3DukeEngineering

Are you talking about the lifting action of the secondary hot end? it’s a standard function of the UM3 print head, there is a little plastic dock attached to the right side panel of the printer, and a mechanical lever protruding out the print head, the print head moves to a position which engages the lever in the teeth of the dock, then by moving the print head fwd/backwards the lever is actuated which lifts/lowers the second print core. I imagine it would be quite hard to incorporate the flex drive extruder, as it would likely have interference with hot end 1’s bowden tube (the nozzles are very close to one another). That and the x/y movements of the print head are actually not ideal for a flex drive unless the flex cable can be positioned at 90deg (like the zesty nimble). The orientation of the flex drive’s cable means the filament is constantly advanced/retarded depending on the print heads position. The effect will be minute but print artifices can usually be observed. IMO, doing the MKII conversion is probably a more affordable solution, unless you happen to have access to a full um3 print head and print cores. Or better yet, put the money towards a new printer.

Hello, I am curious to know how, or if, Cura accounts for the time taken to switch nozzles/tools on their dual extrusion printers, and incorporates that accumulated duration into print time estimates. Print time estimates are very accurate for all of my dual extrusion Ultimakers, although a nozzle switch had to maybe be sub 5 seconds(?). The Problem however: I have a Toolchanger printer that picks up / drops off tools, and runs a purge and wipe sequence every time a new tool is called, this sequence takes apx 18 seconds, so it can easily throw print time estimates off by several hours. I am thinking if Cura already accounts for a predefined nozzle switch time for Ultimaker machines, then I may be able to edit a value to get a better time estimate on a custom machine profile. I have had a poke through the various files in definitions but nothing jumped out at me.

What printer and firmware are you running? there is a line in Marlin, something about “do not return to previous position” or something along those lines, seemed to do the trick

This can happen when you have two sort of concentric patterns that touch the build plate, and support required inside of that. Uncheck "Brim replaces support" and away you go. I realize this is an old thread but its still an issue, so maybe will help someone else in the future.

I am interested in learning how to write a script for post processing gcode and then implementing it into Cura. Can anyone recommend any good resources or examples to review that are specific to this? I know there is probably plenty online about script writing in general, but given the wide spread application I’d like to avoid picking up a degree just to do what I want. Reason being, I have just build my first IDEX printer, and having a fleet of Ultimakers, I am an avid Cura user. But the way Cura handles certain things in relation to tool changes, is just not very smart/efficient. a couple of thing I would like to do with a script: 1) edit the extruder 1 + 2 start gcode the first time it’s used in the gcode. 2) Remove every second M109 command after every T0/T1 command 3) Increase the feedrate for the first G0 move after every T0/T1 command

Hey @Bracco, are you able to describe what exactly your post processing script does? I have installed it (cura 4.8.0) but I don't know what the script is supposed to do or possibly what its not doing (for me)? what is the "layer start x/y bug"? I have just finished a IDEX build and the problem I now have is that on tool change T0 or T1, the finished extruder gets parked and the new extruder comes into play... however it moves to the last position of the previous extruder and sits there oozing until the nozzle hits temperature. ie, if I print two isolated cubes side by side, a white cube on the left with extruder 1, and a black cube on the right with extruder 2, once E1 has finished printing the first layer of the white cube, it is parked, then E2 moves right ontop the white cube layer and sits there waiting to reach reach temp, before then jumping over to the right to print the first layer of the black cube. I am a little lost on how to change this...

I too have this question, I have just put together an IDEX build. When a tool change happens, the selected extruder moves above the print into position and waits a bit to hit required temp whilst oozing, then immediately starts printing.

I wanted to resurrect this issue again as time has passed, Cura has had some revisions and this still drives me nuts. Does anyone know a UM contact they could tag that my able to look into my original post, as I think I have an idea that could be fairly easy to implement and would also fix other users issue of secondary nozzle ooze in their prints. If just to hear some imput from UM would be great. The idea is to have a prime tower consisting of an inner and outer rings/perimeters. The outer perimeter of the prime tower is printed during the nozzle warm up phase as is it does currently, then the model layer is printed as per normal, but as the nozzle enters the cool down phase in preparation to switch nozzles it would return to the prime tower and print the inner perimeter before retracting to a parked state. Each nozzle would print the prime tower both before and after the part layer whenever a nozzle switch is going to happen. The idea is to have the print head printing the prime tower during the warm up/cool down phase of both nozzles rather than having temperature changes of either nozzle taking place while the print head is printing/hovering above the part.

@geert_2, we haven't specifically tried any colorfabb materials. Unfortunately its not readily available in our country and is in the upper price bracket. I may order a spool as I have a job coming up where I need to purchase 15kg of steelfill from them. As for the holes, we don't use PVA. My personal opinion of PVA is that its terrible lol! My favorite is Ultimaker breakaway, and I use it solely as a support interface layer, it works wonders. But you are correct, about the tiny strings embedded into the walls of models. This happens both with PVA and Breakaway. The PLA when dissolved does leave holes. Also, because the PLA and breakaway don't adhere to each other, when there is a string of breakaway embedded in the wall, the string as actually not watertight even though its embedded. Fortunately these areas are easy to visually identify and patch up. What I have found with the UM3, is in preparation to switch to the secondary nozzle the PLA nozzle starts to cool while its still printing. This seem to have an under extrusion effect at the end of each layer before the nozzle switch. The under extrusion is so microscopic that you cant see it, but when I leak test the parts in water I can see bubbles forming in a row across particular layers. I have some further testing to do. I think I will increase the number of walls, change the wall printing order to print the outside first. And I will experiment with a secondary sacrificial prime tower in the hope of the nozzle cool down / warm up sequence taking place while printing this. For the interim to repair my current prints, I will be brushing the PLA models with chloroform using a sponge tipped brush much like cloakfiend has done with acetone. I had a bit of a play last night and I think the mechanical brush strokes seem to displace the surface layer and (hopefully) seal over the tiny holes. As for the casting photos, grey on the left is straight from the printer and the grey centered is chloroform vapor polished 😉

So I read the very first part of cloakfiend’s post and I was correct in stating PLA basically goes untouched by Acetone. It is only some brands that can be polished because they add modifiers to their PLA blends to improve the material characteristics and printing processes. Whilst I would love to switch to a brand that utilises modifiers in their PLA so it can be smoothed with acetone, for investment casting this is not an option 😞 The reason we use PLA for investment casting comes down to the characteristics of the plastic when it is being burnt out of the ceramic mould. Typically PLA burns very cleanly and without residue, it does leave behind some ash content but this can be washed or blown out of the ceramic mould. Typically any PLA that has been blended doesn’t burn cleanly at all. The modifiers melt/burn and tend to leave behind all sorts of sticky residue in the mould that is impossible to remove and this ruins the metal casting usually with porosity. The only exception we have come across is Polymaker Polycast. This material is PLA based but is polishable with IPA. It also burns extremely cleanly with fantastic results. The only down side is it totally crap for very intricate parts. It’s quite prone to stringing witch isn’t such a big deal but the adhesion between layers is extremely poor. For many basic shapes it’s perfectly fine, but for anything with intricate support where you need to use a bit of persuasion to remove it, you tend to break prints. just thought I would leave this info here for anyone else who comes across it. (I find it fascinating 🤓)

@geert_2, thanks for the info. So we typically chemical vapor polish most of the prints to be used as casting patterns. As we use PLA the polishing chemical we use is Chloroform. We can achieve some phenomenally smooth prints this way, and I had always just assumed it would play a part in sealing up any small holes but upon further inspection, it doesn't really. What is interesting is that anything printed on the UM2's generally don't have issues with small holes. Only parts printed on the UM3s which leads me to believe the heat up and cool down process between nozzle switching is introducing some small anomalies. I was always under the impression that acetone had no major affect on PLA, and in cases where it did soften the material, it was usually due to the additives and modifiers that are used and mixed into the filament when manufactured. But I have found cloakfiend's original thread, its rather large so will sit down and have a good read through. Then probably test it out myself. I also read through some gcode and managed to identify the m104/m109 extruder temperature change commands which appear to just be slipped in between a bunch of g1/g0 travel commands, but I don't seem to be able to identify any reasoning behind where Cura chooses to insert the temp change commands. This leads me to believe there is currently no settings within Cura to let me achieve what I want. But I am very keen to find a solution for this, at least to test results so if anyone has an idea please let me know. until then, here's some polished eye candy for you all. (casting made from solid bronze)

Hey guys, does anyone know if there is an option to park the print head in or near the prime tower while one print core cools down, and the other warms up? I see there are start printing and stop printing temperature that can be adjusted in Cura for each print core but I don’t think that adjusting these achieves what we want. I have noticed when the temperature transitions on particular layers that don’t require much extrusion, the head will park itself until the minimum temps are meet to carry on printing. All we were trying to do is let one nozzle cool to idle temp while the other heats to printing temp but we don’t want this to happen while printing. The reasoning behind this: we mostly create 3d printed patterns for metal casting purposes, and for many of our models we use breakaway for support roofs to achieve a detailed finish on overhangs. While the temperature transition between the print cores is happening there can be very minimal oozing and under extrusion that occurs. While this oozing and under extrusion has no major visual effect on the print, at a microscopic level it tends to create very small breaches in the outer walls or shells. The nature of ceramic slurry used in the casting process is very dense, approximately 2.5 x the density of water. The amount of pressure the slurry exerts on the print when submerged causes the slurry to make its way inside the print through these microscopic holes. This goes undetected until the casting is completed and results in poor quality or even failed castings. I will generally attempt to patch these holes but identifying them is an extremely tedious process and can take hours!

@superdave42 I was contemplating this last night and found this post now seeing you have done it! I have 2 x UM2s I’m planing on converting to this setup do you have any further info on wiring... I’m wondering if you managed to retain the white 10wire print head cable? Or if you managed to utelise a common power supply or ground between any individual heaters and temp sensors? would be nice being able to use the 10pin connector to interface everything with no more than 10 wires.

I recovered the firmware and no change. Also replaced the WIFI antenna and nothing 😞 Off to go get a network switch, wired connection still works...

Interesting. I just restored one of my UM3's from a recovery file on a SD card. The printer was having wifi issues (would not create a hotspot) which did not solve the issue by the way. Have just been calibrating the XY offset which seems to be printing a different calibration layout to my other 2 UM3's (all same firmware). And have now just seen the error mentioned in this post. Have triple checked that core AA 0.4 was used in slicer, AA0.4 is in the machine, and machine also reads that AA0.4 is installed! Been a sad sad week for my fleet so far 😞

Curious to see if you managed to resolve this? I too have the same issue. Would have been an early release UM3+ I have tried everything under the sun including reseating the antenna on the wifi board.

Just had an idea that I’m surprised hasn’t been included in Cura yet. I know we can add a material cost to see what each print is going to cost us in filament but I thought it would be really great if I could use Cura to price up print jobs. just by adding a few inputs we could have cura spit out a total print price. This could include things like filament mark up %, hourly rate, ect. Could get pretty extreme with what you want to base your price on but for me I generally charge a hourly rate plus materials with a mark up. I think this this would be a great feature to easily see what your cost price is and what your going to charge for a print.

Yep, I always check the layer view. Layers look perfect in Cura. When I put some pressure on the build plate so I can see the nozzle while it prints this problem layer, it actually looks like there is just not enough material coming out the nozzle (compared to any other layer). Anyway bedtime here (New Zealand) so I will tinker with a test piece once I have this job finished.

Correct this is not UM PVA, but I don't know what brand it is (not labeled) as for the PLA im currently using, its made by Imagin Plastics (a local company reputable for high quality filaments) This particular print is for a client, and time is ticking.... I have had to slice the model into seperate sections to avoid overhangs and I will glue them together. I will draw up a small test piece in a couple of days and have a play with temps and speed. cheers!

Ok so now I’m dumbfounded... I updated the um3 to the latest firmware, and leveled the bed + callibrated the offset. Deleted all cura profiles and reverted back to all the stock settings. Using the std cura um3 .15 layer height profile I still seem to have the same problem ? right now im totally lost. Perhaps it could potentially be something more sinister than firmware or slicing settings? The wifi board also seems to have failed as I cannot connect to wifi or run any update without a direct wan connection (this has been a problem for a long time though) First image you can clearly see the neat and tidy pva layers. second image you can see the pla it has attempted to print on top. 3rd and 4th images are close ups of what it tried to print on top of the pva, this looks like not enough material coming out of the nozzle to me. any suggestions?

Thanks @tinkergnome I hadn’t considered the nozzle offset, so maybe I have overlooked the obvious! I only ever manually level the bed as I found it to be much more exact. let me check it out and report back

Yeah sorry, the photo is a bit hard to understand without seeing the model. support interface thickness is set to 0.6 at 0.1mm layer height. I have just ramped it up to 1.0mm to see if that makes any difference but I don’t imagine it will. The pva interface always looks nice and sharp and clean. i have the z distance between model and support interface set to 0 is that correct? I have just tweaked it to 0.2 to see if that makes any distance. Perhaps having it set to 0 means the first pla layer is trying to print at the same layer height as the last layer of pva interface? (Surely this wouldn’t be the case?) would explain the underextrusion looking issue though...... although saying that if I put slight pressure on the build plate to lower it while it’s printing this layer, I can clearly see there is not enough material being pushed out the nozzle. no issue with pva sticking onto the pla beneath it. But when the first layer of pla is put on top of the pva interface it doesn’t stick very well. I realise pla generally doesn’t stick well to pva but in my opinion it’s not sticking well because it’s under extruded. i use locally purchased materials, unsure of brand as they are rebranded by the company who sells them. Been using them for a very long time without any issues. have printed with the same material combinations before without any troubles but that would have been way back on the firmware out of the box when first purchased (12m+ ago)

Yep it sure does. I have tried to highlight it in the photos. One photo clearly shows the pva support interface. The other shows what the first layer looks like. Sorry the photo isn’t the greatest. But you can see in the area I have circled what the problem is (hopefully)

Cura version 3.1.0 Not sure what you mean by "show as an error"? The printer doesn't realise its an error, but yes it is definitely an error. This happens with all files that require PVA support material

Hey guys, I have a odd one here with my UM3E. This seems to happen every time I print with PLA+PVA. Basically any PLA layer that is printed directly on top of a PVA support interface is extremely under extruded. Note its only the first layer thats under extruded, once it does the next layer up the extrusion rate is back to normal. For the life of me I can't figure out why, I haven't been able to find a relevant setting in Cura that relates to this specific layer and a google search hasn't helped me either. Im wondering if it might be a firmware glitch so am about to try a update but this has really got me stumped. Anyone have any thoughts?