, too.  And again, monkeys may fly out of my @$$.

Other than than, I’m contemplating some additional changes to the design:

  1. Moving to BallScrews.  This is tough one — I like belts, and using belts means good safety. Right now, bad GCode ( as in crashing the mill head into something or other mistake ), using a 1/8″ endmill, stalls the Proxxon ( it’s only a 40 watt tool ), and stalls the motion axis where the mistake happened.  It’s noisy and scary — but doesn’t do any real damage.  I like this — it makes the mill inherently safe.  You just can’t do something stupid enough to damage yourself or the machine.  Worst case, you damage the part you’re working on.  Great!  But, it has a downside as well.  I’d like to get to mild steel, and the current motor/belt/router configuration will probably stop at plate aluminum.  Belt is very accurate and has little lash.  I want to keep those characteristics if I move to screw.  BallScrew meets all the requirements, but is expensive as hell.  Even a cheap ballscrew means an expensive, and large, ball nut.  And talk about your mounts!  Woo!  You can literally double the price of the entire machine for a single ballscrew, nut, and mounts!  I’d have to find a great supplier, or design my own mounts to make this an affordable option.  But it would allow Jewelry-level work, so it could be worth it.  I could also use fast travel ACME. But that eats 50% of the motor power.
  2. XZ plate changes, gantry height changes.  This would allow printer mode to be more space efficient, if I add 2 more MakerSlide wheels in the middle of the plate.  It could then extend it’s reach another 1.5 inches.  I have unused space at the top of the MakerSlide Z beam in print mode. But these changes would make it no longer fit in the trunk of my car.  I think keeping it “trunk-able” is important, so I’m mulling over this change.  EasyMaker can already over-flow its frame, both in milling and in printing, so I’m not sure this is an important change.  But the software, ooh man — that’ll be a massive change to handle printing via overflow.

5 thoughts on “

  1. You want hype for the kickstarter campaign, and no better way than to give the best possible demos and show samples. So maybe focus on that before the kickstarter if you are short on time, I think. And make the campaign last as long as possible.

    For the samples, I really like things like the simple tree carving described here at the end:

    He uses pyCam so it should be good.

    You could even generate new trees live at Maker Faire if they let you mill…that would be cool, I love procedural generation.

    As for the drivers overheat, I hear the boards with integrated drivers generally do not have that problem…you might want to look into that, for example the printrboard/Teensylu, or 4pi / smoothieboard.

    For ballscrews, maybe you can make that an optional modification and show both options?

    Didn’t you have a design for the Z-axis that could be folded or easily dismounted for easier transportation? Did you quit that design?

    Good Maker Faire!

  2. I forgot about that Tree! I could even use a 1/16 end-mill on that tree, and get some nice detail. I was hoping for a sharp color contrast sample that is both milled, then printed.

    I’ve been hearing good things about pyCam. I’ll examine it later today. I couldn’t get all its dependencies installed yesterday, even with MacPorts. I’ll have to dig in some more to the issues. So far, I’ve found no GCode generator that “Just works”.

    The Z Axis is easily removed right now. I eject it every day, and re-insert it every time I use it. But carrying it around in the trunk is a different problem. I normally put it into a “tub o parts”, then put the entire tub in the trunk. That tub is a fixed size, and can’t really be changed. Larger tubs exist, but no longer fit in the trunk. Carrying it raw spills shavings everywhere.

      • Almost works for me 🙂 It requires Py25-ODE. That doesn’t compile for me — there’s a build break in the version pycam uses on my box. The newest version of PyODE requires Pyrexc, whose dependencies don’t compile. I haven’t yet figured out why. In the meantime, I’m using Skeinforge. I’ll probably just build a Ubuntu VM, and do GCode generation in there. I might even use the VM to run some of the tools. Being a Mac user, you get used to VMs, bootcamp, etc… It’s just a matter of time — I have lots of calibration to do with the new hot-end, and I’ll be doing that during the setup day at Maker Faire.

      • Try the homebrew tutorial instead of macports (the second link from shapeoko wiki).

        Many people have replaced macports with homebrew and are happy with it.
        I also have a mac and have used homebrew before with no problems.
        I avoid the VM but sometimes it can be useful..

Comments are closed.