title_body
stringlengths
20
149
upvoted_answer
stringlengths
32
13.1k
TPE warping problems
I have not given up yet! And I figured it out, I think. I tried many things, dried the spool at 60C for an hour, Simplifiy3D's slicer, but eventually playing around I found a setting that resulting in no warping during the print. Final solution: - Original Prusa Powder Coated sheet. - Changed the nozzle to 0.6 mm. - 20 mm/s print speed, slow. - 260 C nozzle temp, make it liquid to get best bed adhesion. - 90 C bed, this keeps it soft at the bottom. Do not change the temperature, as this will detach the print. - 0.95 Extrusion Multiplier instead of 1.2 The extrusion multiplier did the trick. It kept pushing slightly too much stuff out, especially on the bottom layers causing ridges that is will pull on the next layer. It did still warp a bit, I found that this stuff shrinks about 2% when cooled, and when you start infill, it will pull itself loose. Hence the 90C bed temp. It still prints poorly, but it did print without creating "the blob"!
Folgertech FT-5 : Does not seem to be reading G-Code correctly
The image shows that somehow steps in Y direction (under the assumption that you have taken the photograph from the front) are missed. Missing steps can be caused by too high settings for acceleration and jerk, too much friction on the undercarriage or the belt assembly, undercarriage catching something at a certain point, too less current through the stepper. You could increase the torque on the stepper by increasing the Vref of your stepper driver is you have separate stepper driver modules, or a adjustment screw on the printer controller board. Furthermore reduce the acceleration and jerk settings and check the undercarriage for stick-slip movement by disconnecting the belt, then reconnect the belt and check for friction of the pulleys.
G-Code for resetting to 100% speed
You can put the speed to 100 % by G-code command: M220 S100. The M220command is described here. Know that speed changes sent to the printer have an effect on the next printed layer, it first finishes the current layer at the speed commanded before starting printing the layer.
First layer - not even infill
You're not sticking to your bed. Adjust your bed height. You are too far. Otherwise you are going too fast and or too hot. Calibrate the bed. Reduce speed. Then adjust temps. Also could be material contamination See this link for a visual troubleshooting. http://reprap.org/wiki/Print_Troubleshooting_Pictorial_Guide http://support.3dverkstan.se/article/23-a-visual-ultimaker-troubleshooting-guide In addition use some gluestick. That will often solve these issues. Last but not least add a raft if it continues. Or just ignore it. My bet is temp is too hot. I also like to smash my first layer, but not everyone likes that technique as it causes elephant footing.
Step motors vibrate, don't move at all (Prusa I3)
Usually if they get too hot you have to adjust the amperage by turning the little potentiometer on the A4988 (turn left until you can move the motor by hand, turn right until you cant, add a little bit like 1/8 to 1/4 of a turn). But this: I used a different stepstick RAMPS card and Mega, they worked perfectly. it is not completely clear, so: the A4988 work on another board -> your new board is broken another A4988 worked -> the A4988 is broken HTH
How to 3D print an object with variable printing speed?
As @fred_dot_u mentions, Simplify3D has the capability to do this, but you can achieve the same effect by slicing the file twice, once at 40 mm/s and once at 50 mm/s, and then manually combining the generated G-code using a text editor. You should be able to find the point where it transitions from printing the cube to printing the circular structure by looking at the Z-height, and you can simply copy-paste the G-code from that point onward from the 40 mm/s file into the 50 mm/s file. For instance, you might look for a point like this in the G-code (Cura): G1 F3600 X113.543 Y94.098 E105.96969 ;LAYER:25 G0 F7800 X113.743 Y91.369 Z2.700 or this (Simplify3D) G1 X175.981 Y74.019 F2880 ; layer 25, Z = 1.869 ; inner perimeter G1 X162.982 Y57.870 F7500 G1 Z2.069 F900 Note that the both include a comment line (a line that starts with ;) to indicate the layer. Everything before that you should copy from one file, everything afterwards from the other.
What setting do I need to change to get rid of this gap between the shells and the inside?
It's generally called infill overlap. In terms of first layer it's good to set extrude rate to 120% or even more so infill will overlap perimeters and itself. There is also the issue of "overlapping" layers which is not managed by any specific parameter. it's the issue of layer thickness and HE temperature. Unfortunately Matter Control has kinda bug or at least an issue with extrude rate. If you set higher extrude rate then overlap will be automatically reduced in some way. So it's good to increase extrude rate manually during printing and then set it back to normal. this will cheat slicer.
Slic3r under extrusion
It looks like you overtune it. I think it's because top layer of hollow object doesn't have any support (which is obvious) but because of that the filament is not oblate that's why it looks like underextruded. If you really need hollow object then set number of top layers to 3..5 then check if it helps. I know it's not really what you expect but IMO this could help here. (I usually set 2 bottom layers and 3 top and it's quite enought.) You can also experiment with speed of top leyer. I would say the faster the better in this issue as the the first top layer should be as flat as possible then next top layer will have better support. And of course cooling should be set to max (as for the bridges). Here is an explanation why you get underextrusion. A is wall (perimeter) B is (ceiling) top layer. Empty outlines are what you expect but filled shapes are what you really get. Differences are of course slightly smaller but it's to clearly explain the issue Please also check Printer Settings > Advanced > Top solid infill.
How do I remove a stuck resin print
There are generally 3 ways, in order of least to most desirable, and at times you need to combine them in an escalation: Toss the vat. This is the most expensive and generally should only be the last resort, for example, if you damage your film. Careful Brute Force. This can damage the film, but carefully getting a wedge between the film and the print might be the only option to remove the last bits sticking to the film. If you have to resort to this, take your time. Using thermal expansion. We all know that materials expand when you heat them. The reverse is also true: materials shrink if you cool them. Different materials shrink differently. So by cooling the vat with as little liquid resin and the print inside can lead to the print 'popping' free, just like in an FDM printer. Note however that this might not work on its own in case you made a suction cup (then you need to resort to brute force or drill a hole into the top), and you should take extreme caution not to contaminate any food with resin, as the stuff is not safe for consumption at all. With the right setup, you might be able to heat the build plate carefully while the print itself is kept cool.
Advantages of PETG filament?
PETG is great stuff to work with. It is stronger than ABS also. It prints slower than ABS and PLA. The formulas vary quite a bit from vendor to vendor. I have used 3 brands, and each of their properties vary. From my experience you do have to be careful with moisture. You'll be able to tell you have moisture in your filament if you start hearing a slight hissing and popping and an increased number of structural zits on the object. Moisture will also increase the problem listed in Mark's post below regarding the accumulation of filament on the nozzle.
MakerBot replicator 2x glitches
I don't have a profile that has settings for the dissolvable filament anymore, but this is one I use for thin layers (second extruder at 232C my first extruder isn't working so just ignore that one). You may want to try printing small simple objects with each extruder independently first to confirm that you have good settings, then try both together after you know you have good settings. ABS is a pain but mostly for warping and sticking to the build plate. The dissolvable filament I believe is PLA if you're using Makerbot materials. { "_attached_extruders" : [ "mk8", "mk8" ], "_bot" : "replicator2x", "_extruders" : [ 0 ], "_materials" : [ "abs", "abs" ], "adjacentFillLeakyConnections" : false, "adjacentFillLeakyDistanceRatio" : 0, "anchorExtrusionAmount" : 5.0, "anchorExtrusionSpeed" : 2.0, "anchorWidth" : 2.0, "backlashEpsilon" : 0.050, "backlashFeedback" : 0.90, "backlashX" : 0.0, "backlashY" : 0.090, "bedZOffset" : 0.0, "bridgeAnchorMinimumLength" : 0.80, "bridgeAnchorWidth" : 0.80, "bridgeMaximumLength" : 80.0, "bridgeSpacingMultiplier" : 1.0, "coarseness" : 9.999999747378752e-005, "commentClose" : "", "commentOpen" : ";", "computeVolumeLike2_1_0" : false, "defaultExtruder" : 0, "defaultRaftMaterial" : 0, "defaultSupportMaterial" : 0, "description" : "External Definition", "doAnchor" : true, "doBacklashCompensation" : false, "doBreakawaySupport" : true, "doBridging" : true, "doDynamicSpeed" : false, "doDynamicSpeedGradually" : true, "doDynamicSpeedInteriorShells" : false, "doDynamicSpeedOutermostShell" : true, "doExponentialDeceleration" : false, "doExternalSpurs" : true, "doFixedLayerStart" : false, "doFixedShellStart" : true, "doInfills" : true, "doInsets" : true, "doInternalSpurs" : false, "doMixedRaft" : false, "doMixedSupport" : false, "doOutlines" : true, "doPrintLayerMessages" : false, "doPrintProgress" : true, "doPurgeWall" : false, "doRaft" : true, "doSplitLongMoves" : true, "doSupport" : true, "doSupportUnderBridges" : false, "endGcode" : "", "exponentialDecelerationMinSpeed" : 0.0, "extruderProfiles" : [ { "bridgesExtrusionProfile" : "bridges", "feedDiameter" : 1.820000052452087, "feedstockMultiplier" : 0.9300000000000001, "firstLayerExtrusionProfile" : "firstLayer", "firstLayerRaftExtrusionProfile" : "firstLayerRaft", "floorSurfaceFillsExtrusionProfile" : "floorSurfaceFills", "infillsExtrusionProfile" : "infill", "insetsExtrusionProfile" : "insets", "layerHeight" : 0.20, "maxSparseFillThickness" : 0.20, "nozzleDiameter" : 0.40, "outlinesExtrusionProfile" : "outlines", "raftBaseExtrusionProfile" : "raftBase", "raftExtrusionProfile" : "raft", "restartExtraDistance" : 0.0, "restartExtraDistance2" : 0, "restartExtraRate" : 25.0, "restartExtraRate2" : -1, "restartRate" : 25.0, "restartRate2" : 25, "retractDistance" : 1.700000047683716, "retractDistance2" : 0, "retractRate" : 25.0, "retractRate2" : 50, "roofSurfaceFillsExtrusionProfile" : "roofSurfaceFills", "sparseRoofSurfaceFillsExtrusionProfile" : "sparseRoofSurfaceFills", "toolchangeRestartDistance" : 18.50, "toolchangeRestartRate" : 6.0, "toolchangeRetractDistance" : 19.0, "toolchangeRetractRate" : 6.0 }, { "bridgesExtrusionProfile" : "bridges", "feedDiameter" : 1.769999980926514, "feedstockMultiplier" : 0.9300000000000001, "firstLayerExtrusionProfile" : "firstLayer", "firstLayerRaftExtrusionProfile" : "firstLayerRaft", "floorSurfaceFillsExtrusionProfile" : "floorSurfaceFills", "infillsExtrusionProfile" : "infill", "insetsExtrusionProfile" : "insets", "layerHeight" : 0.20, "maxSparseFillThickness" : 0.20, "nozzleDiameter" : 0.40, "outlinesExtrusionProfile" : "outlines", "raftBaseExtrusionProfile" : "raftBase", "raftExtrusionProfile" : "raft", "restartExtraDistance" : 0.0, "restartExtraDistance2" : 0, "restartExtraRate" : 25.0, "restartExtraRate2" : -1, "restartRate" : 25.0, "restartRate2" : 25, "retractDistance" : 1.399999976158142, "retractDistance2" : 0, "retractRate" : 25.0, "retractRate2" : 50, "roofSurfaceFillsExtrusionProfile" : "roofSurfaceFills", "sparseRoofSurfaceFillsExtrusionProfile" : "sparseRoofSurfaceFills", "toolchangeRestartDistance" : 18.50, "toolchangeRestartRate" : 6.0, "toolchangeRetractDistance" : 19.0, "toolchangeRetractRate" : 6.0 } ], "extruderTemp0" : 228, "extruderTemp1" : 232, "extrusionProfiles" : { "bridges" : { "fanSpeed" : 0.50, "feedrate" : 40.0 }, "firstLayer" : { "fanSpeed" : 0.50, "feedrate" : 10.0 }, "firstLayerRaft" : { "fanSpeed" : 0.50, "feedrate" : 50.0 }, "floorSurfaceFills" : { "fanSpeed" : 0.50, "feedrate" : 40.0 }, "infill" : { "fanSpeed" : 0.50, "feedrate" : 40.0 }, "insets" : { "fanSpeed" : 0.50, "feedrate" : 40.0 }, "outlines" : { "fanSpeed" : 0.50, "feedrate" : 10.0 }, "raft" : { "fanSpeed" : 0.50, "feedrate" : 90.0 }, "raftBase" : { "fanSpeed" : 0.50, "feedrate" : 10.0 }, "roofSurfaceFills" : { "fanSpeed" : 0.50, "feedrate" : 90.0 }, "sparseRoofSurfaceFills" : { "fanSpeed" : 0.50, "feedrate" : 90.0 } }, "fixedLayerStartX" : 0.0, "fixedLayerStartY" : 0.0, "fixedShellStartDirection" : 215.0, "floorSolidThickness" : 0, "floorSurfaceThickness" : 0, "floorThickness" : 1.0, "gridSpacingMultiplier" : 1.0, "infillDensity" : 0.3000000119209290, "infillOrientationInterval" : 90, "infillOrientationOffset" : 0, "infillOrientationRange" : 90, "infillShellSpacingMultiplier" : 0.70, "insetDistanceMultiplier" : 1.0, "jsonToolpathOutput" : false, "layerHeight" : 0.1199999973177910, "leakyConnectionsAdjacentDistance" : 0.0, "maxConnectionLength" : 10.0, "maxSparseFillThickness" : 0.1000000014901161, "maxSpurWidth" : 0.50, "minLayerDuration" : 5.0, "minLayerHeight" : 0.010, "minRaftBaseGap" : 0.0, "minSpeedMultiplier" : 0.30, "minSpurLength" : 0.40, "minSpurWidth" : 0.120, "minThickInfillImprovement" : 1.0, "modelFillProfiles" : {}, "numberOfShells" : 2, "platformTemp" : 110, "purgeBucketSide" : 4.0, "purgeWallBaseFilamentWidth" : 2.0, "purgeWallBasePatternLength" : 10.0, "purgeWallBasePatternWidth" : 8.0, "purgeWallModelOffset" : 2.0, "purgeWallPatternWidth" : 2.0, "purgeWallSpacing" : 1.0, "purgeWallWidth" : 0.50, "purgeWallXLength" : 30.0, "raftAligned" : true, "raftBaseAngle" : 0.0, "raftBaseDensity" : 0.6999999880790710, "raftBaseLayers" : 1, "raftBaseRunGapRatio" : 0.8000000119209290, "raftBaseRunLength" : 15.0, "raftBaseThickness" : 0.3000000119209290, "raftBaseWidth" : 2.50, "raftExtraOffset" : 0.0, "raftFillProfiles" : {}, "raftInterfaceAngle" : 45.0, "raftInterfaceDensity" : 0.3000000119209290, "raftInterfaceLayers" : 1, "raftInterfaceThickness" : 0.2700000107288361, "raftInterfaceWidth" : 0.4000000059604645, "raftModelSpacing" : 0.3499999940395355, "raftOutset" : 4.0, "raftSurfaceAngle" : 0.0, "raftSurfaceLayers" : 3, "raftSurfaceShellSpacingMultiplier" : 0.6999999880790710, "raftSurfaceShells" : 2, "raftSurfaceThickness" : 0.1400000005960465, "roofAnchorMargin" : 0.40, "roofSolidThickness" : 0, "roofSurfaceThickness" : 0, "roofThickness" : 1.0, "shellsLeakyConnections" : false, "solidFillOrientationInterval" : 90, "solidFillOrientationOffset" : -45, "solidFillOrientationRange" : 90, "sparseInfillPattern" : "hexagonal", "splitMinimumDistance" : 0.40, "spurOverlap" : 0.0010, "startGcode" : "", "startPosition" : { "x" : -112, "y" : -73.0, "z" : 0 }, "supportAligned" : false, "supportAngle" : 30.0, "supportDensity" : 0.2000000029802322, "supportExcessive" : false, "supportExtraDistance" : 0.50, "supportFillProfiles" : {}, "supportLayerHeight" : 0.2000000029802322, "supportLeakyConnections" : false, "supportModelSpacing" : 0.2000000029802322, "supportRoofModelSpacing" : 0.4000000059604645, "thickLayerThreshold" : 0, "thickLayerVolumeMultiplier" : 1, "travelSpeedXY" : 150.0, "travelSpeedZ" : 23.0, "version" : "3.9.4" }
Why is cupping bad in SLA prints?
The printer prints, then moves up, then down again. The print surface stays inside the resin vat at all times. As a result, we have this experiment: The "bottle" is resting in a vat of liquid. As we raise it more and more, it does not drain until the lower lid is free of the liquid surface or some point of the shell delaminates. The release of the resin can happen in a rather violent way - which in turn could deform the print in the making. Delamination rips through the part till air can enter the enclosed space, destroying the print in the process. Even if printing the mouth against the plate you'll have cupping if you have a solid plate to print against. This can be mitigated with a little angle but trap liquid in the print at the end or including a couple of small gaps close to the surface to allow air to get into the print - yet unless the resin can flow out at the bottom some will be trapped in any case. To prevent cupping, I would turn the cup to print sideways, that way resin and air can be exchanged.
Is it possible to know which is the correct temperature range and speed for any model?
The first indication for print speed and temperature should be taken from the box the filament comes in. Generally it specifies temperature ranges for the hotend and the heated bed. Sometime, mostly online, more parameters can be found amongst which is the printing speed. Do note that temperature and printing speed are linked, if you want to print faster you should increase the temperature. But, if you are printing small or thin things you should print slower so that the part cools enough for the next layer. Basically, part cooling is then also important, but not all filament types (e.g. the ones with a high melt temperature like ABS or PETG) like being cooled too much. So you have another parameter to consider. It is difficult to instruct you to print at a certain speed and certain temperatures as it is highly depending on the filament (e.g. also the filament diameter), the machine type/make and model, extruder setup (direct or Bowden), the print, enclosure, etc. Because of the many parameters affecting printing, it is usually suggested to calibrate the printer by printing a temperature tower or performing retraction tests to find the print window for your specific setup.
Anet A8 beginning layers not printing correctly
With respect to the filament lifting off, and/or not adhering correctly, on the first layer, see Filament lifts from the hot bed while printing... in particular, you may need to clean the bed, calibrate the leveling and the nozzle height (see my answer). The fact that the prints are fine, after 10 or so layers, could be to do with the heat - at the higher layers the print is warm and no warping occurs, but below that, maybe the print bed is acting as a heat sink and making the filament cool too quickly - hence the distortion. Are you using a heated bed? If not, then maybe consider installing one. Not sure about the bubbles: how do your store your filament? The bubbles could be due to water, which expands as it gets hot, thereby causing the filament to bubble. It is best to store your filament in an air tight container, to prevent it collecting moisture and dust from the air.
Ender 3 Random Axis Shift
The usual suspects are overheating stepper motor drivers and over-tight belts. It can also be caused by belts coming loose in their mounts.
What causes these blobby corners
I suspect you are printing through a usb or network connection, and the communication rate it's to slow for any of many reasons. A curve consists of many tiny linear movements, each requiring a command exchange between the PC and printer. If you can, try printing from an sd card plugged into the printer (I'd the printer is so equipped). This could be worse if the uses a Bowden extruder, since there is now compression and windup in the filament.
Stepper motors making noise after steps/when idle
When idle, the stepper is stationary, no rotation. Normal standard electric motors will start spinning as soon as you apply power to them. However, steppers only rotate when a magnetic field is applied1): Stepper motors effectively have multiple "toothed" electromagnets arranged around a central gear-shaped piece of iron. The electromagnets are energized by an external driver circuit or a micro controller. To make the motor shaft turn, first, one electromagnet is given power, which magnetically attracts the gear's teeth. When the gear's teeth are aligned to the first electromagnet, they are slightly offset from the next electromagnet. This means that when the next electromagnet is turned on and the first is turned off, the gear rotates slightly to align with the next one. From there the process is repeated. Each of those rotations is called a "step", with an integer number of steps making a full rotation. In that way, the motor can be turned by a precise angle. The motor's position can then be commanded to move and hold at one of these steps without any position sensor for feedback (an open-loop controller), as long as the motor is carefully sized to the application in respect to torque and speed. When you power the printer and energyze the steppers there is no movement, but, the magnetic coils in the stepper are activated to hold the rotor in position. This is controlled by the stepper driver. The creation of the signal for the magnetic coils is causing the noise. It is a function of the driver type, micro-stepping setting, the stepper motor inductance, current setting and supply voltage. If the stepper motor, stepper driver and power supply can take it, increasing the current setting of the driver may lower the noise. Turning off the steppers (disabling them with G-code using M2, M18, or M84 depending on your firmware) will stop the noise, but you will easily lose the current position as it is not hold into place anymore. 1) source Wikipedia
Bed wobbling on Ender 3
If the V roller wheels aren't tight on the Y axis beam, it means the eccentric nuts are not adjusted correctly. Two of the rollers are mounted centered on the holes in the carriage frame, but the other two are on eccentric nuts which displace them from center slightly depending on the orientation the nut is turned to, to allow tightening and loosening of the grip on the beam. Since the Y axis ones are hard to see under the bed, look at the X or Z ones to get an idea what to expect. Note that the bolt through the whole roller assembly can loosen when adjusting the eccentric nut. You can probably avoid this by figuring out the right direction to turn it and only going that direction (continuing around just under 360 degrees if you go too far). If you do loosen the bolt then the eccentric nut will move on its own under vibration, so you need to re-tighten it. For the Y axis this might require taking off the bed or taking the carriage off the beam (by removing the belt and tensioner).
Odd print artifacts on same layer as large top surface
Just a guess here. If I understood correctly the deformation is at the first layer of the box bottom, it seems highly unlikely that it randomly happened at this exact position so I am assuming that if you do the print again you will get the same deformation at the same place (you can test this maybe ? I know it is a long print). In this case it is more likely that is a slicer issue than something with your machine and it could also be a problem with the stl model. Maybe try to create a small simple box then slice it using the same setting and try to see if you can reproduce artifact.
precise transformation using meshmixer
It's possible to close over a hole using Meshmixer. I've done exactly that recently, although the "hole" was a depression, the process would be the same. As you've discovered, Meshmixer can be considered somewhat imprecise. Fusion 360 will import the model you wish to modify. You would then turn off edit history, convert the model to BREP, then perform the edits you require. Once you are satisfied with the results, it's an easy matter to export the model as an STL file. I've summarized the steps, which are almost as easy as my description. You'd have to combine Google-Fu with the summary for the detailed portions, but it's something I've done in the past. Fusion 360 is free for hobbyists, renewable each year. There are many YouTube tutorials and text-referenced solutions for the steps required to accomplish your edit. Some will consider Blender to be an alternative method and I agree that it's a viable option, but it's not quite as intuitive as Fusion 360 and was more challenging for me to embrace.
Creality 3D Ender-4, a "extruder kit" does not fit the frame
From the photo on their website, it looks like you are orienting it correctly. Here is a photo to confirm the orientation. Note: I added this as an answer so I could share the photo
Internal thread not sticking to inner wall
I do a lot with printed threads, and find this problem creeps up mostly when the filament is wet. Higher temperature (like 220) can partly compensate but it's better to dry your filament and store it properly to keep it dry. Assuming standard metric thread profile, though, these are also pretty severe overhangs for a concave extrusion (where motion of the head will tend to pull the way you're seeing) unless you use very thin layers. You probably need 0.16 mm or thinner to print reliably (independent of pitch, though finer pitch also needs finer layers for other reasons), and limited acceleration for outer walls (I use 500 mm/s²). If you slice with "outer walls first" turned off, you may be able to get by with thicker layers, provided you have everything else tuned perfectly to ensure the outer wall sticks to the previously extruded inner one (i.e. flow rate perfect). A contributor to your problem could also be underextrusion due to oozing during combing. You might try setting max comb distance very low (like 0.6 mm, around what I use) or turning off combing and see if that helps.
How to stop objects with "floating" parts from breaking
The small filaments you remove that hold the parts up are called supports. The one model I located on Thingiverse clearly requires a number of supports, as the model is not easily designed for 3d printing with FDM printers. It would be better printed with SLS, but that's not the focus of your question. You don't specify how large you printed the model, but certainly a scaled-up version will be stronger at the weak points. You will want to use sharp non-shearing cutters to clear away as much of the supports as possible, without torquing on the model. Another option which also reduces the forces on the model body is to use a soldering iron to smooth and clear/cut the supports. If you are able to use cutters and not damage the model, the soldering iron can remove and flatten the remnants of those supports. Please note that if your careful work has resulted in a model that snaps to pieces, your little brother will soon destroy the successfully cleaned up model just as easily. If you have skill with 3d modeling software (Meshmixer and Blender come to mind for such organic models), you can add insignificant items to the model to provide functional support. Would the Rayquaza look fiercer if you 3d printed a cage as an integrated part of the model, using the bars of the cage to provide support? I successfully printed a model that was created by an artist unfamiliar with 3d printing restrictions. The support material was wash-away PVA. I provided the model to the "owner" who washed away the support material and snapped the legs in two. It's sometimes impossible to solve poor designs. You have a good chance if you build a cage for this one.
PLA print warping on heated bed
Several things I've done to stop warping when it occurred: Use a wider brim. If the brim isn't sticking, use a higher bed temperature for the first layer. If the brim comes up only on one side or warping is only on one side, make sure the bed is level. Slowing down the print will keep you from having as long of a strand cooling down. This will lower the contraction force and reduce warping. A hotter bed temperature will reduce the temperature difference between the extrusion and the bed, thus reducing the contraction force. 200 °C is a good temperature for getting PLA to stick. I've tried lower extrusion temperatures to reduce the temperature difference between the extruder and the bed, but this decreases adhesion to the bed and is counterproductive.
RAMPS 1.4 losing power with more than 3 drivers mounted
Your power supply may not be able to provide enough current. Check it's output voltage, the input at the RAMPS and the supply voltage (both VMOT and VDD) on each driver. If the supply is good, you can suspect the polyfuses on the RAMPS board. These are yellow rectangles, designed to protect against over-current. They age with stress, and typically fail to pass their full rated current (but you need to understand if there is a fault which damaged them). The link describes how to remove them.
Is it possible to attach a laser engraver to a SLA / resin 3D printer?
Not with consumer equipment. FDM printers are functionally three-axis motion control machines: "extrude quantity W material at position X Y Z, continue to next W, X, Y, and Z." Adapting it to control a laser position is a relatively simple task, because most of the logic is the same. "fire laser for W duration" instead of "extrude quantity W material". Moving the laser around is functionally similar to moving the print head around. Consumer SLA (aka "DLP" or "mSLA") is an entirely different logic, a different process. It is effectively layer-by-layer snapshots, rather than multi-axis motion control. They do a whole layer at a time, then lift the whole object up for the next layer: "flash layer 1 for W duration, lift bed Z, proceed to layer 2." Professional/industrial SLA printers that use lasers exist (like SLS), but they're tuned to precisely cure photopolymer resins, not cut things.
What are the tradeoffs of a cantilever printer design?
Cantilever printers constrain the X-axis carriage only on one side and have a free hanging side. This means, that the X-beam has to be rather sturdy and is limited in length, making them usually quite small machines. Another tradeoff stemming from this source is, that they have to stiffen this one connection as much as possible. Any play on the connection leads to increasingly large errors the further the printhead is out from the Z-Axis and can lead to bad prints, banding and total failures of prints. This behavior, at least on the TronXY-X1, is sometimes called Z-Wobble and has brought up some aftermarket modifications. For example, this bracket constrains the movement of the X-cantilever. While all axis would be affected by bad constraints, a cantilever suffers the most: The cantilever (if not well constrained) can deflect in both XY and XZ plane, pretty much opening a cone with the tip at the leadscrew nut. A one side-leadscrew portal design can only deflect in the XZ plane, opening an angle along with the center in the leadscrew nut. A double leadscrew design has, constrains both ends of the X-axis and prevents deflection but could suffer from bending the portal bar. The main benefit of a cantilever design is that being of a rather simple design with few moving parts, understanding as well as repairing and modifying the motion control system is fairly easy. The ease of access and low component number also allows reducing manufacturing costs to some degree. Another thing of note is, that these printers usually have a fairly small form factor, allowing them to be used in small workshops without taking up too much space. However, their build-volume to occupied-volume ratio usually is not as good as other designs - printers with a much larger build volume just get a much better ratio there.
Can K-type thermocouple disconnects be used with thermocouple based 3D printers?
Not sure if this is the reason, or just A reason, but thermocouples create a pretty small voltage for a given temperature, whereas thermistors have a larger (easier to read) change in resistance. I would expect that a thermocouple would require slightly more sensitive instrumentation to read it than a thermocouple. And seeing as space requirements aren't that significant on a 3D printer using the easier to read thermistor, despite its size, seems like the better option. I prefer thermocouples myself, since they don't need calibration and effectively lifetime guaranteed (aside from mechanical failure). However almost every project I've ever worked on, the Sparkies always seem to push for thermistors rather than thermocouples.
First layer height problem after changing nozzle
this is a common case in my TT.... Please check the y-carriage wheels - on the 2040 and 2020 profile. Mostly you shall find a little wobble on one of the ends, that introduces instability in leveling, so you could level it, home it and another level is needed. In my case, I decided to upgrade for dual-z drive - please see this Another interesting Z upgrade: option two We can observe z-wobble by watching the x-axis on the left side (no lead screw), it will have a little delay in response when moving up/down comparing to the other side.
Bizarre Underextrusion Issue
From what I can find, the SecKit Go has a direct drive extruder, so 0.15 is a very reasonable linear advance K value. For comparison, I have 0.6 for PLA with a bowden. Values in the range 1-2 are way too high for direct drive except possibly for really soft TPU. Geometrically localized underextrusion almost always means you have either loss of material due to oozing elsewhere, or filament slippage usually due to whatever's being printed just before the part with the underextrusion. If it's oozing you may need to increase retraction or disable or limit combing in your slicer (to prevent unretracted moves; these can ooze a lot of material when crossing infill areas). If it's slippage it could be from trying to print too fast.
20x20x20 calibration cube size change after switching filament
Filament type should have nothing to do with the issues you are facing. This is a mechanical issue or a slicing/scaling issue. The hotend should, if instructed to go to 20 mm height actually go to 20 mm (it did do that before!), it cannot "lose" 2 mm on the way up unless you have a lot of lost steps (e.g. too much load on the carriage pressing it down, but that seems pretty unlikely). Typical variations for Z are in the order of a few tenths of a mm for a properly dialed in printer. Please note that the Z axis is usually controlled by a leadscrew (your printer type and brand is not known, but if you have a Prusa style printer or some sort of a cube you'll have a leadscrew, Delta's have belts), so once you dialed this in, the head will go to the instructed height. Reasoning to your observations: If it was under-extrusion because of lower diameter filament, the last few layers should have been falling out of the hotend (considering the head goes to 20 mm and the print is 18 mm high). This is not the case; you have not described that kind of behavior. The most likely problem is a slicing or scaling problem. The answer to your question is therefore: "No you do not need to calibrate the Z height.". But when changing to another filament you could calibrate the extruder for this new filament; if the diameter is not far off you do not need to do that, but you could if you are into details. Preferably questions like these need to include a picture of the print and the printer type!
Why am I not seeing an effect from an M42 command on Marlin?
I looked at the current Marlin code and the P24 command should work as you expect it unless the pin you are trying to use in listed as the "SENSITIVE_PINS" list: #define SENSITIVE_PINS { 0, 1, \ X_STEP_PIN, X_DIR_PIN, X_ENABLE_PIN, X_MIN_PIN, X_MAX_PIN, \ Y_STEP_PIN, Y_DIR_PIN, Y_ENABLE_PIN, Y_MIN_PIN, Y_MAX_PIN, \ Z_STEP_PIN, Z_DIR_PIN, Z_ENABLE_PIN, Z_MIN_PIN, Z_MAX_PIN, Z_MIN_PROBE_PIN, \ PS_ON_PIN, HEATER_BED_PIN, FAN_PIN, FAN1_PIN, FAN2_PIN, CONTROLLER_FAN_PIN, \ _E0_PINS _E1_PINS _E2_PINS _E3_PINS _E4_PINS BED_PINS \ _H0_PINS _H1_PINS _H2_PINS _H3_PINS _H4_PINS \ _X2_PINS _Y2_PINS _Z2_PINS \ X_MS1_PIN, X_MS2_PIN, Y_MS1_PIN, Y_MS2_PIN, Z_MS1_PIN, Z_MS2_PIN \ } These pins are printer specific; so, without access to your Marlin build, I can't see if pin 4 corresponds to one of these. If this is the problem, the command should be returning an error. If there is no error, I would look closely at the hardware.
Help with a Tool Change Script
As far as I can recognize, the g-code file cannot be evaluated by printer firmware. It is expected that slicer software will be aware of tool head selection and process tool changes according to g-code provided in an on-tool change in most slicers. The best way to think about g-code (in RepRap word) is a one-way communication as the file is not treated as a script that can read and evaluate parameters on the go. (firmware capabilities). references: RepRap G-code list Printer quick start guide
Should I consider health impact of ABS or PLA when printing cookie cutter?
If it's one-time-use, both ABS and PLA are perfectly safe for use as a cookie cutter. The "food safety" of 3D printed parts is fairly controversial. In fact, whether any particular material is approved by regulators (such as the US FDA) for food contact is much more complex than most people realize. Materials can be accepted for some uses and not others. But there are two major considerations: Does the material leach into the food, or does the food leach into the material? Can the material be adequately cleaned and sterilized between repeated food exposures? Whether leaching is a concern will depend on the kind of contact. For example, nylon will absorb acids but is often used for potable water service. ABS is fine for cold foods but not hot foods. PLA (injection grade) is often used for disposable plastic forks and cold beverage cups, but rarely reusable containers. There is some anecdotal evidence that lots of 3D printing filaments, particularly cheap Asian filaments, contain toxic chemicals in the pigments and additives. Heavy metals like lead, cadmium, strontium, and all sorts of nasty organic chemicals have been found. You probably WON'T get any meaningful leaching of these chemicals in the brief exposure with cookie dough, but to be safe I would only use reputable US and EU filament manufacturers for food contact. I would also stick to filament colors that use non-toxic pigments (eg plain white is almost always safe titanium dioxide) or no pigments at all ("natural" color). When it comes to cleaning a cookie cutter for multiple uses, ABS is probably a much better choice than PLA, because ABS can withstand fairly hot water without losing shape. PLA probably won't survive a cycle through the dishwasher. PETG is perhaps a better choice than ABS for longer contact with wet foods, but may or may not survive the dishwasher. However, the porous structure of 3D printed parts is a serious problem for cleanliness. The tiny grain structure and voids between extruded strands can act as a safe haven for bacteria between uses. This is particularly an issue with cookie dough containing raw egg. A porous surface for raw-food handling would not be acceptable in a commercial kitchen. Now, the odds of actually transmitting salmonella via poorly-cleaned cookie cutter may be fairly low. Lots of people cut meat at home on wooden cutting boards (which are also porous) and never get sick. Properly cooking the food immediately after handling will go a long way towards reducing the odds of harmful cross-contamination. You'll have to decide for yourself whether the risk profile is worth it. One-time use is the conservative choice.
Is there any opensource software that controls the RAMPS RepRap board directly from the computer?
No, there exists no software like that. You can't communicate with a RAMPs board because a RAMPs board has no logic built-in; it's just a dumb breakout board that connects the Arduino Mega to your printer's components (such as stepper drivers, MOSFETs for controlling heaters, endstops,...). If you wanted to connect a RAMPs board "directly" to your computer, you'd need a way for your computer to generate the pulses that the Arduino normally generates. You could potentially use a parallel port for this but you'd need many more outputs than a single parallel port can provide and you'd need to find a way to do the analog to digital conversion needed for the thermistors. It's just not very feasible (considering how few computers have parallel ports nowadays).
Setting Junction Deviation in firmware has no effect
I've bought an SKR Pro V1.1 which is suffering from the same problem. It has actually to do with a shortcoming or design flaw (of the "HAL" or "Hardware Abstraction Layer" addressing of SPI devices) of Marlin and access to the SD card; I quote: The problem: On board SD-Card on SKR-PRO 1.1 can't be used. After the SD card has been inserted, the firmware binary is loaded into memory, and all access to the SD card is gone. The initial connection is by hardware design of the SKR boards. Basically, there is a bug in the Marlin code/SKR boards preventing storing (to the SD card) any alterations after the firmware is loaded. There is a PR on Marlin that fixes it, but it has not been merged into Marlin yet. There are 2 options, Every change to the firmware has to be compiled and uploaded to the board by placing a hex (firmware.bin) on the SD card; Compile a firmware version from the sources from the fore mentioned PR.
Monoprice Select IIIP Plus - Best speed settings and speeding up prints
In my experience a print speed of 50-70mm/s is ideal. Even if you set the speed to 150mm/s the print head still changes directions often and rarely will have enough time to accelerate from 0->150 before changing direction again. Some more effective ways of speeding up prints is to adjust Layer height Infill percentage (15-25% for regular prints, more if they need to be more sound) Supports Number of shells, etc
Temperature settings not changing in Cura
You're right in that the profile overrides the material. Cura's overriding precedence is as follows, where higher numbers override the lower numbers: Cura's base defaults. Printer specific settings. Adjustments to the printer made in the Machine Settings dialogue. Nozzle profiles. Material profiles. Quality levels (aka built-in profiles). Custom profiles. Settings not yet saved to a profile. In the Cura GUI, once a setting has been saved to a profile, there is no way to really get it out again. You can only create a new profile and copy over all of your overrides. However for the temperature you can still click on the little fx button next to the setting. This will restore the link from the setting to inherit again from different settings. In the case of printing temperature, the material profile actually adjusts the "Default Printing Temperature" setting. Clicking the fx button makes it go back to "use the Default Printing Temperature" so that your material's temperature is used again. The same does not go for fan speed, sadly. The Cura developers are planning to remove the fx button and replace it with a button to remove a setting override from a custom profile. That would fix your problem. It's not implemented yet though. Source: Am Cura developer.
Klipper BED_SCREWS_ADJUST and SKEW_PROFILE interaction
The skew correction works in X/Y axis while bed screws helps you correct the ax axis. Since they don’t work in the same axis, I don’t see why they would interfere with each other.
Are there differences in filament consumption between ABS and PLA
I’ve never actually printed anything other than PLA, but... Searching the web for filament density, I found this article , among others, that indicate that PLA is nearly 20 % denser than ABS, on average (1.24 vs 1.04 g/cm3). This indicates that a 1 kg reel of ABS will have 20 % more volume (therefore length) than a 1 kg reel of PLA. This does not quite explain the difference you see as 300 g vs 220 g is 36 % but maybe your measurement is off or the particular PLA mix you’re using is slightly denser.
Under-extrusion after a certain amount of successful layers?
This is partial answer and comment as it was too big to fit the comment section, it will be a proper answer once the question is updated by the OP. What you call a weird spiral pattern is the result of under-extrusion. When there is under-extrusion, the resulting print is sparsely filled. The reason for this under-extrusion is most probably partial clogging of the nozzle, or an issue with the extruder itself. One explanation could be that you are suffering from heat creep, but the information you provide is not enough to explain this or any other solution further in detail. Clogging can e.g. be caused by retraction settings and/or temperature settings, without your settings it is a guess what this is actually causing this. Looking at the print deteriorate the higher the Z becomes this is most probably a heat creep issue, so you print too hot with too large retraction settings.
Extrusion is not continuous
and welcome to 3D Printing Stack Exchange. From the picture and the video, it appears that you have significant under extrusion. It is impossible to determine from this information what is causing this. It could be temperature, such as a nozzle that is too cold, slicing configuration being wrong, such as by slicing for 3mm filament but using 1.75mm filament, electronics, such as the extruder stepper not generating enough torque, or a jam where the filament simply can't be pressed into the hot zone. I listened to your video but didn't hear anything that might help straighten this out. But, what I didn't hear was a "clicking" extruder. The extruder "clicks" if it is trying really hard to push the filament but it is not able to come out of the nozzle. It might be worth checking the toothed roller that drives the filament to be sure the teeth are clean. If the teeth are full, the drive roller can't get a grip on the filament to push it forward. Also, I am concerned that plastic is may be coming out somewhere other than the top of the nozzle. That shouldn't happen. Could you attach a photo of plastic coming out of the wrong place? --- Added after comments, and this photo os the nozzle leak was attached. 3D printing needs a long chain of conditions to be right. Any problem in that chain can cause unacceptable results. When trying to get good results, it is always best to fix any problem you find, and the plastic leak is a big problem. Other problems may exist, but this must be fixed for a print to succeed. The plastic leak is a big problem for a few reasons: It allows plastic to leave the melt zone other than through the nozzle. Since 3D printing depends on controlling the plastic leaving the nozzle for the print, this extra escape path is plastic which should have been neatly deposited on your print. The control system has no way to know that the plastic it fed through the extruder is not on the print, so you get under extrusion. There is an extra chamber inside the hot end which must be melted. In the ideal case, the filament directly enters the nozzle, in which it melts. Pressure from the cold filament at one end pushes the molten filament toward and through the nozzle. The filament doesn't not need to change directions or flow particularly easily because it always moves linearly through the hot end. It necks-down to the output hole size, but that is assisted by an entry cone to the nozzle aperature. In a working hot end, there is not hot filament inside the hot end that is not extruded. There are no eddy currents. There are no places where filament can rest. PLA is prone to degradation when kept at printing temperatures. This isn't a problem in an ideal hot end because no filament sits idle. Where there is an opening, or any extra gap, even if filament weren't leaking past the threads, there would be a pool of filament that was not in the extrusion path which would be hot for tool long, and would degrade into a hard, burned mess. Disassembling the hot end can be tricky. The cold plastic is a great glue which has filled and sealed the threads. Do not use the internal heater to heat it. The wires will get in the way when you are taking it apart. There can (will) be broken connections, and possibly uncontrolled heating. Use a hot air gun or hair dryer to soften the plastic. If you can not remove the heater and thermistor before taking it apart, find where they are connected to the electronics and disconnect them there. Do not leave the wires connected. The heater and thermistor are fragile, and too much bending or force on the wires will destroy them. You might purchase spares because you are likely to damage one or both. PLA softens at a relatively low temperature. When you have a plastic encased mess, disconnected from the rest of the printer, you can soften the plastic either with hot air or hot water. A "relatively low tenperature" can still give you serious burns. Use gloves, pliers, toothpicks, and any other tools you need. When you reassemble the hot end, the nozzle should screw tightly against the upper tube that delivers the plastic. Inspect the top of the nozzle and the bottom of the tube, and be sure they are flat, smooth, and are cut at a 90 degree angle to their axis. They must fit together tightly, leaving no gap. One assembly order is recommended by E3D. I mention them because they deserve credit, although I am embellishing with details they are unaware of, have not suggested, and are not responsible for. Start with all parts removed from the heater block (the rectangle of aluminum). Screw the nozzle into the heater block, and then back it off 1/4 to 1/2 a turn. Screw the top tube into the heater block until it is tight against the nozzle. Assemble eerything else -- heater, thermister, wiring, attach to extruder, ... With no plastic, heat the hot end from the control panel to be hotter than the hottest you will need when printing, maybe 250C if you are only printing PLA and ABS, and assure that it is heating normally. WITH THE HOT END HOT, tighten the nozzle with a wrench so that it is tight against the upper tube. Of the three objects in the hot end, the aluminum heater block has the highest coefficient of thermal expansion. It is higher than both the steel upper tube and the brass nozzle. The tendency as the hot end heats is to open a gap between the nozzle and the tube. By making them tight at the highest temperature, they will also be tight at lower temperatures. When the leak is fixed, you may be done, or there may be other problems to solve. Welcome to 3D Printing!
Thermistor reading 190 °C constantly
Yes, in newest Marlin (v2.0) there are configurable thermistor parameters. You need to set custom thermistor: #define TEMP_SENSOR_0 1000 for your only extruder (suffix _0) in Configuration.h, and in Configuration_adv.h modify related section: // // Custom Thermistor 1000 parameters // #if TEMP_SENSOR_0 == 1000 #define HOTEND0_PULLUP_RESISTOR_OHMS 4700 // Pullup resistor #define HOTEND0_RESISTANCE_25C_OHMS 100000 // Resistance at 25C #define HOTEND0_BETA 3950 // Beta value #endif But in Marlin 1.x I couldn't find this section, so change would need to be done somewhere in thermistor definition source files. I mention this older Marlin following the discussion on Wanhao Duplicator i3 Mini & Rebrands Support, which refers to some Wanhao source code, but I cannot judge about value of it. For resistors in the circuit I have similar observations: sometimes they do not give exact readings using multimeter, especially next to capacitor. Isn't heatbed temp probe's circuit the same? - can you compare? But instead... I would suggest doing following things: open circuit test: disconnect the thermistor and check readings (I could then see -19 °C) closed circuit test: connect the heatbed's thermistor instead (in place of hotend's) and check if these readings are reliable (I would try to heat the bed for few seconds to check if changes are noticed, but not too long because it will be uncontrolled process) thermistor test: connecting the hotend's thermistor as heatbed's probe and check on-screen readings for heated bed; are they again 193 °C? do they react to changes? (powering hotend for few secs or warming in fingers?) you may consider to setup USB connection with serial terminal to directly read measurements using G-Code, just to exclude errors in post-processing of the temperature readings before putting it to screen (I know, it's a bit paranoic) The worse thing I can imagine is that you could short wires of hotend heating circuit with the temperature probe circuit, which should be galvanically isolated. Then it could damage readings of the mainboard for this probe when received 12V/24V. I found that the mainboard is most probably all-in-one i3mini 0ne+ motherboard, some revision still available in Wanhao store for another $40. I googled for few photos, there are problably few versions of this board. If you are lucky and skilled then you may find some other pin (connected to ATMega) to reuse. Otherwise you would need to replace steering electronics with new board, which could be of great benefit for future improvements: more ports for extra fans, probes, more memory, better stepper drivers, etc.
Gluing paper to PLA objects
The problem is probably the UHU glue having not had the time to cure through and interacting with either the PLA, the moisture in the garage, or the brick-paper's covering foil. Many of the metal-tube UHU glues - like "UHU Hart" - are resin-based and contain solvents that can interact with some polymers like the coating of the brick paper. Curing times can be up to 48 hours to cure. I have had a very good experience with very basic water-resistant Wood Glue to both smooth over 3D-prints as well as attach layers of wood or paper to it. Most wood glues are made with a chemical that is the same or similar enough to those in Gluestick (usually PVA, sometimes PVP). These usually don't react with the fibers and surface of the NOCH-brand brick paper I know. After having the glue dry thoroughly - for Express type wood glues that is about 2 hours - there is little to no chance that the two interact badly in the garage.
Why is PEEK filament so expensive?
My assumptions about PEEK filament price are: Raw material is more expensive. Compare price of ABS with PEEK pellets. Demand is much lower. There are not many printers able to print peek. If you manufacture PEEK filament you have to store a filament batch for longer time. Manufacturer has to calculate into price storage space, material degradation, ... Filament machine tuning. You have to tune filament extrude machine for PEEK, which takes time because it's a totally different plastic. Maybe there is a cleanup needed after finishing a batch and switching to another material. Working conditions. PEEK is quite smelly and I am not sure if you have to improve work conditions like better ventilation. Research costs. You have to distribute research costs to a filament production where demand is low.
Ender 3 stock Bowden tube inner diameter? Worth upgrading?
I disassembled the hotend side of the Bowden setup to check for deterioration (and found an actual hole in the tube!) for the first time since getting a decent caliper, and after making a clean cut to remove the damaged part, measured the inner diameter at 2.19-2.21 mm. I'm not sure if that's enlarged from wear or the original size it was manufactured as, but either way this suggests that upgrade to Capricorn at 1.95 mm (±0.05 mm) inner diameter should make a significant difference vs my current Bowden tube.
Uploading files using Octoprint REST API
I installed octoprint locally to help debug this problem. Your code snippet says "def def Upload_File" which is a syntax error. If you go into Octoprint, under Settings -> Logging, and download octoprint.log, you'll notice it says "WARNING - 400 POST /api/files/local (::1): No multipart boundary supplied". A quick google search led me to this StackOverflow question, which states to remove the custom content-type header to fix this problem. You actually need to open() the file, and assign that value to the 'file' field in the fle object definition, not just tell it where it's currently at on your machine. Additionally, you need to provide a filename field in the fle object. I got your code snippet working as below, went ahead and swapped your API key back into it. You'll need to point the open() call to the actual location of your file on your hard drive, which I assume is j:/max.gcode, rather than mine which is just opening max.gcode from the running directory of the script. import requests def Upload_File(): fle={'file': open('max.gcode', 'rb'), 'filename': 'max.gcode'} url='http://localhost:5000/api/files/{}'.format('local') payload={'select': 'true','print': 'false' } header={'X-Api-Key': 'FD550BD4DA2442BA906AD1850539D6DB' } response = requests.post(url, files=fle,data=payload,headers=header) print(response) if __name__=='__main__': Upload_File() Finally, if you're going to be doing a whole lot of REST API shenanigans with your Octoprint server, may I offer a prebuilt library?
What is the best way to print a model with support material in first layer?
I had the same problem printing a miniature just recently. As always, settings are somewhat dependent from the object you want to print, but here are some suggestions: Increase the support density: 15% (8% is very low!) Support pattern: zig-zag with "connect zig-zag" option enabled (add stiffness to the "column" of support) Enable support interface (increase adhesion to the plate, and provide a more "beefy" base for the support material) For reference, here's a screenshot of my settings as I tweaked them for that miniature (printed a 0.1mm layer height). ("Support line distance 1mm" is the same as "Support density 15%", or at least it is the same with my nozzle size) If your problem was not only with the "flakiness" of the material, but also with it adhesion to the bed, then prepping your printing bed somehow (with a bit of painter tape, glue, etc... can help. Alternatively using a brim or a raft (as also suggested by another responder) could also help. EDIT: all the above still standing true, it turns out the OP had a hardware problem as well, his timing belt being loose (see comments).
'Sine wave' deviation in X-Axis on moving bed printer.
This is the issue of your z-axis rather than x- or y-axis because printing speed doesn't have any influence on the results. I bet: your threded rods are bent or nuts on these rods have too high clearance or nuts on these rods have eccentricity or couplings (between motor and threded rods) are bent or these couplings are badly fitted All above suggestions concern z-axis. [edit] because it's PrintBot its z-axis has the following disadvantage
LCD full graphic smart controller, no character display and screen blinking
this problem can be solved by turning the slots on the display, as in the image below. Some Chinese displays are inverted from factory.
Cura Parameter Confusion: Initial Layer Flow Rate
If you're over- or under-extruding by a 30% margin, there is going to be so much excess or deficiency of material in the first layer that it will take a few layers to "catch up". Normally you would only adjust the flow by 5% or at most 10% unless you're using it to correct for compression of the filament in the extruder gear (for flexible filaments), but adjusting first layer flow only is always a hack to work around incorrect bed height or adhesion problems. You'd do much better to figure out the right solution to whatever problem you're trying to solve with flow.
How to prevent z-fighting in OpenSCAD?
The general advice in the OpenSCAD community is to "extend your cuts and embed your joins" (source). The rendering artifacts are one thing but rather just an annoyance; however z-fighting can also cause unexpected errors during STL export (I did not experience that myself so far, just read this somewhere). So you would change the dimensions of your objects very slightly (0.01 mm works fine) so that: for a union(), there is overlap volume between the parts for a difference(), the intersector has volume both inside and outside of the intersected part Now you could adjust both the size and position of your parts to keep the mathematically exact dimensions for the resulting part. But I found that for the purposes of 3D printing, such accuracy is not worth it because it complicates the formulas so much. Instead, I adjust either position or size of a part, depending on what is simpler in each case. A measure in the final design will be off by 0.01 mm, which does not matter. And I keep the 0.01 mm offset in a variable called nothing (picked that up somewhere and liked it …). This keeps the calculations intuitively understandable. Example To create a cylinder and cut a hole to half of its depth, I would do this: //!OpenSCAD nothing=0.01; height=40; difference(){ cylinder(h=height, r=20, center=true); translate([0, 0, height/4 + nothing]) cylinder(h=height/2, r=15, center=true); } Now the hole is nothing=0.01 less deep than half of the cylinder – that's the inaccuracy I accept. (Note: If you don't have OpenSCAD installed, you can also try the above code online by copy & pasting it into OpenJSCAD. Include the magic comment in the first line to switch it to OpenSCAD syntax.)
Ender 3 print bed damage
your print surface is destroyed So, you managed to rip off your print surface in the center. Happened to me too. the corners of my scraper were too sharp, cutting the surface. Another time I did pierce the surface with my nozzle. Damage happens. Replacement surfaces for the Ender3 start at about 5 bucks a piece. So get yourself some spares. Clean your bed before applying the new one. Removal process To properly remove a print from the bed, grab your srcaper blade and do the following: sharpen the edge only on ONE side. soften the corners, they should be slightly rouned. Make sure the scraper is kept sharp. When removing prints, be slow! Push the scraper against the bed with the bevel upwards. Apply careful pressure till the tip moves just a little. Move to a different spot till the blade slips under a little, then work the scraper left and right. You'll hear a sucking sound, that's the print coming free. It gets much easier if you let the bed cool down. For very tall prints with a relatively small area, you might not need the scraper at all. PETG has a tendency to stick too much with glass and fuse with PEI. We don't know if the Build-Tak clone Creality uses contains PEI. I suggest to grab gluestick to add a safety layer for printing PETG.
DLP build plate adhesion
I have heard that adding a few pieces of masking tape to the print surface improves adhesion.
Print from SD and connect to serial host at the same time
This can be done, but you need to have the right order of operations. Octoprint relies on such a setup. Set up the serial connection first, as sending the connect signal from your terminal or Raspberry to Marlin triggers the printer to reboot. After having the connection established, start the printjob either via the control buttons or the remote console (for example: Octoprint or Repetier Server, running on your Raspberry) Note that you have to make sure not to dis- and reconnect the raspberry during print, as that might trigger a reboot of the printer!
What type filament should I use to produce stamps?
I see three options... 1. Print with a flexible filament: Many options: TPU as you pointed out, the flexible PLA that Tom mentioned, and others. Here's an article with a few options from Matter Hackers I'd suggest printing the stamp side down so you get a nice flat stamp with no post processing. For any wide gaps, mind your bridging...use fillets or chamfers so the "roof" of the gap is an upside down V or U shape. 2. Print with any hard filament, but use a rubber mat under the paper. The idea here is to use a semi-flexible surface under the paper to help get uniform contact pressure between the paper and the hard stamp. As long as your stamp holds ink, this should work okay. 3. Print your stamp shape, then use it to make a rubber stamp. Making a model and then copying it with your desired non-printable material is a common manufacturing technique that can be used in many situations. Use of molds for casting and related processes can really expand your possibilities. Check out Smooth On. They have several rubber and flexible compounds that you could use for stamps.
Wrong calculation of E-parameter values
The value you see is the cumulative extrusion, in other words the E value in consecutive G-code commands reflect a position of the filament since you started that print. It is normal that it continuously increases in "absolute" mode. You should see, at the beginning of the G-code, also a M82 or G90 command, which means "absolute extrusion". And you should avoid having M83 or G91 later in G-code, which would change an interpretation of E values to "relative extrusion". It is more accurate than relative extrusion (which is the mode you are thinking about). From your print it looks like the printing bed is not properly tuned, probably too far. Also, it is difficult to see properly, but it looks like you get bubbles in the second part of the print, which are the reason there is an increase of extrusion: hot air pushes too much filament out. Maybe your granules are not properly dry. It is difficult to tell you more, pellet extruders are more difficult to tune than filament extruders.
Lead screws - M8 instead of M5?
as long as you match the parts that is ok. The m8 rod will give you more stiffness and will be harder to bend. As the result, you will have to calibrate the steps/mm settings in the firmware.
Acetone Smoothing an ABS Cone - (New to 3D printing)
is it possible to smooth the surface of my print without loosing shape? That is by definition impossible. Smoothing the surface implies that the geometry is changed. Rather, the question is: how much does acetone smoothing change the shape, and is that change within the tolerances of your application? Acetone smoothing is a fickle process, and how (much) the geometry is changed depends on a lot of factors, such as the concentration of the acetone vapor, the amount of time the print is exposed to it, the exact composition of the plastic, geometry of the print, etc... It is definitely possible to overdo it and melt your print into a puddle, but it is also possible to smooth it out without affecting the geometry significantly. I would recommend experimenting with the process and seeing whether you can tune the process (time/concentration) so that you get a smooth enough part without affecting the geometry too much.
How to generate a STL from a rotational solid of two equations?
The Wolfram Alpha "Data Download" feature supports STL format, but is only enabled with a Pro Account. When I click your link to Wolfram, I see the visualization of your equations at the bottom of the page; when I hover over the image, several buttons show below it. One of them is "Data", with a download icon. This gives a number of download options. The File Format drop down includes "STL".
How to wire an LJ18A3-8-Z/BX sensor with an optocoupler to a stock Ender 3?
Both diagrams are exactly the same, the difference is the +5 V that is not connected on the bottom diagram (displayed in white, labelled "not used"), but since your endstops do not have a +5 V pin in the socket (only ground, GND, and signal, SIG or S) you don't need to attach it. You should source the +5 V from elsewhere from your board. As the left and right side of the module are separated circuits, you need to power the board with +5 V. Read this answer for applying 24 V.
Ender 3 v2 won't read or recognize any SD cards
Format for 32 with MBR for ender 3d I had similar issue. Check Reddit
PET-G grid print problem
You mentioned coasting. Absolutely make sure that's off. Coasting inherently extrudes less material than what's needed to print the model correctly, and this underextrusion will accumulate. PETG is really sensitive to underextrusion because it doesn't self-adhere well without a lot of effort to get things right (temperature, speed, etc.) and unless/until the amount extruded is enough to adhere well, it'll just get dragged by the nozzle. Also 220°C is way too low for PETG, and low temperatures will exacerbate this kind of problem. I'd consider 230°C the absolute minimum for PETG, and 240-250 preferable. Ignore what the filament manufacturer or vendor advertises about lower temperatures being okay; this is normally a marketing ploy to ensure they can sell to users with printers that don't recommend high temperatures or have hard limits in firmware and users who are worried about PTFE off-gassing when operating at higher temperatures (not a real issue at or below 250°C, and probably not even up to 260-275, but still common FUD). Even if they don't want to be deceptive they're competing with other vendors who do. Further, make sure fan is off, or at least set to very low speed. Rapid cooling will completely mess up PETG adhesion. PETG is still very viscous at printing temperatures (unlike PLA) so there's no need to start cooling it right away to ensure it keeps its form. Finally, PETG is just a real pain to print with. At first I thought it was a nice material, because it holds shape well and can stand up to fairly high temperatures (which are still true), but there are lots of gotchas and it's really hard to tune your settings to get good adhesion and strong parts. Be prepared to do a lot of experimentation beyond the things I've written in this answer to get good results if you want to keep using it.
Make smooth curves in GCode
Curves are normally approximated with straight-line segments. However, some firmwares (e.g. Marlin) support arc commands. See: Marlin: Controlled Arc Move RepRap: Controlled Arc Move Note that the RepRap Wiki is not always accurate in its list of supported commands.
What are ways to avoid heat creep?
Things to consider (This is expanded from How is heat creep characterized?): The air volume of the cooling fan on the hotend heat sink affects the temperature gradient across the heat sink. Usually the size/geometry of the fan depends on the printer design, so the main parameter of a fan that controls air volume is the rpm. I've noticed fans fitting hotends from 6,000 to 10,000 rpm. The higher rpm the better as far as preventing heat creep. Of course, make sure the fan is spinning properly. The raising the temperature of the extruder increases the chance of heat creep. A high resistive connection to the sensor will cause a lower that actual temperature reading. Thus, the controller will raise the hotend to a higher temperature than set, which can cause het creep. Razing the temperature of the heated bed, especially when the printer is in an enclosure, increases the chance of heat creep because this raises the temperature of the air the fan blows across the heatsink. What can be useful is using higher extrusion and bed temperatures for the first layer, then set lower temperatures on subsequent layers to reduce the chances of heat creep. The slower the print speed the higher the chance of heat creep. That's because a slower print speed gives the filament more time to heat up in the extruder. Many slicers have a setting that slows down the print speed if a layer will finish below a given time. If this slowed down time causes heat creep, the extrusion will stop when the printer reaches these smaller area layers. Too much travel of the extruder without extruding has a similar effect because the filament isn't moving. I verified that connecting isolated structures on a design with heat creep can eliminate it. Using a little more material for thin walls is better than scraping prints. Example: This is easy to demonstrate with PETG. Design a table printed top down 100 mm square with four legs 5 mm square and 100 mm long. Set the parameters so that you start to get heat creep when printing the legs. Now put four 1 mm thick walls between the legs to get the heat creep to go away. Sometimes if you start a second print without letting the extruder cool down from the first print, heat creep will cause the extruder to jam. The heater block might be too close to the heatsink so that they bypass the heat breaks reduction in conducting heat. Hotends for temperatures 245 °C and below can use such things as Teflon tubing in the heat break to isolate the filament from the metal. But, +250 °C nozzle temperatures need all metal hotends. The PID settings in the firmware might be letting the hotend temperature overshoot too high.
CoreXY with 8 bit board problems
You can use an 8-bit controller board for a CoreXY kinematics 3D printer. The calculations are not so complicated opposed to those for a Delta. My Hypercube Evolution uses a RUMBA controller board that hasn't failed me past years.
Laybrick - Gaps on top layer
When using Simplify3D, you may try referring to their awesome troubleshooting guide: https://www.simplify3d.com/support/print-quality-troubleshooting/#holes-and-gaps-in-the-top-layers 3 reasons (from the guide) for gaps in the top layers: under-extrusion: you may try to increase extrusion multiplier to see if this helps low infill percentage: not likely in your case if increasing number of top layers did not help not enough top layers: you already tried increasing number of top layers. In addition to that, you may change certain parameters (such as extrusion multiplier) for a given number of layers by using Simplify3D features. It can be helpful to avoid problems to for the rest of the model.
Platform support up to a certain Z height in Ultimaker Cura/G-code
Is there a software that will enable this? Yes, as of Ultimaker Cura 3.3 Beta, Ultimaker Cura allows you to specify an area which will not be considered for adding supports. In your case you could define everything above 4 mm to be excluded from building support structures. You can look here for this very new feature, it might be what you're looking for.
Applying Kapton Tape To Bed Plate
Try using a solution made for applying window tint, maybe even soapy water. Squeegee it out, heat the bed for a bit and you should be good to go. This video demonstrates the process.
Complete list of fasteners for the Sintron Kossel
I found two lists for the Kossel mini (the documents you referenced are for a mini, so that is what I am assuming you got). One as a Google Doc and one as an order sheet Both seem to list all the fasteners. Here is a Sintron mini carriage assembly and it lists the hardware included I found an ebay listing for the Sintron 2020 Rails. The recommended screws are M3x20mm and here is the infor on the required t-nuts. Happy building.
Filament doesn't retract for some travel movements using PrusaSlicer
Minimum Travel after Retraction is exactly what it says on the tin: if the travel after a retraction would be less than 2 mm, it does not retract. You'll want that to be short, but not 0, because retraction can lead to under extrusion at the start of a new line, and every swap from one shell to the next shell right next to it is classed as travel.
Which material should I use to print a mounting adapter for the y-carriage?
Any thermoplastic is a bad choice to carry a heated bed. For an unheated bed, monolithic (cast, not printed) acrylic plate of sufficient thickness can be used. To get the needed stiffness and temperature resistance, steel plate of about 2+mm is the best choice, followed by aluminium of a higher thickness. The latter can be sawed with woodworking equipment, if done slowly.
How do I get a replacement PTFE tube to fit inside my nozzle?
In the end, I was able to get the tube in the nozzle. It took a lot of patience, and it seems to help if the nozzle is warm as well.
Monoprice printer: no files found on SD card
I've no idea why 3D printers don't like SD cards formatted by operating system utilities, but they don't. Use SD Memory Card Formatter to format your memory cards. Unfortunately, you have probably lost any data that was on the original SD card. The SD cards that are supplied with 3D printers typically contain documentation, test files, and device drivers. However, these files do not need to be present for the correct operation of the printer. All that is needed is a correctly-formatted SD card containing the G-code files that you want to print.
3D Printed Lead Screw
Expanding on some previous comments which are probably enough to warrant an answer: What Trish said is completely right. Leadscrews are readily available parts and any dimensional errors in the leadscrews will be reflected in the output of your CNC machine unless you have some sort of compensation for them. Moreover, if the material is not highly rigid, the dimensions are subject to change over time, so any compensation would have to be ongoing manual adjustment or closed-loop rather than a one-time calibration. "PLA+" is an especially bad choice because it usually means PLA that's been modified with additives to make it less brittle, deforming under stress instead of holding its shape until it breaks catastrophically. CNC Kitchen's video on PLA+ elaborates on this. With that said, if you don't need a high level of precision, or if you're in a situation where you're unable to obtain manufactured components, I think 3D printed leadscrews would work ok if you print them in the XY plane rather than along the Z axis. While nozzle width and discrete layers produce a sort of "stairstep" quantization of printed threads in the Z direction, that doesn't happen with the threads in the XY plane; the nozzle width limits feature resolution (oscillations per unit length) but the positioning of the threads is quantized only to the X and Y (micro)step size, which is typically on the order of 10 microns. Moreover, the strength and rigidity of the part printed in this direction can be very high, due to the offset-layered zigzag structure. Back to accuracy of the part, though, it's important to note that whatever flaws your printer might have in XY positioning accuracy will be reflected in the resulting leadscrew. This includes non-linear effects such as belt paths being slightly trapezoidal instead of having perfectly colinear points of attachment to the carriage. In general, when manufacturing parts that will affect the accuracy of the resulting machine, you want to use processes that amplify the precision your tooling was manufactured with rather than processes that reproduce or amplify its flaws.
Question about retraction distance and stringing/print quality with PETG
You cannot endlessly increase the retraction distance, doing so leads to different problems as you encountered. As a rule of thumb, the retraction distance should not exceed the length of your nozzle. Depending on the type of extruder, many printers use a value between 2 and 7 mm (e.g. the Ultimaker Cura retraction length is 6.5 mm at 25 mm/s, this is for a Bowden style extruder). In fighting stringing (in which PETG is very subjected to happening) retraction length is only one of the parameters that you need to adjust to optimize the printing process. Other important parameters (besides retraction length) for stringing are: retraction speed travel speed minimum travel distance Furthermore, the type of extruder (direct or Bowden) is also important and other parameters that may reduce the stringing are: coasting (stop extruding before the end of the print line is reached using the residual pressure in the nozzle to extrude the last filament) pre-loading/priming filament after a move Usually it suffices to tune the first 4 parameters. As you see in your results, the retraction of 5 mm is close to the result you would expect, increasing it, may lead to clogs and in your case leads to a sort of stringing. This latter result is probably due to the fact that the filament does not reach the print in time (you definitely see under-extruded cylinder columns), while leftover pressure after printing the column layer oozes some of the filament creating strings. Just play with the retraction length in steps of 0.25 mm (around the 5 mm) and you may try to change the speed with 5 mm/s increments if necessary.
Are control boards switchable?
Yes. Stepper motors, heaters, temperature sensors, fans, etc... are all standard parts. Pretty much any control board can be used with the hardware from pretty much every printer. Some more expensive printers might use more exotic parts (especially temperature sensors), but on the budget printers you're looking at I would be very surprised to find anything non-standard.
How can I optimize my print for casting in metal?
Print a two part negative (mold) of your objects. http://www.thingiverse.com/thing:31581 https://pinshape.com/blog/how-to-generate-a-3d-printed-mold-for-an-object/ Melt and pour wax into the mold. Praffin wax melts at only 37C, not an issue for both PLA and ABS. Use the wax object for casting, not ABS/PLA/etc. The mold is also reusable this way. To prevent the wax from sticking to the mold, something might be applied to the surface. Oil maybe?
XYZPrinting Premium Metallic PLA Conductivity
Most PLAs are non-conductive: they contain little to no metal in comparison to plastic and can't make a conductive path. Even metallic PLA does not contain enough metal to be conductive. There are some rare filaments that are conductive, they are sold as conductive PLA. I have yet to encounter a non-black conductive PLA. One filament I suspect to be possibly conductive but have no hard evidence is made by The Vitual Foundry and contains about 80% metal. It is intended to be treated in the oven to burn out the plastic part, leaving behind a somewhat pure metal structure, but might be conductive without this process already. I have no experience with this and all my exposure to it was this answer. It should be highly abrasive though.
Unable to Determine Proper Print Scale of STL model
The STL format does not define unit information. So there is no way of knowing, from a STL file alone, what size it should be. However there are only a handful of units people design with. So the unit is most likely one of inches, cm or mm. So if you inverse convert between those combinations you'll likely find the real size. Your image is somewhat inconclusive. But it seems to indicate your object is in fact larger than your bed. As it it is 18-19 pips of height and i would interpret each as 10 mm from the image. Which would indicate your model is correct. Split and glue? Print corner to corner you should have 207 mm length that way (but not necessarily width).
How to fix transparent outer face?
You've not offered the source of the model, nor the software from which it was created, but I suspect that you'll find the "invisible" faces have reversed normals. This is a characteristic of modeling software that believes the outside is the inside and vice versa. If you are comfortable with using another program, you can check/confirm those problematic sections. I'm fond of Meshmixer, a multi-platform 3D modeling and editing program. It will display reversed normals with a bit of zebra striping. One can use the Select tool, highlight those faces and use the Select Edit, reverse normals menu entry to repair the problem. Some slicers will automatically repair this type of problem. For a different approach, download and import your model into Slic3r, which will indicate that it has corrected xx errors.
Stuck getting ANet A8 Printer to build layers
Your nozzle is way too far from the bed for the first layer. You should adjust the bed leveling screws to bring the nozzle closer to the bed, to the point where a piece of paper just barely fits between the nozzle and the bed. You should closely watch the first layer as it is being put down, the plastic should get squished slightly and there should be no gaps between adjacent lines.
My DIY printer only boots when all my endstops are triggered. Why is that?
From your fault description of printer logic board losing power when endstops are released it sounds to me like those endstops are shorting VCC to GND when not pressed. That the fault also goes away when the endstops are unplugged reinforces this. I guess the logic board has its own protected regulator because otherwise you would be shorting out the power supply and that would also turn itself off after likely causing a lot of damage to the logic board.
PLA wood filament: a special treatment?
PLA and wood fibres = wood filament Most wood filamet consists of about 60-70 % PLA and 40-30 % wood fibres. This basically implies that PLA temperatures should be used. It can be printed with standard 0.4 mm nozzles, but it is adviced to be printed with a larger diameter nozzle. A larger nozzle will less likely to cause nozzles to clog (more area for the fibres to pass through). Basic printing advice Start experimenting with relatively large layer heights (0.25-0.3 mm layer heights). Printing speeds should be held high to ensure relative short residence times in the hot end (in the range 50-80 mm/s). Short residence times prevent degradation by heat resulting in clogging. It is best for wood fibres filled filaments to not have the filament to stagnate (e.g. pausing). A higher retraction speed and distance might be needed because the filament usually is a little more runny compared to plain PLA. Beware After printing, you should retract the filament from the hotend so that during next printing jobs heat up cycle, the filament doesn't degrade and clogs the nozzle. Alter the end script G-code and also use priming scripts at the start and/or do not forget to use a large skirt of multiple lines or a considerable length. As always, you shouldn't leave the printer unattended, regularly check the printer and shut down the printer when there is something wrong, e.g. not extruding. Also note that wood filled filament is reported to be abrasive to brass nozzles, a (hardened) steel nozzle or a Ruby nozzle should mitigate wear problems. Basic settings These settings have proven to work, but are not a guarantee, these settings should provide a reference to start experimenting on your own printer: Temperature: 190-205 °C Layerheight: 0.25 mm Speed: 50 mm/s Minimal layertime: 8 s Bowden type extruder Retraction speed: 40 mm/s Retraction distance: 5.5 mm Direct extruder Retraction speed: 25 mm/s Retraction distance: 1.3 mm
Photon S DLP first layers adhere but support will not
I finally reached out to anycubic and their tech support answered and was awesome! They had me update the firmware, tighten the fep film more ( instead of using the small lever I used the big lever this time), then level the bed again. I printed their test print and it worked! After that I looked at their parameters and I saw that mine were very different. The main differences was in the lift distance, speed and retract speed. I only had 1 mm lift distance and much slower. Here are the correct ones in the end
Minimum material deposit threshold in slic3r
It does this in small corners. For example, this model, when sliced has this on layer ~7: Tiny tiny bit of blue infill. There seems to be no feature to explicitely disable this. But sometimes an additional perimeter solves it.
Is it safe to use a 3D printed icing smoother?
As mentioned, FDM 3D printed parts are flawed for food service and prep use regardless of the material, because of the small gaps between lines where bacteria can grow, and because your printer isn't used in a way to keep it from introducing contaminants into otherwise clean material. However, there are some ways around this. Single-use I'm pretty comfortable using my printer to make items for one time use. If I wanted an icing smoother with a fancy shape for a special cake that I'm not likely to need again soon, I'd go ahead and do that. I would apply all the icing at one time, and then I'd discard the piece, rather than try to clean and save it. Also note that I'm not talking about a commercial kitchen; this would be for a cake I'd eat myself with friends, rather than sell. The big thing I've done this way so far (I've had my printer less than a year) is make shaped cookie cutters. I'll print the cutters, use them, and then throw them away. If I want the same shape again some time, I'll re-print. Lining The other thing you can do with food prep items is print them with the intent to use liners. For example, here is a 3D-printable taco train, where a train car has grooves to hold tacos (yum!). It wouldn't be good to put a taco directly in here, but you could use napkins or similar food-safe liner to separate the 3D part from the food. In the case of the icing smoother, you might be able to print the part and wrap it in aluminum foil.
SLA printers: safety with resin contaminated disposals
The IPA with the resin in it is a chemical waste and toxic to nature. As such, it needs to be given to a chemical waste handler. But those handlers do charge by volume, so you need to find a way to handle volumes reasonably. On the one hand, this means to not use huge volumes for washing the prints but reasonably small batches, and on the other hand to try to saturate them as much as possible before giving them to disposal. It is not safe to leave the IPA in the sun or close to a source of flame, and it would certainly be not a good idea to try to torch the IPA to try to reduce the volume, but you might manage to try to concentrate the toxic waste if you might have access to a vacuum distillation apparatus - the IPA would be able to be distilled over to be reused while the original material would end concentrated. At the moment I am waiting for an MSDS of eSun's Water-Washable resin, but the Eleegoo one is available, and it reads that it is classed as "Aquatic Chronic Class 2" - that's better than the Class 4 which IPA-washing resins have, but it is still not Sewer-safe: Do not allow product to reach sewage system or any water source. However, water is much less of a dangerous base for the waste than IPA, making it easier to handle. This mixture too needs to go to a chemical waste handler or be made inert but the waste also can be concentrated with less danger than IPA using a similar apparatus (cold distilling out water) or even a dedicated waste cookpot/evaporation vat that is exposed to heat - though you clearly should do this in a chemical-grade air filtration unit to mitigate the chance of resin fumes escaping into the atmosphere or your work environment if you heat it more than some. The dissolved resin might not be able to cure anymore as the photoinitiator might be used up during it evaporating.
Print quality of Kossel clone for Print in Place models
Delta printers are considered to be able to be accurate printers cause of the limited weight in the head (using Bowden extruder setup). The positioning can be very accurate (limited weight, limited overshooting) and because of the limited amount of weight, the print speed can be increased. An interesting paper has been written on a comparative study between a Cartesian and a Delta machine. The paper concludes that the Delta style printer produces "a better surface finish". However, I'm a bit skeptical as the images of the Cartesian printed part they have printed are of far less quality I can produce with 2 of my printers. Fact is that Delta machines have no Z-wobble (also called banding) that is a common problem with Prusa i3 style printers for instance.
Still getting "no layers detected" after repairing model in PrusaSlicer console
Still don't have an actual answer as to why repairing the stl file didn't seem to work, but since my stl file was a combination of different stl files I tried repairing the individual stl files before combining them, which resolved the problem.
How does a layman get a 3D printed replacement part?
If you have the remaining pieces of the gear and enough remains to determine certain measurements, one can either engineer the gear using a number of gear modeling designs, or one can take measurements directly from the parts and engineer a raw design. If the gear you have is not particularly peculiar, it is possible to use a gear generator plug-in, template, or library to make the "foundation" of your gear. The modeling software would then be used to add the appropriate bosses and key ways required to complete the design. If you are considering to create the part yourself, you have a wide selection of programs from which to choose. I'm fond of OpenSCAD, and it does have a number of gear libraries. Simple bosses and key ways are easily accomplished using OpenSCAD. Another package available on the internet which includes the option of using a gear generator is Tinkercad which has a reputation of being easy to use. You'll find many tutorials for this program as well. Tinkercad requires an "outside" program to generate the gear design which is then imported to the model workspace. Even a program as simple as Inkscape can create gear profiles to be imported into many design packages. Fusion360 is available free for hobby or non-commercial use, but may not be the easiest to learn in a short time. So many others as well. Use your favorite search engine for "gear generator modeling software" or similar wording and be overloaded with links.
Cura 3 speed selection
No you didn't do anything wrong. The problem with time estimates in slicers is that they don't know anything about the printer's firmware and physical limitations, so they estimate time based on the (whoefully wrong) assumption that the printer will do only and exactly what told to. So, if the gcode says "move 100mm at 1m/s" the slicer will assume your printer will take 100ms to perform that action. In reality though, your printer will take some time to accelerate, and it may actually be unable to reach the target speed of 1m/s altoghether, taking a lot longer to perform the full action. Typically, Cura estimates are OK for makerbots, and Slic3r PE for Prusa printers, as the software is mainly developed towards those machines, but for most other printers the estimates will be sensibly off. You may also be interested in trying to use octoprint to feed the GCODE to the printer: octoprint monitors the actual elapsed time against the gcode and try to adjust the estimate of the remaining time accordingly, also telling the operator how reliable that estimate is.
Prusa i3 MK3S 3D printer PETG printing problems
230 °C is way too cool for PETG and will result in underextrusion unless you print really slow, and poor bonding. Underextrusion in turn leads to stringing because of pressure build-up. I print PETG at 250 °C.
TPU Filament in Dremel 3D20?
TPU wants in general two things of your printer: A Printing Temperature of (over many makers) 195-230 °C A Direct Drive (extruder on the printhead) Bowden extruders are not ideal for printing flexible filaments such as NinjaFlex due to the excessive distance between the stepper motor and the extruder head. However, some users have generated successful prints using reduced speeds.ninjaflex handout Check the temperature you can reach, and you are lucky, as some of the smaller Dremels use Bowden but the 3d20 is apparently direct drive. If you want to try to run a Bowden with flexible filaments, dial down speed down really low (20-30 mm/s at most) and pray.
Tool to measure in STL files
I suggest blender. It's not the simplest of tools but it is free and learning it will improve your 3d printing skills. :-) (I write this answer also for future viewers of this question so I start basic). Check here for another answer: https://blender.stackexchange.com/questions/19772/how-do-i-measure-a-distance-between-two-points Import your STL-file. Press the Home key to view everything. Select the model by clicking on it with your left mouse button. (Blender changed to left-click-select as of version 2.80) Hit tab to enter edit-mode. Press N (or use View | Properties) until the Properties panel shows up. Select the "Length" checkbox in the "Edge Info" section of the Properties panel (see image below). Select "Edge Select" mode (see image below) Select the edge to measure by clicking on it with your right mouse button. If you need to measure the distance between to vertices with no edge. Create the edge by selecting them and pressing F. If you need to measure the distance between a vertex and any other point, select it and press E to extrude.
Octopi connecting just once to wifi
What worked for me: changed SD card from 2 GB to 16 GB (do not need so much) power the pi with 5.4 V instead of 5 - 5.1 V
Can you cure resin with sunlight through a window?
The glass will block most of the uv light; but not all. It will depend on the type of light that the resin is sensitive to; in order to determine if it will continue to cure behind a glass window in direct sunlight. Some resins also sensitive to blue light. You will need to look at the material data sheet for the resin to be able to know for sure. Be advised though, that the resin does not stop curing, and will continue to cure slowly over time, just sitting on the desk. https://www.thoughtco.com/does-glass-block-uv-light-608316 From the link: Glass that is transparent to visible light absorbs nearly all UVB. This is the wavelength range that can cause a sunburn, so it's true you can't get a sunburn through glass. However, UVA is much closer to the visible spectrum than UVB. About 75% of UVA passes through ordinary glass. UVA leads to skin damage and genetic mutations that can lead to cancer. Glass does not protect you from skin damage from the sun. It affects indoor plants too. Have you ever taken an indoor plant outside and burned its leaves? This happens because the plant was unaccustomed to the higher levels of UVA found outside, compared with inside a sunny window.
Can G-code scripts be run automatically on inserting an SD card when using Marlin Firmware?
I am not entirely sure if this is what you are after, or if it will work, but from this post on Printing From SD Card when printer is turned On While an LCD controller is not needed for auto#.g to work, if you have an LCD controller you can use the MENU_ADDAUTOSTART option to add a menu command that will run the auto0.g, auto1.g, auto2.g whenever you want. This (final) post ended up closing the thread so it appears to answer the OP's question - however, whether it will answer your question is another matter. This post, on the Marlin forum, replicates your question and asks about purely insertion of an SD card on a printer which is already powered on. A reply was received and apparently it should be possible by making a code change in ultralcd.ccp: It does not seem like an auto print, when inserting a SD-card, is to be found in the current program code of Marlin. You might consider making a feature request here: [github.com] I do not think that it would require much work to implement a "card.autoprintfile()" function in Marlin. It seems to me that a call to such a function could happen at line 5172 right after: else LCD_MESSAGEPGM(MSG_SD_INSERTED); in the file ultralcd.ccp A subsequent feature request has been made on Marlin's GitHub, see [FR] Can G-code scripts be run automatically on inserting an SD card when using Marlin Firmware? However, from this reply, it appears that a feature request is already pending: I'm not in favor of this as a general feature. We already have a feature request to be able to run custom G-code in response to various events, with SD card insertion being one of those events. Would that feature serve your needs? However, as yet, I do not have a link to this other feature request. Will update as and when. As an aside, in might be worth mentioning that the SDFat related code could be the place to code this in, see if SD card removed, how to flag and restart?, although after looking at the Marlin codebase, in particular SdFatUtils.cpp, this would appear to be a red herring.
Should I include fillets on my 3d printed parts?
If your part needs structural support, then the word is: absolutely. Fillets provide the added support when you need it. If your part has a meeting line which is sharp - 90° (or perpendicular), there is a natural stress riser in your design. This is a weak spot where a crack can form. If strength is needed and the fillet won't interfere with the design, it's definitely something you should include with your part.
Why won't makerbot accept an STL file from Blender?
Here is what I suggest you try. If you have a file that you can view/edit in blender I would export it as both STL and OBJ formats. Then take those files and upload them to Netfabb (https://netfabb.azurewebsites.net/) and get a "repaired" file. Have the library try again with the repaired STL and OBJ files. If this doesn't work try to get the exact error message/dialog that the makerbot software is giving them as well as the version of the software that they are using.