Jump to content

EdgE

Member
  • Posts

    17
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by EdgE

  1. Generating gcode for this solution is pretty simple, like you do this with an UM3.

     

    After a  nozzle switch (T0/T1), cura is executing the additional gcode automatically.

    This could be found in ultimaker2s_1st.def.json:

    "machine_extruder_start_code": { "default_value": "G0 F9000 X219 Y17.5\nG0 F1000 X223\nG0 F9000 X219" }.

    Here you could place the commands for doing the switching movements.

    See working example here: http://www.khwelling.nl/3d/pic/um2s_cura_definitions.zip

    (These definitions are partly copied from UM3 & UM2. Be aware to save the right files in the right folders.)

     

    So the only thing you have to figure out, where your exact switching positions are and write them in both extruder files.

    (So this should also be possible on an UMO+.)

  2.  

    On 3-1-2018 at 11:50 AM, jtronics said:

    Hello, that's a nice solution (like UMO3). Has anyone else tried that?
    Is it possible to generate the needed G-Code with Cura 3.X? Are there any scripts or something else?

    Thanks, well a colleague has also one printer converted to this solution. He also created the Cura 3.x definition files. :)

    His switching solution is activated by hitting the slider blocks. That will be a little bit quicker, but you loose also some mm printing space in x-direction.

    I know also, that some other people are busy with this mod, but have not seen any pictures yet..

     

    So there is a Cura 3.x machine definition available. (I will update this on my webpage.)

    The Cura modifiction is working very simple. Within the extruder definition files (ultimaker2s_1st.def.json), I just changed the line:

    "machine_extruder_start_code": { "default_value": "G0 F9000 X219 Y17.5\nG0 F1000 X223\nG0 F9000 X219" }
    So here are the coordinates stored to do the switching action for lowering the left nozzle. A similar code is added for the right nozzle.

    So the good news is, you only have the update the marlin firmware once and then you could just tune the switching coordinates in the json files.

     

     

    21 hours ago, conny_g said:

    I'd be a bit worried that cutting the bottom and removing the screws from the bottom plate makes the head much less stiff than before. Does that have any effects on the print quality?

    Good point, I did add additional mass (extra heater block etc) and the bottom plate is indeed not bolted anymore. On the other hand, there is a big spring, which presses all play out of the switching construction in one direction. Also the screws are still connected to the upper plate, compressing the plastic parts a little. From my point of view, there is not noticeable more ringing, compared to the unmodified version.  If you know a very good test stl, send a link and I love to do some tests on this issue.

    (I'm also modifying another printer, for (5x) higher printing speeds and here I do have issue's with stiffness resulting in ringing... But that will be a different topic.)

     

    This picture shows hardly any ringing, but you do see some minimal layer shifting mainly in Y direction once in a while. I guess this will be around 0.1mm.

    IMG-20170324-WA0004.jpg

  3. @Gudo, nice design, that open C-frame. For inspiration: have a look to this nice DIY printer:

    http://well-engineered.net/index.php/en/projects

    With all these mods, the only original part remaining is the frame...

    Well, that's not completely true. You do need to buy longer belts and some more pulley's for sure. You also need to print some new brackets and optional create a second top plate (wood will do). But all other parts could just be re-used in a different order..

  4. Hey Gudo, very nice build!

    I also like the way you solved the object cooling fans, I have to copy that :)

    Anyone has ever tried UM Marlin on a corexy setup? ...

    I did a year ago. Transformed UM2 having corexy inside the frame, while using standard UM2 firmware. I did have some problems with "define Quick_Home". You should not use this option in combination with corexy.

    I'm also thinkering on dockable um3 cores. Downside of using ultimainboard, is having only 2 heaters available (+build plate ofc). While using corexy, there should be many more cores inside... ;)

  5. New file location: 

    https://github.com/KoosWelling/UM2_switch

     

    I got this switching mechanism idea already a year ago, but never took time to realize. Since I'm printing now a couple of months w/o any switching issue's, it's time to share this upgrade here on the forum.

    It's working with a simple cam-plate, which is moving both nozzles at the same time. The cam-plate is bi-stable, by using the existing (UM2) springs which also hold the teflon isolator in place.

    http://www.khwelling.nl/3d/pic/um2s_part_modifications.png

    Benefits of this modification, besides enabling 2 material prints:

    • Simple modification on the existing hardware
    • Switching is done above bed clips, so no loss of additional print space
    • Modification works both on UM2 and UM2+

     

    Switching in action:

     

     

    Some prints:

    http://www.khwelling.nl/3d/pic/duplo_abc_rik.jpg

    • Like 9
  6. 1.) Can I change the firmware to push 2 separate strands of material at the same time.

    As far as I know, only one E-motor could be the active feeder.

    Although it's possible to activate both heaters.

    4.) Any suggestions questions?

    So, if you want to have 2 simultaneously running E-motors, my guess, is to re-direct the E1 step & dir signal on the big pcb, directly to E2 step & dir. (Do this on your own responsibility..)

  7. Updated design, no need for screws anymore.

    Just use a piece of 2.85 mm filament instead.

    Download the design: urobot_2hh_x1.25.stl

    Improvements:

     

    • 100% plastic: using piece of filament instead of M3 screw
    • Increased weight of feed & longer toes
    • Increased lower body weight
    • Decreased upper body weight
    • Scale: 1.25
    • Increased walking angle: around  10 degrees

     

    urobot_2hh_x1.25_assembled.thumb.png.3f63c4401dda774348e06e1be522bb0c.png

    urobot_2hh_x1.25_assembled.thumb.png.3f63c4401dda774348e06e1be522bb0c.png

    • Like 2
  8. Maybe some more tips:

    * The legs should move without any friction(!). So when holding your robot horizontal (face up), the legs should drop downwards. When rotating (face down), the legs should immediately again drop downwards.

    * So when using a screw (or something else), drill all the holes. So when I did created more walking robots (scale 1:1.5), I used for example a M3 screw:

    -drilling the body: D2.8 (okay fit for holding this M3 screw)

    -drilling the feet: D3.2 (some clearance around the M3 screw, for 'frictionless' rotation)

    * I guess some space between the legs is important, so add some washers or a (bigger) nut, or print a small cylinder, which fits in between.

    * About walking movement: of course you need a ramp. Secondly this walking needs also a sideward motion at the start. So it will move in some kind of circular motion around the feed downwards..

    * The ramp: play a little with the optimal angle. The ramp it's self should be smooth, even a sticker is high enough (~0.2mm) to stop de robot walking. (So a printed ramp is a challenge!)

    It's not that easy to make it run, but with a little modding & creativity it's definitely possible!

×
×
  • Create New...