Thanks Thanks:  4
Likes Likes:  4
Dislikes Dislikes:  0
Page 12 of 13 FirstFirst ... 210111213 LastLast
Results 111 to 120 of 123
  1. #111
    3D Printer Noob
    Join Date
    Mar 2018
    Posts
    20
    Post Thanks / Like
    Quote Originally Posted by JohnTee View Post
    boilerbots!!
    If you are still here, please elaborate a little.

    What is a 'smoothie board', what does it do, where does it do it, what does it do it to??

    Is this a Cube Pro thingy, or do you have a different printer now?

    Curious minds want to learn!
    As bolsoncerrado pointed you to, http://smoothieware.org/. It is a project that I believe started from a kick-starter project and various other opensource printer software bases. There are many boards that companies use to run their 3D printers and then there are lots of software projects that target various boards. The main difference between the hardware and the software is the embedded processor that they are designed around. I am converted 32bit ARM guy and so I wanted a board that is using an ARM processor, Smoothie is the project for that. Another famous firmware/software project is Marlin and that is based on very old 8 bit Atmel microcontrollers. The Marlin project is used by Prusa and you can buy the Rambo boards that Prusa used because they are designed and manufactured by a different person (https://ultimachine.com/products/einsy-rambo-1-1).

    You can buy a Smoothie board designed and directly supported by software project through their website or you can buy other boards designed and built by various people that also work with the Smoothie code because they use the same processor and mostly the same components. For me the "MKS SBASE v1.3" board seemed like a good choice, it has drivers for 5 motors, including 2 extruders, 2 hot ends a bed heater and a fan. It also comes with ethernet and can run everything off a single 24V supply which is what the Cube uses internally.

    You can buy the board through Amazon for $60 delivered!

    The complication is with the retrofit of all the cables. The original Cube board used a different series of JST connectors, wrong pins, wrong pitch. So I had to cut off all the board side connectors and crimp on new pins (these are very tiny pins). I had to remove the "expansion" board from up where the 3 extruders are located and rewire that area. I got the steps per millimeter from the Cube XML files so that part was easy. I had to calibrate the acceleration parameters for the motors and play around with homing and limits. Also I had to determine the calibration points for the hotend thermistor because it is not documented anywhere.

    If you bought the board, a dc/dc, relay board, connector kit I think you could convert a Cube Pro like I did for around $100.

    So now that I am free from the shackles of 3D systems I can just get on with printing using whatever materials and software I want.

    When I purchased the printer used (for not a lot of money) I new that it would require hacking to make it useful and based on the threads in this forum I was betting that I could help unlock this thing but in the end the hours (many hours) where just adding up with no end in sight, so I decided to go the route that I new would ultimately work, gut it and replace the controller.

    If there where a lot of Cube Pro owners interested, it would be possible to make a "smoothie" compatible board that was designed to replace the original board and be pin compatible with the existing cables, minus the LCD touch screen perhaps. However the production costs in the US would probably cost more than $150 to produce the boards. The same could be done for the Cube series of printers.

    Sorry to derail your BFB thread.....

  2. #112
    3D Printer God(dess)
    Join Date
    Nov 2016
    Posts
    1,579
    Post Thanks / Like
    <think adapter cables>

  3. #113
    3D Printer Noob
    Join Date
    May 2018
    Posts
    6
    Post Thanks / Like
    if someone can help me I have a problem I try to print with two materials at the same time with infinity rinse as supports and Pla, the problem is that the change of color between layers is eternal and in doing this each layer of melts the other and remains horrible, some way to change the color change times

  4. #114
    Regular 3D Printer
    Join Date
    Feb 2018
    Posts
    101
    Post Thanks / Like
    Thank you for the great explanation, Boilerbots!

    I totally get your frustration with trying to decode the black box that is the 3DS controller.

    What you've accomplished and the knowhow that came with it is invaluable.

    Would you be willing to share details about the parameters you've worked out for the smoothieware and the MKS board? I know it would take some more of your time, but I suspect there are a few more folks that would like to follow you down that path! Perhaps offering it as donationware might help defray some of your cost (as eychei/tomuro did with ReCube.) I doubt there are enough to justify making a plug-compatible version of the board, though.

    (Oh, and if you are willing to continue the discussion, this would make a terrific new Thread!!)
    Last edited by JohnTee; 06-06-2018 at 01:58 AM.

  5. #115
    3D Printer God(dess)
    Join Date
    Nov 2016
    Posts
    1,579
    Post Thanks / Like
    This may be a good place for this...
    Very little has changed over time in the .cube3 initialization procedure.

    The highlighted cells represent additions and subtractions from V1.00 to the current version 1.14B.
    Of note is the removal of a G92 command and an apparent need for a second M107 (fan off) command.
    The G92 command assigns a new home position. In fact, it removes some of the offset by going negative in Z.
    This means that the the system should be using the following as an initial gap:
    the offset from the gauge... of much controversy but lets say it is 0.25 for number purposes. This is in essence "zero" in Z.
    The first layer of V1.14B files starts at Z0.25 without a G92 offset.
    The test print V1.00 starts at 0.38mm and has two different offsets, -0.10 for ABS and -0.08 for PLA.
    This means the test print Z-values are offset by the G92 code.
    So if the gap is set to 0.25 and you add the 0.25 for the first z-move, you get an effective gap of 0.50mm using V1.14B.
    However, if you have the same 0.25 gap and you run the test prints, you get an effective initial gap of:
    0.25+0.38-(0.10 for ABS or 0.08 for PLA) ending up with an initial gap of 0.53mm with ABS and 0.55mm with PLA with V1.00.
    30-50 microns in variation... That's a difference of as much as 10%!

    I believe that these facts makes the test print pretty much useless to confirm gap

    Cube_init_profiles.PNG


    Otherwise, sidewalks came into being as well.

    Now I really want to see what comes out of 3D Builder!

    - - - - - - - - - -

    What do you suppose 3DS used for M240?

    If should mean "take a picture" according to G-code... (?)

    Notice that the values swap between the two sets of calls.

    - - - - - - - - - -

    I recommend this posting for understanding a few things about the Cube3 slicer:
    http://www.print3dforum.com/showthre...ll=1#post44861

    Toward the bottom I discuss the voids 3rd party ABS leaves. This is something that could be resolved with a bfb change.
    Anyone know which encoder works generically for the Cube3 decoded files?
    I'd really like to do some manual file tests.

    - - - - - - - - - -

    Thanks JohnTee for clarifying encoding. Definitely hot to figure out how to make this MakeShaper filament work.

    - - - - - - - - - -

    I'm pretty sure this will be welcome here... Old info but quite useful. Source: http://www.kisslicertalk.com/viewtop...&t=97&start=10

    Its used in the select extruder section to set the M240 (acceleration?) and extruder speed for the material in use.


    Printer gcode:
    Prefix:
    CODE: SELECT ALL
    ^Firmware:V1.00
    ^Minfirmware:V1.00
    ^DRM:000000000000
    ^PrinterModel:CUBEPRO
    ; [mm] mode
    ;G21
    ; absolute mode
    ;G90
    M404 S<BED-20>
    ^ set chamber temp to bed temp take 20 deg
    M227 P250 S250 G300 F700
    ^ retract and reextend, global.
    M228 P0 S250
    ^ disable m227
    M231 P0 S0
    M232 P2500 S2500
    M233 P1850
    ^home z, drop z to glue posn, display glue message and wait for ok?
    M106 P100
    ^ fans on, max speed
    G4 P0
    ^ dwell disabled/dwell for 0 seconds
    ^M601 P2 S30 F5
    ^ fairly sure this is select tool offset 1 - the following commands use this offset when interpreting g commands. P0 is extruder 1, P2 is extruder 2. s30 and F5 don't change between tools
    M228 P0 S1
    ^ disable m227
    M227 P1 S1 G1000 F1000
    ^ change global retract settings again
    M404 S<BED> P1
    ^ set chamber temp to bed temp and wait till it gets there
    M107
    ^fan off
    ^InitComplete




    Select extruder:
    CODE: SELECT ALL
    ; Select extruder, warm, purge
    ; BfB-style
    M106 P100
    ^fans on
    M<EXT+1>04 S<TEMP-10> P1
    ^heat new ext to priming temp
    <MATL>
    ^select acceleration settings for current matrerial
    G4 P0
    ^dwell for 0 seconds
    M<EXT+1>04 S<TEMP-10> P1
    ^heat new ext to priming temp and wait
    G1 X108.000 Y136.000 <Z> F9000.0
    ^posn for wipe
    G1 X108.000 Y161.000 <Z> F9000.0
    ^wipe new ext into bin
    G1 X108.000 Y157.000 <Z> F9000.0
    ^posn new ext near bin lip
    M<EXT+1>04 S<TEMP-10>
    ^heat new ext to priming temp
    M55<EXT+1> P650 S50
    ^prime new ext 650 steps at 55rpm
    G1 X108.000 Y136.000 <Z> F9000.0
    ^wipe new ext out of bin




    Deselect extruder:
    CODE: SELECT ALL
    ; Guaranteed same extruder, cooling down
    ; BfB-style
    <MATL>
    ^select acceleration settings for current matrerial
    M<EXT+1>04 S<TEMP> P1
    ^keepwarm old extruder
    M56<EXT+1> P450 S500
    ^reverse old extruder 450 steps at 50 rpm
    G1 X118.000 Y133.000 <Z> F15000.0
    ^posn old ext for wipe
    G1 X118.000 Y157.000 <Z> F9000.0
    ^wipe old ext into bin
    G1 X118.000 Y133.000 <Z> F9000.0
    ^wipe old ext out of bin

    - - - - - - - - - -

    Take a very careful look at this information from the BFBlib...
    This seems to be the generator of the .cube3 files. it clearly shows what is a variable and what is fixed.


    M106
    M107
    M104
    M204
    M304
    ^Time:
    ^Time:%d
    ^Material:
    ^Material: %.3f
    ^LayerTime: %s %f %f %f
    M%d04 S%d P1

    M103
    G1 X%.3f Y%.3f Z%.4f F5000.0
    G4
    P1
    ^LayerTime: %.4f %.3f %.3f ^

    M601
    M101
    M201
    M301
    M227 %s %s
    ^ModelHeight:
    ^ModelHeight: %.3f
    ^LayerCount:
    ^LayerCount: %d
    ^MaterialLengthE1:
    ^MaterialLengthE1: %.3f
    ^MaterialLengthE2:
    ^MaterialLengthE2: %.3f
    ^MaterialLengthE3:
    ^MaterialLengthE3: %.3f
    M106 P%d
    G4 P%d
    ^InitComplete </Summary>
    #QuickLayer P%0.2f O%0.2f D1
    #QuickLayer P%0.2f O%0.2f D0
    M228 P%d S%d
    M227 P%d S%d G%d F%d
    #Vector T
    M108
    M108 S%.1f
    G1 %s %s %s %s %s

    G1 X%.3f %s %s F%.1f
    M228
    M103
    M240 S0
    ^INITCOMPLETE
    #VECTOR T
    G4
    G1
    ^FIRMWARE: V
    ^DRM: :
    ^CUBE:
    ^TYPE:
    ^TIME:
    ^LayerDensity: ^Supports:
    ^Raft:
    ^MaterialCodeE1
    ^MaterialCodeE2
    ^MATERIAL:
    ^LayerHeight: P S
    ^LAYERTIME: %f %f %f
    ^Time
    ^Type
    ^Cube :V
    ^Firmware
    ^Minfirmware
    ^DRM
    ^Material

    ^LayerTime P F G S
    M%d04 P1
    M56%d
    M55%d
    M542
    M543 Z Y X 0 M
    M108 S
    #Vector
    ^MaterialCodeE3
    ^MaterialLengthE1
    ^MaterialLengthE2
    ^MaterialLengthE3

    ^ModelHeight
    ^LayerCount
    ^LayerHeight Hollow Strong Solid
    ^Density Hex Lines Diamond
    ^Pattern
    CUBE1
    CUBE2
    CUBE3
    CUBEX
    CUBEPRO EKOCYCLE

    ^PrinterModel
    ^Supports
    ^SupportBorders: TRUE
    ^SupportBorders: FALSE
    ^Sidewalks
    ^SidewalkBorders: TRUE
    ^SidewalkBorders: FALSE off: on:
    ^SidewalkPerf: ^Version
    ^Raft
    M231
    M232
    M233
    M404
    M240
    ^XSize
    ^YSize
    #ShortSupports
    #NormalSupports <Summary> >
    MaterialCodeE1 </ <
    MaterialCodeE2
    MaterialCodeE3
    MaterialLengthE1

    MaterialLengthE2
    MaterialLengthE3
    MaterialWeightE1
    MaterialWeightE2
    MaterialWeightE3
    CokeBottlesE1
    CokeBottlesE2
    CokeBottlesE3
    ModelHeight

    LayerCount
    LayerHeight
    Density
    Pattern
    PrinterModel

    Supports
    Version Raft
    Sidewalk
    XSize
    YSize
    DRM Material
    Time
    Firmware >V
    Minfirmware
    _distance>

    Layer <!--
    Below layer distances are calculated with following code from bfblib.cpp:
    bte(): dx = (x1 - x0) dy = (y1 - y0)
    mag = sqrt(dx * dx + dy * dy)
    float Dist = (mag / drawSpeed) * (extruderSpeed /10.0f) * 25.397f
    25.397f = filament pulled in per extruder stepper revolution
    Layer0 is sum of all other layers --> %p eE pP

    - - - - - - - - - -

    New finding...

    Indeed it is M227 that causes the retraction and sets the parameters for every M103 instance.

    Reducing F from say F600 to F60, this causes the retract to act very slowly.
    There are two instances I find that control this throughout the print and they are both in the initialization stage:
    M227 P450 S450 G450 F600

    Still taking them apart to find what the other 3 are really doing.

    - - - - - - - - - -

    More tests on the M227 command.
    M227 P450 S450 G450 F600

    You will find these in the bfb file.

    I've changed all the parameters and only one is clear...
    The F is a velocity call.
    Slowing F down will act like a dwell but it does cause the potential of drool.
    I am not sure how to label this M227 option from an effects standpoint.

    I am finding P450 to be somewhat benign.
    P150 and P10 seem to work just as well.

    The S value on the other hand is easily defined but probably has another component to consider.
    What is easily noted by the amount of retraction that happens.
    At the very least, this is the value in steps that the filament retracts during any M227 call or any M103 (all extruders off) calls.
    This should make short works of several things we should be aware of.

    The setting of S450 is part of the characterization of the Cube3 from 3DS. With heavier bowden tubes, the stretch within the tube is reduced.
    Changing this value to make the system "the same" would be warranted. Maybe something like S150 or so.
    How this value inter-plays with the other 3 is not yet known.

    At the very least, I can now test the MakeShaper filament to see if I can aleviate the turn-back voids using this one setting.

    - - - - - - - - - -

    Oh yea... the G450 doesn't make sense. It has something to do with a rate of extrusion.
    I went for G10 and the dwell at the start was very long... and the subsequent trace was very thin.
    Anything over G450 seems to create normal parts. This one probably has some good info behind it but I can't see it yet.

    - - - - - - - - - -

    I love it when things actually work!

    I took this well tested part (x2) and had issues with the MakeShaper ABS filament.
    The part on the right was how the MakeShaper version came out... voids after de-string operations.
    The part on the left is one where I adjusted the de-string operation in the .cube3 file.

    The difference is feak'n HUGE!
    The change was minimal.
    And I didn't even lower the temperature for the MakeShaper. It seems to work fine at 260*C.

    In short, I exported the .cube3 file from the library.
    I decoded the file and removed the binary hash at the bottom in Notepad.
    I did a global replace of 2 command line:
    M228 P0 S450 was replaced with M228 P0 S225 (don't ask me what the Snnn really does at the moment.)
    M227 P450 S450 (Gnnn Fnnn) was replaced with M227 P225 S225 (Gnnn Fnnn)

    These commands show up in a lot of places, but they are all associated with the M103 (stop all extruders).

    If you notice, I opted to change all the values to 1/2 the previous value.
    I chose M228 S225 because of it's mirror associativity. I figured that the S-value for M228 should match the S-value of the M227.

    I also decided the change the P-value of 450 to P225 for the same reason. The P and S seems to work together so they probably should match.
    I couldn't tell what changed when I changed the P value, but the S value makes a significant difference in the retraction length.

    The result was insanely positive. I went from a failing filament to a seam so clean it is hard to find.

    Now, I have a reason that this is happening... Not the ful pronouncement, but in part at least.

    The original bowden tubes on the Cube3 cartridges are very "weak"... they stretch when you put pressure on the filament.
    That means the 1/2 meter of cartridge tubing actually stretched some 6-8mm.
    And when the pressure is relieved, the retraction is dampened by the tube shortening again.
    As i now stare over to the CubePro printout, I also notice the much lower value for this M227 setting.
    There too the CubePro uses M227 P250 S250 G300 and F700 along with M228 S250...
    This gives me good reason to keep these 3 numbers at least matched.
    I am however hopeful someone can prove me wrong, or at least provide exact meanings in detail for each of these settings and how the Cubes use them.

    So, for my part, and since the advent of 4mm OD bowden tubes for the Cube3 printer, I've been watching this scenario for a while.
    Now I know that one, we can compensate and two, even the smallest changes we make to better our system, we affect their performance.
    This is easy to understand when you notice that 3D Systems spent quite a bit of time "formulating" and "optimizing" each system with very few commands and switches.

    ms1.PNG

    - - - - - - - - - -

    ...and if you were wondering if the build-plate still drops during a retract set to zero, yes it does in all cases.
    There is a special case for the Hershey... M227 P1 S1 G1000 and F1000. Yes, you can reduce or increase the Hershey with the G-value.

    - - - - - - - - - -

    from the BFBAxon manual... Not complete but nice to have!

    Gcode Outline

    Some basic knowledge is required in order to identify which extruder is used for a particular purpose i.e. raft, model or support material. It is also a
    requirement to identify which material the Gcode extruders are running by observing the temperatures in use.
    Raft layers
    The default raft generated in Axon consists of 2 layers at 0.7mm and 1.0mm.
    Material Temperature Ranges:
    ABS: 220 degC to 260 degC, the lower the temperature the less likely the layers will stick. 260 degC is the default temp and anything hotter than
    270 degC could potentially damage the extruder head over time.
    PLA: 180 degC to 220 degC, the lower the temperature the less likely the layers will stick. 195 degC is the default temp.
    Supported Gcode Commands
    Movement Codes
    G1 Coordinated Motion - supports X, Y, and Z axes
    G4 Dwell
    G90 Absolute Positioning
    G92 Sets an origin offset, which is currently used to set the extruder 2 X,Y offset from extruder 1. This can be overridden in the firmware.
    G93 Set ignore G92 setting
    Special functions (M Codes)
    M101 Turn extruder 1 on forward at the RPM specified by a previous M108
    M103 Turn all extruders off
    M104 S145.0 Set Extruder 1 target temperature to 145.0 degC
    M106 Turn part cooling fan on
    M107 Turn part cooling fan off
    M108 S400 Set Extruder 1 speed to S value/10 = 40rpm
    M201 Turn second extruder on forward at the RPM specified by a previous M108
    M204 S145.0 Set Extruder 2 target temperature to 145.0 C
    M208 S400 Set Extruder 2 speed to S value/10 = 40rpm
    M210 AUX 1 off (not on all hardware configurations)
    M211 AUX 1 on
    M220 AUX 2 off
    M221 AUX 2 on
    M226 Pause the printer until ESC key is pressed
    M227 S1000 P1000 on extruder stop (M103) reverse the extruder stepper for S turns and prime on extruder start (M101)
    M228 Disable M227
    M229 Same as M227 but P and S parameters specify prime and reverse extruder drive screw revolutions (not stepper steps)
    M230 Disable / enable wait for temperature change
    M301 Turn extruder 3 on forward at the RPM specified by a previous M108
    M304 S145.0 Set target temperature of extruder 3 to 145.0 degC
    M308 S400 Set Extruder 3 speed to S value/10 = 40rpm. (Note to turn off extruders 2 and 3 use M103 which turns off all extruders)
    M542 - Enter 3000 nozzle wipe / move to RapMan rest location
    M543 - Exit 3000 nozzle wipe / does nothing on RapMan
    M551 / M552 / M553 - Prime extruder 1 / 2 / 3. Parameter P = stepper steps, S = RPM * 10
    M561 / M562 / M563 - Reverse extruder 1 / 2 / 3. Parameter P = stepper steps, S = RPM * 10
    For a reference sheet, you can download a PDF of these commands



    Missing of note is M601 and M240

    I suspect M240 is "servo #-steps X/Y/S" and M601 are predefined positions at P# S60(?) F5 (velocity-slow)
    Last edited by TommyDee; 09-07-2018 at 11:30 PM.

  6. Thanks bolsoncerrado thanked for this post
  7. #116
    Regular 3D Printer
    Join Date
    Feb 2018
    Posts
    101
    Post Thanks / Like
    Sweet -- What a great improvement on the example part!

    If I remember right the Axon manual refers to a predecessor of the various Cube printers, prior to 3DS taking over BFB - right?
    It's pretty clear that 3DS modified some of the Axon commands a bit, and maybe added a couple of their own.

    My take is that the commands worth modifying - and therefore needing to be understood - are the temperature settings (Mx04), the extruder retract/prime (M227), and the relationship between extrusion rate (Mx08) and travel speed (the F value in G1) for various materials.

    (Are the Cubes unique in having the travel speed tacked on to the movement (G1) command?)

    The other commands used in the Init and Finish routines don't seem to change, so they can just be used as-is without needing to dissect them, IMO.

    It would be nice to be able to translate the S and P and G and F values into some standard notation (e.g., mm/second, microns, steps etc.) to make it possible to design parameter files for other slicers. (I know Simplify3D already has this covered, but it's too pricey for my hobby budget. And I'm kind of too lazy to get a Windows 10 running to use 3dBuilder.) I've gotten part way through modifying Slic3r to generate Cube3 bfb code, but sort of ran out of steam; I'm not really a C++ programmer...) Part of what I got stuck on is questions like "how many extruder rotation steps does it take to equal G1 x y z F1232.0 with G228 S23.0?"

    I say all that to express my appreciation to TD for the work on understanding 3DS bfb. And I know it IS work!

  8. #117
    3D Printer God(dess)
    Join Date
    Nov 2016
    Posts
    1,579
    Post Thanks / Like
    Man, I have numbers spinning around my head in all directions, but I know that is stated in these postings.

    This is right out of the 3DB DLL:
    Below layer distances are calculated with following code from bfblib.cpp:
    bte(): dx = (x1 - x0) dy = (y1 - y0)
    mag = sqrt(dx * dx + dy * dy)
    float Dist = (mag / drawSpeed) * (extruderSpeed /10.0f) * 25.397f
    25.397f = filament pulled in per extruder stepper revolution
    Layer0 is sum of all other layers --> %p eE pP

    That multiplier with the division by 10 corelates to the M55n and M56n commands.
    M551 / M552 / M553 - Prime extruder 1 / 2 / 3. Parameter P = stepper steps, S = RPM * 10
    M561 / M562 / M563 - Reverse extruder 1 / 2 / 3. Parameter P = stepper steps, S = RPM * 10

    DLL's is where machine specific calibration is managed.

    The M231/232/233 are also missing. But if you notice, this list is far more informative than any on RepRap Wiki!
    This information is the GENESIS of the Cube3! Go back and check out their manuals.
    Everything the consumer department at 3D Systems ever did brought about the Cube3!
    There is a soul there, and as long as that is true, we can gain knowledge.

    How about this... you can still install Axon right from the 3D Systems support site... and it requires Python which is free(?)... and you can output BFB ready files ready to be dropped into the heart of standard headers and footers. All the tweak settings in Axon that are there should be valid today! This means massive system control in settings rather than a canned cop-out by 3D Systems. They thought people would be joining their lot and they never came! So the entire slicer for 3DS thing just flailed as a profit center. Do CubePro people have any more access?

    And as to the extruder speed you were talking about... yes, you can easily make compensation changes for different filament drivers. The one above is specific to the Cube3 driver. You would simply need a multiplier in a spreadsheet to systematically tweak any filament feed call with the M108. M108 S400 Set Extruder 1 speed to S value/10 = 40rpm <= there's that /10 again! ...but 40 is from the RapMan.

    Care to take on Axon and see what their output files look like? Right up Windows' 7's alley!
    https://www.3dsystems.com/shop/support/legacy/downloads

    - - - - - - - - - -

    Duh! M231, M232, and M233 are the limits of the current platform!

    - - - - - - - - - -

    I have another find within the .cube3 files worth noting...

    When you have M227 P450 S450 G600 F800
    You could have M227 P400 S400 G750 F1000 commands as well.
    This happens when layers are not meeting minimal layer times.
    Guess how I found this little bugger!

    Also, if there are M227 P400 S400's...
    There are also remaining M228 P0 S400 calls that must be updated.
    Again, notice that the S values seem to remain paired between M227 and M228.

    This drove me nuts all day with terrible results.

    Also, there may actually be lower limits to these values as well.
    This requires more testing.

    I have not been able to simply disable the retraction as yet.

    - - - - - - - - - -

    Love it when a plan works!

    Almost no discernible seam on the right.

    noseam.PNG

    I only replaced the M227 P&S values with half their original value ...if value is greater than 1
    ...and the M228 S value when the S value matches the M227 S value.

    This is made simple with notepad, but confirm your data in Excel.

    You can also copy -from- Excel and replace the tab with space.

    This -is- the cleanest ABS print I've ever gotten form the Cube3!
    And this is without any heat shield.
    Last edited by TommyDee; 09-09-2018 at 11:05 PM.

  9. #118
    Regular 3D Printer
    Join Date
    Feb 2018
    Posts
    101
    Post Thanks / Like
    Quote Originally Posted by TommyDee
    Care to take on Axon and see what their output files look like? Right up Windows' 7's alley!
    https://www.3dsystems.com/shop/support/legacy/downloads
    I installed Axon 3 and built a little print file. It came out with a .bfbe suffix, which I take to mean "bfb-encoded".

    Trouble is, our .cube3 decoder doesn't decode it. We need a different encryption key.

    We are going to need some help from swwinterry I think.

  10. #119
    3D Printer God(dess)
    Join Date
    Nov 2016
    Posts
    1,579
    Post Thanks / Like
    I thought I saw a bit about output in binary or ascii option (?)

    Thanks for doing that. I ran into that trying to convert cube pro files too
    Did you know the Cube3 will open .creation files? Nothing else will!

    More tease:

    This is looking at the seam of the core... on the build plate... sole print... no cooling tower... at 260*C
    Remember the part is a hex/star profile...

    noseam2.PNG

    - - - - - - - - - -

    And one more. This time, I offer you to test this part for yourself along with the original. Link to the download below.
    Remember that the modified version has to be run from the USB port.

    seam_Q.PNG

    You're looking at the seam of a 3rd party ABS. Even with 3DS filament, I could never get this so clean!
    These are some of the best prints I've even gotten from the Cube3.
    Nearly perfect turn-backs and absolutely minimal blobs on layer starts.
    And no need for cooling towers to manage drool.

    I turned down the temperature on this one because the filament's rating.
    Therefore, the print file is the original and the modified version.

    Just to remind what this test is all about; this is to test BFB modifications to accommodate the 4mm Teflon bowden tube.
    My assertion is that the original cartridges have stretchy bowden tubes.
    The Cube3 was optimized for this by exaggerating the retract function normally related to destringing.
    Evidence from other Cube models shows that this is the right place to be looking for fine tuning.
    Therefore, yes, in order for you to test this part, you would need a rigid solution with one of the hub solution (doesn't matter which).

    Download link: https://drive.google.com/open?id=1ID...xz2TLtVFBeItbM
    Prints are left side ABS just to keep apples, apples.

    Mods:
    M227 P450 S450 G600 F800 => M227 P225 S225 G600 F800
    M227 P400 S400 G750 F1000 => M227 P200 S200 G750 F1000
    M228 P0 S450 => M228 P0 S225
    M228 P0 S400 => M228 P0 S200
    M104 S260 => M104 S250
    M204 S260 => M204 S250

    Notice that I've cut the modified values for retraction in half.

    - - - - - - - - - -

    If you wanted to write maintenance .cube3 files, what would you write?

    I'm thinking that testing simultaneous dual extruders for gap-matching, for instance.
    Or maybe something other than a test print to pull filament from the nozzle...
    Maybe a super retract to actually pull the filament completely from the nozzle automatically?
    This should be easy.

    Did you know that your test print is a simple .cube3 file that can be replaced on the internal SD?
    One is PLA and one is ABS. These two files should be updated just for semantics. The rook?

    Anyway... if anyone has a dire desire to help identify some useful routines, maybe we can implement them in some custom .cube3 procedure.

    Just remember that there is a very limited command set.

    - - - - - - - - - -

    For a moment I thought the decoder was limited to 12mb but I was wrong. It failed to decode a specific file. Keep that in mind if things look awkward.
    I just processed nearly 19mb through notepad and the coders with the above translation. No fuss from any of tools.
    I am using Excel to confirm no unwanted entries get through. Open Office works too but not quite as good on the filters unless I'm missing something.

    I'm going to suggest that these settings will correct the offset for the more rigid bowden tubes over the stock 3D Systems cartridge tubes.
    Not to suggest there isn't a hardware fix to deal with this as well. I started a solution long ago when people thought I was nuts for pursuing this anomaly.
    Don't worry, I'm still nuts And it is still worth pursuing.
    Last edited by TommyDee; 09-09-2018 at 05:54 PM.

  11. #120
    Regular 3D Printer
    Join Date
    Feb 2018
    Posts
    101
    Post Thanks / Like
    Quote Originally Posted by TommyDee View Post
    I thought I saw a bit about output in binary or ascii option (?)
    I haven't been able to find any option or setting for the file format in Axon 3 Alpha 1, 2 or 3, which are the only versions I can find.
    The manual that is available in 3ds archives is for Axon 2, and it refers to using any text editor to edit the files.
    From what I've found, it appears Axon 2 produces text .bfb files, while Axon 3 produces only encrypted files.
    I've not been able to find a download site for Axon 2.

    Does anyone know where to find an Axon 2 setup file?

    (It may not be worth the trouble, since the Cube 3 uses a different flavor of bfb. But it might be interesting to look at.)

 

 

Similar Threads

  1. [ HOW TO ] Use other slicers with cube 3
    By Mythandar in forum Cube3
    Replies: 309
    Last Post: 05-24-2018, 08:59 AM
  2. [ THING ] Printable Cube 3 Filament (re)Winder
    By chaman in forum Cube3
    Replies: 25
    Last Post: 08-01-2017, 04:35 PM
  3. Cube 3 enclosure
    By Bleky in forum Cube3
    Replies: 31
    Last Post: 03-31-2017, 04:34 PM
  4. ERROR Code T24: Temperature Error
    By bolsoncerrado in forum Cube3
    Replies: 3
    Last Post: 03-02-2017, 04:35 PM
  5. Converting 12v/24v power supply to remote (G-code) switching
    By bolsoncerrado in forum 3D Hardware
    Replies: 9
    Last Post: 06-30-2016, 04:24 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •