Torgeir

Hi netminder, Just looked at your print project as I've also have an UM2. If I'd read your project file right, you're printing with a 0.6 mm nozzle and speed at 60 mm/s? Right? If so, this is way to fast for such a nozzle with such a detailed object I'll say. Also the flow request for this is very high. In order to print all this details you should use a 0.4 mm nozzle, a much better match for your print. Your printer can print this object very well. For Cura 4.6.1, use default fine, select your infill and select skirt. Reducing "print speed" is very effective in order to avoid issues in the object (resonance/shadows etc.). This should work well for your printer. (Actually I'll think your printer worked very good in spite of this to high demand 🙂). When it come to elephant feet, it is much better to have those -than a print come loose from the bed, oh yes, tried that.. Some use to reduce first layer in the original 3d drawing in order to reduce this effect. Hope this help Good luck Torgeir

Hi gr5, @gr5 You certainly made me thinking.. Those axis types have been using "multi coils" for years.. I wrongly used the pitch 2 deg, -but it should be multiplied by 4 as an UM2 use quad coils. So the real pitch is 8 deg., to be used for such a calculation. I'm going to correct the numbers to avoid confusion.. Thanks for your hint. Regard Torgeir

Hi alexgan, Hmm.. Seems there is something going on at the coupler. How is the coupler, the white insulator that's right above the heath block? If the coupler is too much compressed in between the fixed alu spacer and the heat block, the coupler may shrink and the passage for the filament will decrease. This will create a massive resistance for the "free" feed of the filament. This pressure is set by the stainless tube screw that's attached to the heat block, however if this screw is to loose, it may lead to a leakage of filament into the extruder that will ruin you day.. Do not ask me about how tight is enough, cause I never seen any info about this. I just turn it down until it is only in contact with the coupler, then I add about (5-10) degrees more turn -so that I feel "some" resistance. I usually put one mark on the screw and heat sink so I can see that it is held in the right place and do not move. So yes, there is a critical balance in tightening this screw.. These couplers is a consumable part and we should always have some in spare. Hope this help Torgeir

Hi there, No triplex, but a standard one with 2 deg. pitch. Regards Torgeir

Hi Folks, This tread is kind of old, but here's at last 3 different issues that can happen with any kind of printer.. So, to the first one that's very interesting, cause this is a relatively small round print with an overhang. This problem is not related to the breakaway support, but it is because of to much heat radiation to this part/side of the printed object. As the object is round, you will see that one sector will suffer for lo blast from one of the fan and this is the problem here. This could be cause one fan is not working properly, or just stopped. Also such object may not have sufficient cooling due to the position of the fan and the fact that the object is very close to the bed. To avoid this latter, the only way is to print with blue tape and heat bed off. The break away should have some distance to the "roof" as it should not stick to hard to the print, so around a millimeter or less is what I'm using -but depend a little of filament type as well. The second print with white PLA (I guess), is something else. This color is the most difficult one to print due to it's color, -white. From the thermodynamics this is well known, object with this color do not radiate it's heat energy as well as the object darker colors. So here it is better to be on the low end of temperature advised by the producer. In this print you see some brownish remains that are in the skin of the object, this also indicates to high temperature. With a small content of water present in the filament, very small holes might bee seen in your print. Well, this is just about the old issues up there. Thanks Torgeir

Hi Raymond, This is great information. Thanks Torgeir

Hi Johnny24, Welcome in here. I'll think 48 deg. C., on the outside metal cover is pretty high! This may lead to overheating of the small stepper driving ICS. If it's this hot I would have installed fan just to blow under there while printing. I'll think that there was planned to be a fan or two in the original edition of UM2. Somehow this was skipped, but maybe this was an optional thing for operation in hot countries? On the main PCB there is a connector to be used for a 24 V DC fan.. You may also check that the current settings for the stepper motors is at correct level that's normally 1200 mA. Good Luck Thanks Torgeir

Hi Alex, If the filament is soft and climbing up the tube, check that the aft cooling fan is working. This fan is keeping the cold side of the hot end "cold" in order to prevent such happen. Such thing may also happen if the printer is turned off directly after printing without cooling down. Thanks Torgeir

Hi 3D_rockstar, The best thing to do is just to get a new stepper of same type as used in the standard UM2. Other alternatives will not improve your Z resolution very much. Let me explain, it is not easy to just to swap the axes, this is because they operate with different shaft diameter vs pitch angle.. Edited: The pitch used for the Z shaft calculation 2 deg. pitch and this is to be multiplied with the number of coils in “parallel” in the shaft. An UM2 have a 4 coil shaft. So pitch to be used is 8 mm. Lets look at the shaft UM2 is using. The shaft diameter is 8 mm with 8 mm pitch (I.E. the axis move 8 mm per rev 360 deg.) Meaning we have 8 mm/200 step. Each full step is then 0.04 mm or 40 micron. Standard setting for Z use 8 bit micro step. So standard setting give; 8 mm/200/8=0.005 mm or 5 micron! If you like to improve this, you can install a jumper to use 16 bit micro step. With 16 bit setting; 8 mm/200/16=0.0025 mm or 2.5 micron. By looking at this numbers one can understand why Ultimaker selected 8 bit.. The reason I’m using 16 bit on Z and X/Y geared by two, is simply because I’m mostly print very small parts and most of my parts could be printed with an UM2 Go.. However, printing with this setting is very interesting -but sure kind of experimental. So, all in all, a small mechanical play will easily ruin this resolution by numbers.. But the very good thing here is that most of the resonance shadows in my printed model's is gone. 🙂 The above numbers is recalculated with correct pitch 8 mm/rev. Thank you @gr5 Hope this help you. Torgeir

Hi 3D_rockstar, First I’ll explain my interest in improving the mechanical tolerances. As thermoplastics is very difficult to keep within a given tolerances, I wanted to keep all this other mechanical play at a minimum level. I’ve also tried to get read of as much weight as possible from the moving extruder and the gantry system. This is important in order to have as little resonance as possible, as this often give your print object some shadow/echo effect. This is a typical after rapid accelerating/decelerating movement of the extruder head. Much of this problem is caused by the short belts stretching/flexing during heavy load during this process. This two short belts are the ones suffer due to the high torque that’s needed for this rapid movements. As I used a mechanical “belt gearing” 16:32 tooth pulleys on both X/Y axis, I’ve got higher resolution on those axes, as well I halved the torque on those two axis. You can see how this was done here: https://community.ultimaker.com/topic/31254-changing-an-um2-from-125-micron-to-625-micron-resolution-on-x-y-axes/?tab=comments#comment-257342 When it comes to the height axis Z, I’ve found that the pitch angle is important to leave as it is. This is because G play an important role in keeping the level as steady as possible, if this angle is less steep it is much easier to have issues due to this. I would rather go for the Igus type and the choose a step motor with higher step rate per/rev. But here ofc., using same pitch as the original Z-shaft. I have not printed the 3Dbenchy after I did this modification, but I've one other from before and after mod. I've printed the last one yesterday. An important note here, this two model is printed with the exact same gcode file from 2016! The one on the right side is printed after modification. Click on the picture to zoom in. I've splited the pictures into two part, not as good as it is originally, but you can see the ringing resonance after the /S (sss).. 🙂 Here is another picture of this flow test print of the eight step upward the speed increase. If you zoom in you will see the reduction of the resonance. The only issue after this mod is that the steppers X/Y now running at the double speed from before to keep same speed of printing as before. So the noise level is increased! I'll see if I can find some references from others I've used to collect. Regards Torgeir Edited, by some reason my pictures is much reduced in quality here...

Hi 3D_rockstar, Go for the old type h-stepper with fixed rod, -but if you really want to change this, -the only type that match good enough for an UM2 is the Igus type. The original step/unit for an UM2 is; X80.00 Y80.00 Z200 E282 To improve the resolution on the Z axis you just have to install a link on the 8/16 bit contact on the PCB, then the resolution is doubled to 400 step/unit for the Z axis. As a result of this link, your bed will only rice half of fully up position, so you need to change the number of step/unit in the firmware. By using Pronterface (or other method) and sending the following to the printer: M92 Z400 M500 The first string M92 Z400 change motion step for Z to 400 Step/unit and the second string M500 store the changes to the firmware. (Note: A factory reset will restore to the original setting.) This will improve the Z axe stepping much! I've doubled my step/unit for all tree axis, and my printer is much improved.. 🙂 So my printer setting X / Y is now 160 step/unit and Z is 400 step/unit. My resolution is for X/Y down to 6.25 uM (old 12.5 uM) and h down to 2.5 uM (old 5.0 uM). The result is amazing and I've never been able to make as god print as I do now. Also I can print overhang better due to this modification. Good Luck Torgeir

Thank you. I'm going to investigate this thing for sure.. (But disagree very much in naming our printed objects 2.5 D, they are really 3D object, so I'd rater said; welcome into the digital world... 🙂 ) Torgeir

I've to agree much with gr5 that's probably a faulty driver circuit.. There's 4 DMOS (power MOSFET) for each of the four wires going to the steppers, if just one of them fail, this kind of behavior is typical. If it was the other side failing like this, the extruder head would hit the wall directly. As you have another "spare" driver E2 (you have an old UM2 right?), it is possible to redefine the firmware to use the extra stepper driver serving the X stepper motor. You could always try to increase the current to 1400 mA, just to confirm this is not a friction problem. Running the printer with this high setting for long time can destroy the drivers, as there is no additional cooling to prevent overheating those integrated circuit. In a previous version of S3D I've had, the gcode files from this slicer always increased E1 to 1400 mA! Good luck. Torgeir

Hi Folks, How about this tread? Thanks Torgeir

To bad, the gcode monitor in S3D is the best there is IMHO. When it come to details, there is nothing as I'm aware of, that shows details as good as S3D. Also the process of replaying the "head track", as lines build up give you much information. Hmm. Lots of people name this part as "simulator", but it is just a replay function. Here is an example of a detail from a Fan Duct Holder print. The print is sliced very well, but have some artifact that is not a failure of the printer.. Here is a picture of the actual printed (detail) part: Here you can see the problems, yes small but... ----------------------------------------------------------------------------------------------------- Here is the detail picture of this sliced model shown with gcode viewer in Cura 4.6.1: No fail is shown here, so you may think this is your printer? --------------------------------------------------------------------------------- Here is the detail picture of this sliced model shown with gcode viewer in S3D: Here you can see the problem, so my printer is ok.. This is just an example of how important it is to use good known tools to avoid long waiting and many tries before a success. Well, some said good tools is half of the job.. Just my thoughts. Thanks Torgeir

Great Raymond! The changes in this PCB might only be the interface toward the stepper drivers, so this should be very easy to modify in software I'll think. This is a question for @tinkergnome. I've copied your picture for the ease of seeing details in here.. Regards Torgeir

Hi davidbitton, I'd never said so, but if you need to avoid traveling marks in "some" model you make, this old slicer may be the one to use for such a model. I do really also use Cura 4.6.1 ! Actually a number of versions of Cura, including S3D. Regards Torgeir.

Hi Raymond, Sorry to hear this info. That not a good start, -however, I would just buy one of that small PCB with a A4988 installed. Just swap and off you go.. Will this fan be possible to put outside the shielded box, in order to blow inside with suction from outside? My shielded box (to go over the PCB) had two holes for fans to be installed into. I'm using two 24VDC (in serial -less noise) fans mounted with white foam (to avoid humming noise from the chassis). This white plastic foam is taped with aluminium tape for sealing/attatchment and hold very well. Don't let this stop you, you are to close now.. Good luck. Regards Torgeir

Hi John, Yes the newer Cura is very good for some of the problematic issues, as "adaptive layers" for the overhang (among many thing), -there is just nothing compare to this printing without support up to 90 degrees!... However, for the issues you've address here, there is a "secret" (IMO) version of CUra that's outstanding and this one is hiding the travel lines inside the wall very well. An Old version -right, this is Cura 2.3.1.,! A high number of my very best prints is sliced with this version and I'm still using it very much. Note: Cura 2.3.1 do not have its own gcode viewer. But else a winner.... But make sure, if you'll try this version,; that's combing mode is off and that Z hop when retraction is not enabled. Very important! If you like I can show the difference between the versions is alike. Good luck Regards Torgeir

Hi Leporello, I'm sure my explanation is wrong, the initial layer is just one layer. I've missed this one as when they say initial layers, they mean all kind of first layers as brim, skirt, support and printed object.. Sorry by this was my bummer.. I've tried this, with increased value and then my squeezed first layer just would not stick! 😞 So this speed is for first layer only. So using the plugin, as Greg suggested is the key here, -or the hard way of hand "assemble" a gcode addition he also suggested, will work I'm sure. Thanks Good Luck Torgeir.

Hi, Did you ever check the coil current setting for X and Y axis. If the current is set to lo, it might loose step esp. in the beginning of print as there is some added friction during this part of printing. Thanks Torgeir

Hi clinton, This is a right triangle that's rotated 360 deg around the coiled rods "center axe" forming a cone. It can cut material outside or inside. (Find some tutorials about this issue in the help section.) Thanks

Hi clinton, Try to use a cone cut, instead of chamfer. Have never tried your way.. Thanks Torgeir

Hi DiogoLopes, Just want to warn you that the gcode viewer sometimes (-quite often) make your model with error that do not exist. My experience is that the Cura slicers become better and better in later versions, but this is not true for the gcode viewer. Try to use the viewer in Cura 2.7.0, much better IMO. This one is "quite" a lot like the one in S3D. Thanks Torgeir

Hi there, Can't you just adjust the initial speed, as this is for the first number of layers set by initial height ("initial layer height" divided by "layer height" give the number of layers at this speed)? Corrected Initial layer is ONLY the first layer printed. Just a thought, that's was wrong.. Thanks Torgeir