Page 3 of 9

Re: Cubepro gcode

Posted: 24 Jan 2015, 16:44
by martini
Update: I have printed many times using fixed bfb files and noticed that I can set everything to 0 and the printer won't even ask for a cartridge which is fantastic. On the other hand, it looks like the cubepro needs to know the material, not just the code but at least 0.001 length otherwise it gets really fast and messy.

Re: Cubepro gcode

Posted: 25 Jan 2015, 13:46
by Epop
What are you calling a 'big' file? (build size or file size and how much) I've run ~15MB builds through it without issue... it took a few min though.

CodeX.exe is the only utility I've found to turn *.bfb files into *.CubePro files(compile them so they'll run). The machine will read *.bfb, but it won't print them unless they've been processed. (*.bfb seems to be the file extension 3DS used to use)
martini wrote:Update: I have printed many times using fixed bfb files and noticed that I can set everything to 0 and the printer won't even ask for a cartridge which is fantastic. On the other hand, it looks like the cubepro needs to know the material, not just the code but at least 0.001 length otherwise it gets really fast and messy.
With all the header/ini settings at zero(or -1) my Cubepro v1.08a will not run without chipped cartridges. What settings did you get yours to run on (or do you just have firmware v1.00-1.03)?

I didn't notice a difference on mine... I just ran a ~25mm dia. puck. I could imagine that the firmware picks up some of the information for speed/ acceleration /deceleration, though, I don't think that would be the case.

One of the (few) things I picked up is the screen shows the build options. It has little icons for build strength, color, pattern etc. I thought that's all it was doing.

Cubeit is converting the files wrong, it's removing ALL M227(global retract/return?) words from the program. At the moment it looks like all its doing is removing kisslicer comments and code for reprap machines that CubePro isn't using. I'll try doing it manually and see what I get.

Unless any of you have seen it doing something else?

Re: Cubepro gcode

Posted: 26 Jan 2015, 09:56
by martini
Epop wrote:What are you calling a 'big' file? (build size or file size and how much) I've run ~15MB builds through it without issue... it took a few min though.

CodeX.exe is the only utility I've found to turn *.bfb files into *.CubePro files(compile them so they'll run). The machine will read *.bfb, but it won't print them unless they've been processed. (*.bfb seems to be the file extension 3DS used to use)

With all the header/ini settings at zero(or -1) my Cubepro v1.08a will not run without chipped cartridges. What settings did you get yours to run on (or do you just have firmware v1.00-1.03)?

I didn't notice a difference on mine... I just ran a ~25mm dia. puck. I could imagine that the firmware picks up some of the information for speed/ acceleration /deceleration, though, I don't think that would be the case.

One of the (few) things I picked up is the screen shows the build options. It has little icons for build strength, color, pattern etc. I thought that's all it was doing.
Cubeit is converting the files wrong, it's removing ALL M227(global retract/return?) words from the program. At the moment it looks like all its doing is removing kisslicer comments and code for reprap machines that CubePro isn't using. I'll try doing it manually and see what I get.

Unless any of you have seen it doing something else?
I have tried a big build size, a 20x20x20cm box (I wanted to see if with headers settings on 0 it would still run without enough material) and an organic shape with meshmixer supports of about 100mb but both .bfb files came out with 0 bytes. In fact, any file bigger than 20mb has come out empty.

I am running on v1.03. It doesn't make much difference if I change the rest of the headers or not, I if set material length on 0.000 it won't ask for a cartridge but unless an ABS or PLA icon is there (by setting material length) the printer goes nuts. Does your printer write the chip with filament used even with all settings at zero? Mine does. I wouldn't mind inserting a chipped cartridge if the printer is not writting it.

I am a mac user and kisslicer has a mac version but not cubeit nor codex seem to have one or at least I can't make them work yet; so I got virtual windows but kisslicer crashes. I'd totally buy kisslicer pro but it hasn't been useful for me so far, I just use native cubepro files, decode and recode back again.

Re: Cubepro gcode

Posted: 26 Jan 2015, 14:00
by Yodajammies
Just dropping in to confirm the current procedure for running Kisslicer files on the cubepro.

Running v1.08 using stock materials

I've replaced the header initialization values and the file is accepted by the machine but throws a filament error after warming up.

To confirm, is this about as far as we have progressed in getting a KS file to properly run on a cubePro?

Re: Cubepro gcode

Posted: 27 Jan 2015, 00:19
by Epop
martini wrote: I have tried a big build size, a 20x20x20cm box (I wanted to see if with headers settings on 0 it would still run without enough material) and an organic shape with meshmixer supports of about 100mb but both .bfb files came out with 0 bytes. In fact, any file bigger than 20mb has come out empty.

I am running on v1.03. It doesn't make much difference if I change the rest of the headers or not, I if set material length on 0.000 it won't ask for a cartridge but unless an ABS or PLA icon is there (by setting material length) the printer goes nuts. Does your printer write the chip with filament used even with all settings at zero? Mine does. I wouldn't mind inserting a chipped cartridge if the printer is not writting it.

I am a mac user and kisslicer has a mac version but not cubeit nor codex seem to have one or at least I can't make them work yet; so I got virtual windows but kisslicer crashes. I'd totally buy kisslicer pro but it hasn't been useful for me so far, I just use native cubepro files, decode and recode back again.
... Okay... big file. :)

I'm currently running a 20x20x20 cube file from Kisslicer(no cubeit)... for an hour... performance mon. tells me its currently using 640.5MB of memory and climbing... *.bfb is 125.6MB... so... memory leaks?

I'm running v1.08A... I can't look into why it would flip out on yours. I haven't tested to see if anythings being written to the chip. but I did start a run with ONLY my 'empty' chip in, and it started to print then alarmed out 'not enough material' even though it was in slot #3 and I was running the part from #1. it's a possibility its writing to any chip installed or... it just caught the '0%' flag... :/

OSX hmm... My brother's a Mac guy but he's about as handy as a bucket of sand in the desert(not knocking Mac... just saying). I'm going to try cubit one more time re-inserting codes that are missing. see if I get a better print with no other changes.
Yodajammies wrote:Just dropping in to confirm the current procedure for running Kisslicer files on the cubepro.

Running v1.08 using stock materials

I've replaced the header initialization values and the file is accepted by the machine but throws a filament error after warming up.

To confirm, is this about as far as we have progressed in getting a KS file to properly run on a cubePro?
I got the first layer to print! :)

and it went like this...
THIS file was *NOT* run through Cubit
Kisslicer init&post removed entirely
comments changed from ; to ^

XY off center!!! because...? kisslicer offset -OR- odd change in wipe positions from Cubepro.exe(?)
***Will "Crash"(belt skips teeth) intermittently on return to 'home' positions, had a few instances where it recovered correctly after filament error. the other dozen times it didn't... if you try it, be ready on a kill switch!
First Perimeter ran okay->good
First infill too slow or too much feed-squished/oozy

ran off to a weird place in +x and +y, came back and printed in ~correct location
Second perimeter(?) -still in 1st layer according to kisslicer
Filament fail- (did not modify start/stop throughout file)
so... Filament fail at layer change(start) is from no M228 (M227 off), or M227 (is removed or not actually added by Cubeit at first layer). I also added M108 (extruder deciRPM<?>) It was in the Cubepro.exe build at the same points.

Code: Select all

...
^InitComplete
M103
M561 P400 S500
G4 P1
^Select extruder, warm, purge ---------yes... its purging... again... :B
G4 P0
M601 P2 S30 F5
M228 P0 S1
M227 P1 S1 G1000 F1000
M240 S1400
M104 S220 P1
M104 S265 P1
G1 X108.000 Y136.000 Z0.4000 F750.0
G1 X108.000 Y161.000 Z0.4000 F9000.0
G1 X108.000 Y157.000 Z0.4000 F9000.0
M104 S250
M551 P1500 S50
G1 X108.000 Y136.000 Z0.4000 F9000.0
M601 P2 S30 F5
^ExtChangeComplete

M227 P1 S1 G1000 F1000  <<<<added

M106 P100
M304 S260 P1

M228 P0 S250  <<<<added
M227 P250 S250 G300 F700  <<<<added
M103

^ 'Perimeter Path', 45.5 [RPM], 10.0 [head mm/s]
G1 X-23.6 Y5.8 Z0.20 F2175.0
G1 X-23.6 Y5.8 Z0.20 F2175.0

^ extruder on
M108 S35.0  <<<<added
M101

G1 X-23.67 Y6.21 Z0.20 F2175.0
G1 X-24.38 Y6.82 Z0.20 F2175.0
...
...
...
G1 X-23.38 Y5.94 Z0.20 F2175.0
G1 X-23.79 Y5.97 Z0.20 F2175.0

^ extruder(s) off
M228 P0 S1  <<<<added
M227 P1 S1 G1000 F1000  <<<<added
M103

^
^ 'Loop Path', 45.5 [RPM], 10.0 [head mm/s]
G1 X-23.97 Y5.8 Z0.20 F9000

^ extruder on
M108 S35.0  <<<<added
M101

G1 X-24.72 Y6.44 Z0.20 F2175.0
...
<skipped to error>
...
G1 X-24.02 Y5.17 Z0.20 F2175.0
^ extruder(s) off
M228 P0 S1  <<<<added
M227 P1 S1 G1000 F1000  <<<<added
M103
^
^ 'Solid Path', 45.5 [RPM], 10.0 [head mm/s]
G1 X-28.32 Y7.41 Z0.20 F9000

----no M227/228

^ extruder on  
M108 S35.0
M101
<<<error>>>
G1 X-34.92 Y7.58 Z0.20 F2175.0
...
It ran through to the 2nd layer perimeter? Completed a fill, ran off to a weird place, came back, did another perimeter and failed.

Yay! Progress... :)

P.S. Cubit fails to exit. stays running in background. no warnings from anti-virus, yet... but... :x

**Updates CodeX is @1.5 hours and 800MB... shutting it down might try to run it later for giggles. :)
Cubeit is changing F words as high as 30000(thirty-thousand)

Code: Select all

M103
G1 X-27.28 Y-13.12 Z1.5 F30000
G1 X-27.28 Y-13.12 Z1.5 F3285
G1 X-37.24 Y-12.87 Z1.5 F3285
G1 X-37.27 Y-12.85 Z1.5 F3285
G1 X-37.27 Y-12.85 Z1.95 F210
inspecting Cubepro.exe files further into the programs I have yet to see an F word higher than 9000.0 (nine-thousand)... FYI

Re: Cubepro gcode

Posted: 27 Jan 2015, 00:56
by martini
... Okay... big file. :)

I'm currently running a 20x20x20 cube file from Kisslicer(no cubeit)... for an hour... performance mon. tells me its currently using 640.5MB of memory and climbing... *.bfb is 125.6MB... so... memory leaks?

I'm running v1.08A... I can't look into why it would flip out on yours. I haven't tested to see if anythings being written to the chip. but I did start a run with ONLY my 'empty' chip in, and it started to print then alarmed out 'not enough material' even though it was in slot #3 and I was running the part from #1. it's a possibility its writing to any chip installed or... it just caught the '0%' flag... :/

OSX hmm... My brother's a Mac guy but he's about as handy as a bucket of sand in the desert(not knocking Mac... just saying). I'm going to try cubit one more time re-inserting codes that are missing. see if I get a better print with no other changes.
Thank you so much for all your testings :D
I thought I had v1.3 but actually it's v1.5. Please let me know how it goes on your empty cartridge, do you think I should upgrade?
When there is no ABS/PLA icon it drops material at specified x, y points but doesn't run smoothly from one point to the other, instead the bed levels up and down to get to each point. If you look at the following code (from codex) you will see that the model starts to extrude at Z0.4000 but layer height was set on 0.200mm.

Code: Select all

^InitComplete
#Vector T22
M204 S260 
M228 P0 S250
M227 P250 S250 G300 F700
M103
G1 X12.160 Y13.167 Z0.4000 F9000.0
M108 S40.0
M201
G1 X12.970 Y12.375 Z0.4000 F1650.0
G1 X13.732 Y11.527 Z0.4000 F1650.0
G1 X14.435 Y10.635 Z0.4000 F1650.0
G1 X15.080 Y9.697 Z0.4000 F1650.0
G1 X15.662 Y8.717 Z0.4000 F1650.0
G1 X16.185 Y7.710 Z0.4000 F1650.0
I also sliced the same model on Kisslicer and it starts to extrude at Z0.2:

Code: Select all

^Firmware:V1..38.276
G21
G90
M542
M304 S230
M227 P296 S118
M228 P44 S44
M553 P700 S80
M543
M308 S346
M103
G1 X25.7 Y-44.27 Z0.45 F12000
G1 X25.7 Y-44.27 Z0.2 F210
M301
G1 X25.61 Y-44 Z0.2 F896.5
G1 X24.12 Y-42.63 Z0.2 F896.5
G1 X23.03 Y-41.82 Z0.2 F896.5
G1 X21.16 Y-40.64 Z0.2 F896.5
G1 X19.04 Y-39.69 Z0.2 F896.5
I mixed codex headers with kisslicer g1 codes and it worked and moved like printing but no material came out :shock: I cancelled, I wasn't sure if kisslicer/cubeitmod (cubex) are harmless to the printer, are they? I wonder if the bed leveling and messy drops come from missing information of a whole layer, from Z0.2 TO Z.04

Re: Cubepro gcode

Posted: 28 Jan 2015, 00:57
by Epop
martini wrote: ...
Thank you so much for all your testings :D
I thought I had v1.3 but actually it's v1.5. Please let me know how it goes on your empty cartridge, do you think I should upgrade?
When there is no ABS/PLA icon it drops material at specified x, y points but doesn't run smoothly from one point to the other, instead the bed levels up and down to get to each point. If you look at the following code (from codex) you will see that the model starts to extrude at Z0.4000 but layer height was set on 0.200mm.

...

I also sliced the same model on Kisslicer and it starts to extrude at Z0.2:

...

I mixed codex headers with kisslicer g1 codes and it worked and moved like printing but no material came out :shock: I cancelled, I wasn't sure if kisslicer/cubeitmod (cubex) are harmless to the printer, are they? I wonder if the bed leveling and messy drops come from missing information of a whole layer, from Z0.2 TO Z.04
Thanks for your feedback! Out of half a dozen forums/groups I've joined during the trouble with CubePro this has been the only forum that people haven't jumped in with "oh... that sucks. Here! buy *MY* printer!" :roll:

I would NOT upgrade unless you're going to commit to buying 3DS filament- typically any loopholes your version has would be closed if you upgrade to v1.08A.

I concur... I've noticed some funky movements going on in Z though the odd z height might be the 1st layer settings for Kisslicer/cubit (bed roughness is one setting that will change Z)

YES Kisslicer & Cubit (also if they are setup incorrectly) CAN Damage the equipment... wrong coordinates/heights, wrong temperatures, trying to feed way too fast, certain G&M words in the wrong places etc... can damage the parts, especially since they are lightweight.

personally IDC... the 90day warranty is long over and the prints coming from 3DS/CubePro @factory settings are weak IF they run at all, or IF you can remove the only support option you have (yes, or no) without destroying your part.

So...

Todays experiments were...

Home position grinding is... partially my fault it seems. After a fault I haven't been dragging the gantry off the Y limit switch(if you do it, turn the power off first). If I don't x,y is still messed up- if it attempts to return home it will grind depending on where it alarmed out (no m227 is different than no m108) and if I get a 'not enough filament' error it returns home properly if I have reset after fault. on either Cubepro.exe or Kisslicer/Cubit.

I corrected M227/M228 but ran into a layer that M108 was not directly above M101(for extrude start) so it seems it is required. (filament feed error)

My machine doesn't seem to be writing to the EEPROM on the good cartridges I have plugged in, or I just haven't printed enough for it to show.

The tallest build I have made so far was 5.4mm ... at that point there was ANOTHER check for 3D Systems filament. which it failed... I did not see any abnormal G code, and I've swapped out PLA & ABS cartridges and it doesn't seem to care as long as there is a cartridge with 'enough' filament in the machine to start. I don't know if its a timer built into the firmware or tied to my lack of #Vector T(11, 4, 12) words in the file... or its reading the Gcode and catches I have no cartridge in the extruder it's printing from... or what. :?

So, yeah, now I see... 3D Systems is not in the buisiness of 3D printers... they're in the business of selling filament at 400%-700% markup. :x

I didn't get much time to play with it today... that's all I got. :( We'll see how far I go down this CubePro road before getting off on the 'expensive' exit.

Re: Cubepro gcode

Posted: 28 Jan 2015, 08:57
by martini
Thanks for your feedback! Out of half a dozen forums/groups I've joined during the trouble with CubePro this has been the only forum that people haven't jumped in with "oh... that sucks. Here! buy *MY* printer!" :roll:

I would NOT upgrade unless you're going to commit to buying 3DS filament- typically any loopholes your version has would be closed if you upgrade to v1.08A.

I concur... I've noticed some funky movements going on in Z though the odd z height might be the 1st layer settings for Kisslicer/cubit (bed roughness is one setting that will change Z)

YES Kisslicer & Cubit (also if they are setup incorrectly) CAN Damage the equipment... wrong coordinates/heights, wrong temperatures, trying to feed way too fast, certain G&M words in the wrong places etc... can damage the parts, especially since they are lightweight.

personally IDC... the 90day warranty is long over and the prints coming from 3DS/CubePro @factory settings are weak IF they run at all, or IF you can remove the only support option you have (yes, or no) without destroying your part.

So...

Todays experiments were...

Home position grinding is... partially my fault it seems. After a fault I haven't been dragging the gantry off the Y limit switch(if you do it, turn the power off first). If I don't x,y is still messed up- if it attempts to return home it will grind depending on where it alarmed out (no m227 is different than no m108) and if I get a 'not enough filament' error it returns home properly if I have reset after fault. on either Cubepro.exe or Kisslicer/Cubit.

I corrected M227/M228 but ran into a layer that M108 was not directly above M101(for extrude start) so it seems it is required. (filament feed error)

My machine doesn't seem to be writing to the EEPROM on the good cartridges I have plugged in, or I just haven't printed enough for it to show.

The tallest build I have made so far was 5.4mm ... at that point there was ANOTHER check for 3D Systems filament. which it failed... I did not see any abnormal G code, and I've swapped out PLA & ABS cartridges and it doesn't seem to care as long as there is a cartridge with 'enough' filament in the machine to start. I don't know if its a timer built into the firmware or tied to my lack of #Vector T(11, 4, 12) words in the file... or its reading the Gcode and catches I have no cartridge in the extruder it's printing from... or what. :?

So, yeah, now I see... 3D Systems is not in the buisiness of 3D printers... they're in the business of selling filament at 400%-700% markup. :x

I didn't get much time to play with it today... that's all I got. :( We'll see how far I go down this CubePro road before getting off on the 'expensive' exit.
Me too! I am very determined to trick my machine into bulk filament first. Once I get there I'll have cheap material to mess with kisslicer settings. That's my plan :D It is very frustrating but quite fun at the same time.
From my experiments I have started to believe that maybe not all but some strange movements could be coming from Codex, not just from kisslicer having the wrong codes. Every single decoded file has come out with first layer at Z.04 and unless the printer knows what material is using it's like it doesn't know how to fix that. I modified a decoded file to start at z0.2 and no material came out but there was no shaky movements either. Printing starts on heating chamber phase, maybe the material is not hot enough for to it melt but this is not supposed to happen if I am not messing with native codes at all, just the headers, another reason to consider codex is doing something weird.

Yesterday, I came across a website called freelance.com and chatted with a firmware/microcontroller programmer. I explained to him what I was trying to do and gave him every information I have (codex, 1.8 update, gcodes, etc.) and he believes he can do a reinversed engineer. He's supposed to take a deeper look today and let me know price and procedure. If this works, maybe we could ask him for some help on kisslicer settings.

About supports, for most models I find meshmixer really helpful and material saver, specially for organic or complex shapes but for some others cubepro's are better. I usually print the models on ABS and supports on PLA. Then I put the model on a rice cooker until it boils a little. Since PLA softens at lower temperatures than ABS, supports come out very easily.

Re: Cubepro gcode

Posted: 30 Jan 2015, 11:10
by joydisee
Please let us know what's his proposition !

Thanks !

Re: Cubepro gcode

Posted: 30 Jan 2015, 14:09
by martini
joydisee wrote:Please let us know what's his proposition !

Thanks !
Of course!
So far he has checked all data and we have run some tests. I made a video and a pdf with results (pdf is basically because between my accent and printer`s noise it is bit painful to watch. If you are interested (or anyone in this forum :) ), you can give your email address and I`ll send them to you.
We are considering now a way to use voltage to reset the chip and/or using a dummy chip. He thinks we have very limited set of codes with gcode to find any loopholes and it's all about hardware mechanism 3DS has implemented for security. I have been reading the symptoms posted in this thread supposedly caused by kisslicer and I can see that I get the same issues even when I am not changing codes at all. Erase material length is enough to cause wobbling, speedy movements, and problems during chamber/nozzles warm up step on my printer.