Predator Project - The Mjolnir -



  • Hi guys, I would like to show you what I'm working on.
    I'm trying to collect here, in one place the troubles I have and what I've found on this forum.

    some info first:

    I start working on my Anycubic Predator in march, here what will be the final result: Anycubic_Predator_MM.png

    The upgrade consist in:

    -Duet3 Board + Panel I7* + Raspberry Pi4
    -Double extruder
    -Maxiwatt hotend heater *
    -BLtouch V3 *
    -Magnetic joint ยน
    -linear bearings (rails)

    -nema17 (0.9ยฐ)*
    -Liquid cooling for extruders and (if possible) for motors
    -... and other small things ...

    Anycubic_Predator_MM_est.png
    Anycubic_Predator_MM_est_sez.png
    Anycubic_Predator_MM_est2.png
    Anycubic_Predator_MM_rail.png

    *this is already done...
    ยน I placed again the standard rods because as you will read, I've some troubles leveling the base

    Now consider I'm working with the standard Anycubic extruder configuration: 1 extruder with flying motor (only maxiwatt heater and BL touch V3 upgrades are applied)

    Major issues I've now is regarding the leveling, tomorrow I'll try to explain as well as possible what I do, trying to understand with you what is wrong. I've no doubt on rods lenght, but is not clear for me the endstop regulation

    I anticipate you that I've read the following post found in this forum:

    https://forum.duet3d.com/topic/13097/endstop-calibration-for-delta?_=1597433494622
    https://forum.duet3d.com/topic/9741/wavy-ridges-in-mesh-bed-leveling-on-delta-kossel-xl/2

    and this nice document:
    http://boim.com/DeltaUtil/CalDoc/Calibration.html

    but all my trials following the suggestions wrote, goes bad...

    the best result I get with this config:
    config.g
    homedelta.g
    bed.g

    is this Heightmap:
    heightmap.csv

    98d6eba2-6a92-4852-8e58-5c62f066ad13-image.png

    4f2b18d4-3c6c-4b5d-9a71-ecd9d5920c4a-image.png
    and this is the beginning of print:

    0b54a1b6-46e2-4c6a-8cbd-424ccf0d846d.jpg

    in the red area (positive Y) the extruder is very closer to the base, (on big Y values it touch the bed ๐Ÿ˜ฐ ).
    green area is the reverse, little bit far from the bed (doesn't stich well to the surface) (I will take a better picture with white PLA, is more clear)

    I don't understand yet where to work first... firmware or hardware...

    Is the first time I made something at this level, probably i'm missing something stupid (I'm sure ๐Ÿ˜ณ )

    Any kind of suggestion are well accepted.

    thank you guys,

    Marco


  • administrators

    I presume you have read https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer.

    Offset Z probes such as BLTouch do not work well on delta printers. They are sensitive to effector tilt (which alters the relative height of the Z probe and the nozzle, hence the effective trigger height), and it's very difficult to eliminate effector tilt completely. To make them work at all on a delta, you need to measure the trigger height (nozzle height above bed) at every auto calibration point using feeler gauges, and apply the differences between those measurements and the trigger height at the reference position (normally bed centre) as H parameter corrections on the G30 commands in bed.g. This is tedious, and needs to be re-done any time you make a change that might alter the calibration.

    Nozzle-contact Z probes are much more satisfactory on deltas.



  • @dc42
    "I presume you have read https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer." Yes sorry I forgot... it was the first one when I start installing duet board on my predator ๐Ÿ˜Š

    Honestly I thought the BL touch was not the best, also because as you can see, it has a great offset from the effector center axis (46mm), I thing by the geometry this can be one of the reasons it is not a good solution.
    4555e6b4-ea8c-4dd1-8537-2543186812f7.jpg

    Nozzle contact probe (I think you're speaking about piezo probe) was one of the first solution i thought but i was scared about the mechanical application on my effector, especially looking forward for the double extruder mod...

    Here the promised update i've done today:

    As you can see here, the printing is quite well in the center of the bed, become far in negative Y cordinates and too closer on positive Y cordinate...
    64b63d62-9f35-47b8-8811-023c1ec1a50a.jpg

    Here the actual effector configuration:

    6b1e7223-98cc-47be-92d1-6959f126a426.jpg

    3f4d5341-285c-4471-ac04-8d5f2f3b91c0.jpg

    e48d3f18-6fca-4a41-ba3f-29af8084813f.jpg

    ... Next step: design new effector for Piezo probe.

    You think guys it'll need other mod, due to this new design?



  • I forgot the other option, Piezo on bed. Here I think we have 2 roads:

    Piezo under the bed (but i'm scared when i need to probe for ABS (bed @90/100ยฐC, i think piezo suffer at that temperature.

    Piezo under the bed support... but how many and where on ultrabase shape?

    suggestions?

    thanks again guys


  • administrators

    Piezo in the print head mounting or on the effector may be another possibility.



  • Smart effector is also an elegant solution, but with the dual extruder you pretty much have to eliminate effector tilt and in that case bltouch should also work reasonably well?



  • @bearer said in Predator Project - The Mjolnir -:

    but with the dual extruder you pretty much have to eliminate effector tilt

    My experience with dual nozzles is that it's very difficult to prevent one from interfering with the printing by the other.

    I wonder if the 'tiltability' of delta printers can be used to address it. E.g. on tool change, tilting the non active nozzle up such that it is not interfering with piezo measurement or printing by the other nozzle.



  • Hi Everybody, thank you for the suggestions and sharing your experiences.

    what I learned is to say bye bye to my BLtouch (i will recycle it on my ender 3 ๐Ÿ˜ฌ )

    I designed new bed support to place piezo (Precision piezo kit ร˜20mm):

    6f650e35-67b2-4c82-80c9-94556d5720e9-image.png

    46816725-468b-4f0b-abe5-0a118f4b0396-image.png

    30563185-384d-4648-9697-80ea3f40584b-image.png

    1a1b5a31-3cbf-45e2-9082-5d6c84082f99-image.png

    ...and problably i will move to a new print head solution to prevent tilt issues:

    9dab07eb-78f7-4592-b34d-825a55ea8505-image.png

    i will design a special cooler to proceed with my liquid cooling idea and create a compact effector...

    hope you like it and i'm on the right way now.. ๐Ÿ˜‰ thanks guys!



  • I am waiting for the day when our 3D printers will be as pretty as their CAD renderings.



  • here we go....!

    20f83983-cf11-4ffc-be58-83d7e1ed6310.jpg

    3983686e-fbce-4ba2-88de-2043db8507d0.jpg

    instead of what drawed before, I made (due to supplier wrong shipment) the ร˜27 piezo support, probably more sensitive ๐Ÿ˜‰

    here a short preview,
    https://youtu.be/zSb38aC-XgE

    looks good:

    g32.jpg

    But unfortunatelly I don't understand this mesh... really don't...
    dd5e4b36-c529-42f2-87bd-a7b9c8042c2b-image.png

    I will check again geometry and road lenght with a caliber...

    wait for your comments


  • administrators

    @thor3k, that height map with ridges at one end of the X direction turning into valleys at the other end is a beautiful example of the effects of backlash in the joints of a delta printer.

    Try lubricating the joints.



  • Thank you dc42, I thought it depend of bad geometries.. is something I saw on first tests, I remember (se my first post) it goes better when I goes back from magnetic joints to the original ball eye rod joints..

    I will check the lubrification on my mag balls...


  • administrators

    Btw sticktion in the carriage or belts too loose would also cause that pattern in the height map.


Log in to reply