MPCNC Ukraine

This topic contains 55 replies, has 8 voices, and was last updated by  Guffy 2 months, 2 weeks ago.

Viewing 30 posts - 1 through 30 (of 56 total)
  • Author
    Posts
  • #69355

    Guffy
    Participant

    hi all
    I know that probaly a few people from my country made a MPCNC, but it seems I’m first who reported about that here. )

    So. The project is not finished yet, but at least it began moving.
    It’s mpcnc with working volume 600x400x120mm and dual endstops. I have used PLA plastic and 25mm stainless stell tubes. It will use 710W 30000rpm Metabo milling motor.
    Expected usage is milling wood and aluminium.
    And may be I will buy 10-15W led laser head.

    20180916_130625
    20180916_130613

    Also I have tried Repetier Firmaware and attach config.json file. Just a few notes about this:
    – I don’t connected endstops yet, but it can move XY axis to both directions.
    – Somewhy it moves Y axis to opposite side. So you have to swap Y and E1 motors.
    – I’m not familiar with Z-probe yet so may be I didn’t configure it properly
    – Provbably because lack of endstops it dosen’t not play simpliest g-code program properly.

    20180916_130524

    So I will start with Marlin, may be later will switch to Repetier

    3 users thanked author for this post.
    #69373

    Jeffeb3
    Participant

    Interesting. I don’t know why, but I never thought I’d using repetier firmware. I didn’t know they could motor x and y drivers.

    #69377

    Barry
    Participant

    I ran into the same motor issue with repietier firmware.  It seemed like an easier way to configure things with its web based setup, but doesn’t quite work for us.

    #69396

    Guffy
    Participant

    I have solved Y axis direction (Checked invert Y motor and unchecked invert Y2\E1 motor). So manually it moves properly (configuration.h attached).
    But it still doesn’t properly play gcode

    G21 ; Set units to mm
    G90 ; Absolute positioning
    G1 Z2.54 F4800 ; Move to clearance level
    
    ;
    ; Operation: 0
    ; Name: 
    ; Type: Engrave
    ; Paths: 1
    ; Direction: Conventional
    ; Cut Depth: 3.175
    ; Pass Depth: 3.175
    ; Plunge rate: 127
    ; Cut rate: 900
    ;
    
    ; Path 0
    ; Rapid to initial position
    G1 X283.6147 Y148.6772 F4800
    G1 Z0.0000
    ; plunge
    G1 Z-3.1750 F127
    ; cut
    G1 X151.3228 Y148.6772 F900
    G1 X151.3228 Y16.3853
    G1 X283.6147 Y16.3853
    G1 X283.6147 Y148.6772
    G1 X283.6147 Y148.6772
    ; Retract
    G1 Z2.5400 F4800
    
    ; Return to 0,0
    G0 X0 Y0 F4800
    M2
    Attachments:
    1. Configuration.h
    #69430

    Guffy
    Participant

    And I found that repetier doesn’t properly play the gcode because it has move commads over X\Y dimentions.
    So i defined real dimensions 600x400x120 and repetier played the gcode well. It seems Marling just doesn’t limit gcode to move over machine MAX coordinates. Repetier does.

    Attachments:
    1. Configuration-1.h
    #69439

    Jeffeb3
    Participant

    Marlin will limit the size if you enable soft endstops.

    The feedrate on your rapids is F4800, which is 80mm/s. That’s probably close to too fast for X/Y, but definitely for Z. Unless repetier is somehow faster. Since Z needs to move 800 steps/mm, the stepper goes a lot faster to get the same speeds. Something like F300 is a lot safer for Z moves.

    #69440

    Guffy
    Participant

    yep, i know about software endstops in Marlin. Repetier also has them and it for MAX it enabled by default.

    #define min_software_endstop_x false
    #define min_software_endstop_y false
    #define min_software_endstop_z false
    #define max_software_endstop_x true
    #define max_software_endstop_y true
    #define max_software_endstop_z true

    But i didnt’s find it in UI configurator. So you can disabe them, but have to manually edit configuration.h

    This how Repetier plays the code now:

    That’s Marlin:

    #71176

    Guffy
    Participant

    hi again
    So, there is a progress in my build

    I made a table. Made boxes for LCD and electronics. Wired the mpcnc with 6 x 0.22mm^2 cable and crimped connectors. Made drag chan brackets for Z. And all these parts combined together.

    And finally drawed the picture with gсode produced by Estlcam.

    So next step I’m going to do is making tool mounting parts

    20181006_185251
    20181006_185315
    20181006_185342
    20181006_185403

    1 user thanked author for this post.
    #71191

    Ryan
    Keymaster

    Looking good so far!

    #71232

    Guffy
    Participant

    Thank you 🙂
    Two things are a bit difficult when I was building the mpcnc:
    1. Have to search a lot of addition models. As example, your STLs for dual endstops can be included in main package, but they don’t. But this is simpliest case. Some things I had to design by myself, becadse nothing I had found liked me or didn’t fit tomy design. So I designed drag chai mounts with openscad and brackets with fusion 360.
    2. Honestly, it’s a bit annoying that holes aren’t desinged for M4 nuts. Some time it’s enough space for key, some time i had to use screwdriver as wedge. For tool mount I will print nut holders.
    Imho, if you can assume that F version of mpcnc could be named as international it have to be designed for M4 nuts

    #71257

    Ryan
    Keymaster

    Thanks. I will keep that on the notepad for the next revision.

    #71258

    Ryan
    Keymaster

    What are the flat to flat dimensions of a regular M4 nut in your country?

    #71307

    Evan
    Participant

    Ryan, you need a copy of Machinery’s Handbook!

     

    http://www.engineershandbook.com/Tables/nuts.htm

    #71343

    Guffy
    Participant

    What are the flat to flat dimensions of a regular M4 nut in your country?

    143109695_w640_h2048_m

    #71350

    Guffy
    Participant

    so for nuts I have printed this case https://www.thingiverse.com/thing:3037646
    but print it no so easy.
    * I have to replace nozzle to 0.4mm (I printed the mpcnc with 0.5mm)
    * use Slic3r (Cura couldn’t make the proper result).
    * enlarge the model by 4-5% to fit a nut.

    Models I had used when was building:

    Nut trap for M4 https://www.thingiverse.com/thing:2372852

    Parametric drag chain model to print 2 ends (male + female) https://www.thingiverse.com/thing:915487
    I done small modification of openscad code.

    Mega + Ramps case:
    This model won https://www.thingiverse.com/thing:761806
    + minimal size (space where the drag chain is also located)
    + have places for a fan (only not good idea is to make external rims around holes for fan screws)
    + cover could be opened any time without screwdriver
    – boards are fixed not too good. i had to use a few pieces of gum to fix them better

    Smart controller case. I made it from 2 model sets.
    From this one I took top and bottom part of body https://www.thingiverse.com/thing:563515 . But middle part is wrong. Had to desing new one.
    From another one I took legs – with 45 degree angle and holes for screw it on the table.

    Drag chain mount i designed by myself https://www.thingiverse.com/thing:3086172
    You can take it.

    Brackets I also designed by myself (in Fusion 360). Similar to this one https://www.thingiverse.com/thing:1855944

    Holder of a cable for Z https://www.thingiverse.com/thing:1877548
    I used it without modifications, but it’s not best model. It’s easy to broke central pin.

    #71352

    Guffy
    Participant

    for tool holder i used this model https://www.thingiverse.com/thing:2853994
    this only one which has 16.5mm space between 43mm hole and body of tool mount. this is important for my milling motor.
    i think this could be used by anyone who used motors from Bosh (like GGS 28 C or GGS 28 CE) or Metabo (like GE 710 Compact or FME 737)

    #71526

    Guffy
    Participant

    ok, I have mounted the tool. I will move the cnc to garage and try to milling

    20181010_165247
    20181010_165306
    also I made a few changes to marlin configuration
    1. a small fix – I had set proper real sizes of bed. otherwise homing worked wrong when i tried to do it from far corner

    // The size of the print bed
    #define X_BED_SIZE 595
    #define Y_BED_SIZE 395
    
    // Travel limits (mm) after homing, corresponding to endstop positions.
    #define X_MIN_POS 0
    #define Y_MIN_POS 0
    #define Z_MIN_POS 0
    #define X_MAX_POS X_BED_SIZE
    #define Y_MAX_POS Y_BED_SIZE
    #define Z_MAX_POS 120

    2. I try to prevent disactivate Z stepper by timeout. The tool is pretty heavy (1.4Kg)? so with disabled stepper tool mount may slide down.

    // Default stepper release if idle. Set to 0 to deactivate.
    // Steppers will shut down DEFAULT_STEPPER_DEACTIVE_TIME seconds after the last move when DISABLE_INACTIVE_? is true.
    // Time can be set by M18 and M84.
    #define DEFAULT_STEPPER_DEACTIVE_TIME 600
    #define DISABLE_INACTIVE_X true
    #define DISABLE_INACTIVE_Y true
    #define DISABLE_INACTIVE_Z false // set to false if the nozzle will fall down on your printed part when print has finished.
    #define DISABLE_INACTIVE_E true

    3. Most useful thing – I filled custom menu with items suitable for CNC

    /**
    * User-defined menu items that execute custom GCode
    */
    #define CUSTOM_USER_MENUS
    #if ENABLED(CUSTOM_USER_MENUS)
    #define USER_SCRIPT_DONE "M117 User Script Done"
    #define USER_SCRIPT_AUDIBLE_FEEDBACK
    #define USER_SCRIPT_RETURN // Return to status screen after a script
    
    #define USER_DESC_1 "Set zero X Y Z"
    #define USER_GCODE_1 "G92 X0 Y0 Z0"
    
    #define USER_DESC_2 "Home XY"
    #define USER_GCODE_2 "G28 X Y"
    
    #define USER_DESC_3 "Move by X+100 Y+100"
    #define USER_GCODE_3 "G91\nG0 X100 Y100 F1200\nG90"
    
    #define USER_DESC_4 "Move by Z+5"
    #define USER_GCODE_4 "G91\nG0 Z5 F600\nG90"
    
    #define USER_DESC_5 "Set zero Z"
    #define USER_GCODE_5 "G92 Z0"
    #endif
    
    20181010_165323
    
    
    1 user thanked author for this post.
    #71531

    Jeffeb3
    Participant

    2. I try to prevent disactivate Z stepper by timeout. The tool is pretty heavy (1.4Kg)? so with disabled stepper tool mount may slide down.

    // Default stepper release if idle. Set to 0 to deactivate.
    // Steppers will shut down DEFAULT_STEPPER_DEACTIVE_TIME seconds after the last move when DISABLE_INACTIVE_? is true.
    // Time can be set by M18 and M84.
    #define DEFAULT_STEPPER_DEACTIVE_TIME 600
    #define DISABLE_INACTIVE_X true
    #define DISABLE_INACTIVE_Y true
    #define DISABLE_INACTIVE_Z false // set to false if the nozzle will fall down on your printed part when print has finished.
    #define DISABLE_INACTIVE_E true

    That is awesome. We need this in the Low Rider configuration. I think Ryan just sets the deactive time to something very large.

    Also excellent Custom LCD options. I didn’t realize you could enter those so easily. Thanks for sharing!

    #71775

    Bill
    Participant

    You might want to put a second strap around your tool up higher to provide a bit more support when the bit is cutting. If not you risk the existing mount cracking under load.

    #71814

    Guffy
    Participant

    Firstly, I guess that in case I will use too high feed rate a flute with diameter 1/8 or less will breaked firstly then the motor adapter. may be this is wrong expectations, i haven’t enough expirience yet.
    Secondary, if I will decide to use such top ring I will have to make more sophisticated design of it than people design it usualy. because on the level where tool mount has last holes the motor has power switch. Or i have to replace tool mount to some more tall variant.
    Right now I trying to choice dust shoe from variants I see on the thingiverse

    #71822

    Jeffeb3
    Participant

    Firstly, I guess that in case I will use too high feed rate a flute with diameter 1/8 or less will breaked firstly then the motor adapter. may be this is wrong expectations, i haven’t enough expirience yet.

    The steppers will skip before anything breaks.

    #71844

    Guffy
    Participant

    The steppers will skip before anything breaks.

    Never say never 😉
    First try of milling of aluminium. Carbide flute 1/8->2mm12mm.
    feed rate 15mm/s, 0.5mm Z step, climb milling, 5% trachoidal
    Broken (may be due cutting path had been out over edge of the plate)

    20181013_194840

    By the way I tried 2 days ago to try foam and wood (flute 1/8, 20mm/s, 1mm Z step, conventional). Milling succeeded.

    20181010_191949
    20181011_104522

     

    #71870

    Jeffeb3
    Participant

    The aluminum was probably because the but got stuck to the work, right? The chips fused them together?

    But sure, anything can happen. Except i will never say never 🙂

    #71872

    Guffy
    Participant

    The aluminum was probably because the but got stuck to the work, right? The chips fused them together?

    All happened so fastly so all that I saw is that the tool broke when it had leaved the aliminium. Entire milling took just about 5 sec
    And I found that it seems estlcam doesn’t support trochoidal\adaptive milling for 3D mode (I took STL file model)

    #72079

    Guffy
    Participant

    so, it’s like a blog here )
    I tried to milling a piece of laminate and an aluminiun strip (soft aluminium 2 x 50 x 1000mm. I had bought it specially for tests).
    I prepared 2 gcode files. Both for 1/8 tool, milling depth 1,5mm, 24000RMP (in Estlcam)

    Firstly i was set about 20000RPM on the motor and tried to cut a piece of laminate.

    1. 5% trachoidal cut, 4800mm/min XY feed rate, 1200mm/min Z feed rate

    20181015_143820

    2. ordinary cut, 450mm/min XY feed rate, 300mm/min Z feed rate

    20181015_143823

    Imho, both results are excellent.

     

    Then I tried to milling aluminium strip.
    A. Not finished “W”. Trachoidal milling at full depth.
    Failed. Sometime it had cutted holes in the bottom layer. And it looks like steppers lost steps.

    20181015_143445

    B. Not finished “Woo”. I increased speed to 25000RPM. Trachoidal milling at partial depth. I defined Z zero at 0.6mm on top of the stripe. So it cutted not with full depth, about 1mm.
    Semi-succeeded. Bottom surface looks not bad, but edges looks dirty.
    And this trochoidal milling is really slow. This work took abour 23 min.

    20181015_143452

    C. Finished “Woot!”. I increased speed to maximum that the motor can do – 27000RPM. Ordinary milling at partial depth. I defined Z zero at 0.7mm on top of the stripe. So cut depth was about 0.9mm
    Edges look much better then with trochoidal milling. Bottom layer looks not so good, but at right side it more dirty I guess just because the stripe was not enough fixed. The job completed really fast – just in 4 min.

    20181015_143504

    Conclusion at the moment. I think that trochoidal milling probaly could be reasonable choice for harder materials like steel or some kind of aluminium alloy. But for soft aluminium ordinary milling with my motor produce not bad results and much faster.

    20181015_143807

    #72103

    Ryan
    Keymaster

    You need to mess with RPM, I am thinking it is way too high. That ridge of stuck chips on the surface is all bad. you need to cut a big enough chip that it gets thrown clear and takes the heat with it. The chips you have are welded on to the material.

    1 user thanked author for this post.
    #72127

    Guffy
    Participant

    I thought that spiral movements during trochoidal milling maked some wrong cutting direction between tool and material, but then i looked a video and ensured that all ok.
    And I had increased RPM because on that first try I thought that cutting power is not enough, so steppers was stuck and lost steps.

    so, you mean rpm is too high, the tool friction too much to the material and this makes cutted surface too hot and aluminium just melt under the tool.
    ok, thanks.

    I will prepare a gcode files again. With single “W” (a bit smaller, with 20mm height) and changed settings.

    1. For trochoidal milling I will increase feedrate to XY=6000mm/min and Z=1500mm/min. Will try to slow down from 25000 to 20000 and then to 15000. Will mill with DOC=1mm.
    Also I will prepare a file with increased trochoidal=10%

    2. For ordinary milling I will keep previous feedrate (XY=450, Z=300), but cutting calculator hints me that 13000RMP is enough for these feedrate and DOC=1/WOC=3.175. So I will try to cut at 15000 and 13000.

    Hope, I will not break the tool )

    #72518

    Guffy
    Participant

    hi again
    I made geometry test and a few tries of aluminium milling
    20181017_181847

    #72524

    Guffy
    Participant

    20181017_192431
    All tries made with DOC=1mm, feedrate 450mm/min. Ordinary milling with WOC=90% (of 1/8)

    Ordinary milling looks not bad with 23000 and 21000 rpm.
    13000 (as calculator recommended) look bad. 17000 also not good.
    at 20000 i see some vibration on the left side edge of W. i don’t know why.

    Trochoidal milling at 21000 rpm (with 10% and 5%) looks fine from edges and bottom surface quality point of view. But at feedrate 6000mm/min there is huge geometry distorsions. looks like this feedrate is too much, belts work as springs. as result W looks too fat with too rounded corners. 1500mm/min of Z feedrate also produces a bit too deeper cut.
    when I look at my previous try I see that trochoidal milling at 4200mm/min also has similar distorsions. so i don’t know, may be something about 1200 will be good for such heavy head, need to do additional experiments.

     

    #72534

    Guffy
    Participant
Viewing 30 posts - 1 through 30 (of 56 total)

You must be logged in to reply to this topic.