tinkergnome

@georgp - if sliced with Cura - it's a combination of two things: - the start script ends with the first extruder - for some reason Cura adds it's own tool change command before the start script, but not afterwards I don't know if it is possible to change the start script dynamically depending on the used extruders, probably not. I think the easiest way would be to add a second printer in Cura (exclusively used for this case) and change the start script. So that it will only prime (and use) the second extruder. The following is untested, only to give you an idea: G90 ;absolute positioning M82 ;set extruder to absolute mode M107 ;start with the fan off M200 D0 T0 ;reset filament diameter M200 D0 T1 G28 Z0; home all G28 X0 Y0 G0 Z20 F2400 ;move the platform to 20mm G92 E0 M104 T0 S0 M190 S{material_bed_temperature_layer_0} M104 T1 S{material_print_temperature_layer_0, 1} T1 ; move to the 2nd head G0 Z20 F2400 G92 E-7.0 G1 E0 F45 ;purge nozzle G1 X60 Z0.2 F5000 ; move away from the prime poop G1 X20 F9000 M400 ;finish all moves ;end of startup sequence

I think these explanations about the various grades are pretty comprehensive: https://www.supermagnete.de/eng/faq/What-temperatures-can-magnets-withstand https://www.supermagnete.de/eng/physical-magnet-data I guess the mechanism is a bit over-constrained now.... but if it works... who cares? 🙂

There's a "Support Placement" setting - sounds like switching from "Everywhere" to "Touching build plate" would be the easiest solution in this case? https://ultimaker.com/en/resources/52866-support

The switch to temperature offsets in the "tune" menu (instead of absolute values) was necessary because the gcode now contains a lot of temperature changes (for multi-extruder prints). I think there are four or five temperatures for each extruder in Cura. Any absolute tuning value would be permanently overridden by the gcode and therefor pretty useless. The +/- 25C approach was more or less adopted from the brand new UM3 back then. Just to be sure: Do you miss the offsets in the old firmware or the absolute values for the Mark2? BTW: Normal neodymium magnets are losing their magnetic effect irreversible above 80C. But there are neodymium magnets with higher temperature grades available.

100% infill is a special case. Cura will not generate infill at all in this case, but generates "bottom layers" all the way to the top of the model. That's why it looks different at 100% Perhaps the area is too small for the normal infill? I would expect a "gap fill" there instead (if it is activated - i think it is a "shell" setting)? In additiion: does Cura has settings like "minimal infill line length" or "minimal infill area"? I'm not sure - i would make all infill settings visible and take a look at the tooltips.

Well, your screenshot shows the settings for the first extruder, not the second. Can it be that simple?

Das "Troubleshooting" von Erin ist in der Regel sehr viel hilfreicher, oder ist es das, was Du schon versucht hast? https://ultimakernasupport.zendesk.com/hc/en-us/articles/115004169003-Reseating-the-Printhead-Cable-Ultimaker-3-Series- Wenn das öfter vorkommt, kannst Du bei der Montage auch noch ein kleines Stück Isolierband an der "Klemmstelle" um das Kabel wickeln, und außerdem das Kabel möglichst tief in den Druckkopf stecken, dann funktioniert die Zugentlastung besser. Hast Du ein relativ frühes Exemplar von dem Drucker (2017 oder früher)? Soweit ich weiß wurde diese neuralgische Stelle bei später gekauften Geräten verstärkt. Hier auf dem Bild sieht man ein neues Kabel mit extra "Manschette" (geht auch, aber Isolierband ist günstiger... 🙂 ) https://www.igo3d.com/Print-Head-Cable-UM3EXT

You're welcome. That's what we all are... I think regarding the 3D-printing electronics and software we are still experiencing the very first steps of many interesting possibilities. I guess even high end printers will appear as dumb as a stump in a few years... The printers - mind you - not we ourselves hopefully...🙂

Basically yes. There are always exceptions but those are probably negligible under normal circumstances. Every printer firmware supports a specific set of gcode commands and any setting that can be changed with a gcode will use the new value until the next reboot. On most printers many of the changeable settings can be stored in EEPROM and will survive a reboot then. There are fix configuration options as well, those can only be changed if one customizes the firmware. For example axis directions, moving limits, position of limit switches and such...

"nozzle is touching the build plate" is always treated as "z=0" - no exceptions (*) z-axis is moved to the "initial layer height" for the first layer... (0.27mm actual nozzle height in your example) ...and adds the amount of "layer height" to the actual nozzle height for all subsequent layers no magic, it's that simple (*) The manual leveling on Ultimaker printers (with the calibration card or a piece of paper) works no different. The thickness of the calibration card is compensated by the firmware, so that z=0 still means: "nozzle is touching the build plate". The assumed thickness of the paper differs a bit between the printer models though... it's 0.1mm (piece of paper) for the UM2-series and 0.15mm (calibration card) for UM3 and S5. The standard profiles in Cura are using a slightly thicker first layer. As @gr5 said - the intention is to level out unevenness of the build plate (there's no perfect even surface in the real world....). The thicker (but fewer) lines on the first layer are there to increase adhesion (AFAIK). I don't know the actual physics behind this, but i know that Ultimaker material engineers do a lot of tests and i tend to trust their huge experiences in this case... 🙂

as you said... PVA is tricky... (it may vary depending on the brand though...) I always remove the PVA from the printer (if i don't need it immediately again), and feed some cheap PLA through the BB core (just with the normal "load material" function). This removes any PVA residues from the print core and is much more easy and gentle than those hot/cold pulls (IMHO). I never had a clogged BB core since i use this method (no matter how long it rests unused in a drawer). I guess any leftover PVA will slightly expand in the nozzle when it absorbs moisture and then simply clogs, if it is reheated again...(but that's just my simplified theory)? Oh - and double check if the feeder is correct reassembled - especially take sure that this screwhead is back inside the housing.

If there are no options for the export units in your CAD software, just use the scale function in Cura. Cura 4.x shows the size in the lower left corner, so first scale the model to the expected dimensions and you're good to go.

For the sake of completeness: technically all settings have changed. Because every configured printer has a separate complete stack of settings which automatically loads when a printer is selected from the list.

@Joaquin87 well.... the model is very tiny.... only 0.3mm thick Depending on the first layer height it will be printed with only one or perhaps two layers. It seems that your expectations are different?

You can find the types on the printer in /usr/share/griffin/griffin/datatypes/eventType.py (needs active developer mode though) [...] #0x0000XXXX range, system related events SYSTEM_STARTUP = EventType(0x00000001, "System started") CRITICAL_ERROR = EventType(0x00000002, "Critical error {0}") SYSTEM_RESET = EventType(0x00000003, "Cleared all settings and history") SYSTEM_MAINTENANCE = EventType(0x00000004, "The maintenance action {0} was performed by {1}.") [...] #0x0001XXXX range, hotend related events HOTEND_CARTRIDGE_CHANGE = EventType(0x00010000, "Hotend {0} changed to {1} with serial {2}") HOTEND_MATERIAL_CHANGE = EventType(0x00010001, "Hotend {0} material changed to {1} by {2}") HOTEND_CARTRIDGE_REMOVE = EventType(0x00010002, "Hotend {0} removed") #0x0002XXXX range, print related events PRINT_STARTED = EventType(0x00020000, "Print {0} started with name {1}") PRINT_PAUSED = EventType(0x00020001, "Print {0} paused") PRINT_RESUMED = EventType(0x00020002, "Print {0} resumed") PRINT_ABORTED = EventType(0x00020003, "Print {0} aborted") PRINT_FINISHED = EventType(0x00020004, "Print {0} finished") PRINT_CLEARED = EventType(0x00020005, "Print {0} cleared") #0x0010XXXX range, authentication related events AUTHENTICATION_KEYS_ADDED = EventType(0x00100000, "API Authentication added for application: {1} user: {2} with id: {0}") AUTHENTICATION_KEYS_REMOVED = EventType(0x00100001, "API Authentication removed for application: {1} user: {2} with id: {0}") [...]

Yes, that's intended (on my side) - it's the ymmv-part of the example 🙂 You can change the last few movements if you prefer the actual "homing". Like so: ... T0 G0 X23 Y200 F10800 G28 X0 Y0 ;home head G0 Z200 F2000 G28 Z0 ;home buildplate ; disable motors M84 ... I would choose the closest variant (a.k.a. nozzle size) per extruder and adjust the line width to the real size. Line width settings are the ones that count (make all of them visible first - there are quite a few).

@canozerdem Before you think about 3D printing: your model looks very weird, i guess it is not intended to look like this...? In addition: it's not manifold and has very tiny details that will be not printable at all - at least not in this scale on an FDM printer. For those architectural models it is usually a good idea to divide it in several (flat) pieces that are glued together after printing. As said - the model is very small, in which size will it be printed? Don't waste your time with modeling of details that are smaller than the nozzle diameter... There are some good sources about modeling for 3D printing (even with Sketchup 🙄). It's probably not what you want to hear...., but the most important first task is to prepare a usable model. You can start with this link (for example): https://i.materialise.com/blog/en/3d-printing-with-sketchup/

In this case you can add these steps to the end-script in Cura or to the according gcode script in the printing host (if applicable). My "print finished" script in Octoprint looks like this: (it's just an example, YMMV) ;disable all heaters {% snippet 'disable_hotends' %} {% snippet 'disable_bed' %} ;end retract M302 S0 G90 G92 E11.0 G1 E0 F1500 M400 M302 S170 ;reset settings M501 ;move to parking pos T0 G0 X23 Y200 F10800 ;G28 Z0 G0 Z200 F2000 ; disable motors M84 M117 I guess that's a different issue. You can test the fans in the "Advanced" menu. Does it apply to both printheads?

Hi Georg, was the print started from the SD-card on the printer like common? The firmware should perform the finishing steps in this case...? 🤷‍♂️ Were the "cooling down" and "print finished" screens shown as usual?

Hi @Torgeir, that's not because you changed the micro-stepping mode, but because you print with very very thin layers? Or do i got it wrong? Can you explain it a bit more? What's the real issue that you're talking about here? What has material flow and Cura to do with the micro-stepping? More questions.... 🙂 Have you tested the maximum achievable speed? Does it still reach 300mm/s travel speed or will this burn the motors down? Can the poor little mega2560 hold up the pace?

Are those things not explained in the manual? @A3000T dunno if this is related in your case, but be aware that the material station is an "always on" unit, like a refrigerator. Never cut the power - at least not while sensitive materials are stored in it.

sorry, it's called "Merge Models" in Cura... https://ultimaker.com/en/resources/52852-merge-and-group-models I may be wrong, but i think @kfsone wants to achieve the opposite. He expects that both parts are printed only with infill (without top and bottom layers) in areas where they "touch" each other. Like: handle both models in the same way as a (merged) single shell, but use a different color (extruder) for the details. I'm not aware of any slicer that behaves this way. 🤷‍♂️

@MirkoKay As others said - it's nothing that the slicing software can solve for you. The model is not manifold and/or not "watertight". You have to build a manifold object if you want use it for 3D printing. I assume there are a lot of sources that explain how to use blender for those tasks (if you really have to...), here is an example: https://blender.stackexchange.com/questions/7910/what-is-non-manifold-geometry There are other free CAD programs that are much more suited for such technical models. I do not say that it will not be possible with blender, but there are choices... If you import your model in Meshmixer (for example), you can see that it consists of 6 separate shells that are not properly connected to each other, have overlapping areas, wrong normals and all such things. (or as @eldrick said... it's a mess 😃) In Cura: you can install the "Mesh Tools" plugin from the Marketplace. This will give you a warning when you load the model. It's good for a quick check. Don't waste your time with slicing as long as the model has errors. That's the best advice i can give. 🤷‍♂️

I guess you can't prevent this entirely. The base plate and the letters are two separate models after all and cura will always print them as such. Probably more important: both models should not overlap in the first place. The base should have appropriate pockets in places where the letters are located later. This would eliminate the need for "Remove Mesh Intersections" by the slicing software (as well as any other subsequent "tricks" and "mesh fixes"). I think the intended Cura function to align dual color models is called "merge origins", not "group". I'm not sure: is it possible to change the wall line count and/or bottom height in the "per model settings"? So that you can set both to zero, but only for the letters?

The schematics are published on Github: https://github.com/Ultimaker/Ultimaker2/tree/master/1091_Main_board_v2.1.1_(x1) There are a few unused Arduino pins exposed on headers on the main board (J22 - J26 and on EXP1 or EXP2 too). For more ambitious projects it's probably even possible to add more devices to the existing I2C bus (in addition to display and encoder led), but i have not seen that yet.