Jump to content

jsamuels

Dormant
  • Posts

    38
  • Joined

  • Last visited

    Never

Posts posted by jsamuels

  1. Continuation from this thread... viewtopic.php?f=7&t=1484

    I'm printing this thing:

    http://www.thingiverse.com/thing:16813

    Per the SCAD file - OD should be 15 mm, but the printed part measures 14.6 mm.

    ID should measure 8.1 mm but is barely able to be forced onto an 8mm rod with significant pressure. Length is 24mm exactly as it should be.

    There is a seam / ridge on the outside of the part running along where the Z axis moves are made, it measures approximately +.4mm, giving an OD there of 15mm - which is the expected OD. But of course totally wrong still.

    I also still have a slight ridge at the bottom, very minimal at +.1mm.

    Image :

    http://i.imgur.com/oAD7u.jpg

    I have attempted as much calibration as I have possibly been able to do, filament diameter, E steps, adjusted the bed, etc. Many cubes have been produced and I'm no closer to getting accurate and high quality parts. I don't know what else to do and I must say I am very disappointed in my purchase at this point. Please let me know what information I can provide, to help troubleshoot and solve my issue. And thanks in advance for any help.

  2. No, its not the only factor, but it seems to be a big cause though. Other things to check: the bed needs to be accurately leveled and zeroed, and the XY axis needs to be properly square, aligned and lubricated.

    Filament diameter should also be carefully measured and set. The filament I received with the machine from Ultimaker seems high quality and has very little variance in diameter, however some other filament I have from a different supplier is oval by nearly .2mm and varies as much as +-.2mm more when measured in multiple spots along the length.

    Beyond that getting your machine settings dialed in (speed, temperature, etc) will improve your prints. But you need a high quality and well adjusted base machine to start with, to see the best results.

    I'm pretty new here, but it's what I've learned so far.

  3. OK, I have adjusted Z axis, to where I can barely get adhesion to the printed bed. I've changed the initial layer to .2 mm, and lowered the temperature to 195. I added 5 skirt lines to hopefully work out any issues with the first plastic coming from the nozzle, and lowered infill and infill overlap to 10% - though I now have gaps in between perimeter and infill, so I'll need to raise this back up.

    I created a simple test part in TinkerCad, a 20mm disk 5mm thick, with 8mm 4mm and 3mm holes and printed quite a few of these as I performed each adjustment. The best result:

    Thickness of the skirt is approximately .25 mm.

    The part thickness is 4.95 to 5.00 mm; the edges are slightly curled, making them thinner than the center.

    The bottom flared area measures 19.95 to 20.05 mm.

    The straight area on the sides measure 19.65-19.75 mm.

    Also, the holes are undersized, but I understand this is likely due to the low poly count on the circles combined with thermal shrink. I don't have pin gauges handy to get an exact diameter, but I'm unable to fit an M4 screw (OD = 3.9mm) into the 4mm hole, nor will an M3 screw (OD = 2.9mm) fit into the 3mm hole without threading them in. They cannot be easily threaded into or through the bottom due to the flared ridge (which exists inside the holes as well).

    Here are a few images:

    First layer printing :

    http://i.imgur.com/VETPo.jpg

    Bottom of the object:

    http://i.imgur.com/xD4HR.jpg

    Side view of the object:

    http://i.imgur.com/VxucU.jpg

    It's obviously improved, but still a quite obvious ridge. I'm still not happy with it... I will keep tweaking though I'm not really sure what else to try just yet.

  4. Is it possible to have the project planner print per-layer instead of per-object? What I mean is, if you add 5 models to the project planner, currently it prints object 1, then object 2, etc. Is it instead possible to print layer 1 of each object, then layer 2 of each object, and so on? This would enable tighter object packing as you don't have to worry about the head crashing into each one. However a drawback would be if you have a failure at some point it ruins all the objects, instead of just the one it's currently printing.

  5. Thanks for the replies.

    Daid: I leveled the bed while cold, and zeroed the Z axis while hot (after running a few prints actually). The tip of the nozzle is barely touching the bed. I can feel slight interference with a .0005" feeler gage. I'm not sure how to get it any more exact? BTW - it would be nice to have Z-axis compensation in your Cura software, instead of fiddling with a screw.

    MSURunner it's a good thought. I'm not sure how it might account for this effect as the higher speed above the bottom layers should also extrude more plastic? I can test it though.

    I will test these suggestions this evening, and report back. Thanks!

  6. Hi guys – new Ultimaker owner, and I’m having some issues with my prints. The bottom several layers of my prints are flared pretty heavily, so much so as to be unusable without some tedious hand-finish work. There’s also a groove in the side of the prints where the nozzle makes the movement between the inner and outer layers of the perimeter. I took some photos which shows the problem:

    http://i.imgur.com/VTuMI.jpg

    http://i.imgur.com/ZFNxw.jpg

    These objects are sliced with the default settings in Cura and printed using Printrun, I haven't played with the advanced settings yet but I think I should be getting better quality.

    Things I have tried:

    I leveled the bed using a dial indicator, as best as I can – it’s within the flatness of the acrylic build platform at least.

    I measured & set the filament diameter – roundness is quite good, while diameter varies only a bit from 2.79-2.81 mm; I used 2.80 as the average.

    I zeroed the bed using a fine-adjust Z axis limit, a .0005” (.012mm) feeler gage fits with slight interference, while a .0015” is very snug and deflects the bed slightly. I did this in the center of the bed where objects are printed as there is some irregularity in the bed flatness, a “potato chip” shape if you will. But unless I’ve done something wrong, the center of the bed is flat and zeroed.

    I checked the extruder steps per mm as per the first-run wizard, and it’s quite close, within a few mm. I didn’t alter it from the default value.

    I adjusted the Z-axis zero, if I raise the nozzle higher on the bed it’s improved a bit but I can’t raise it very much or the parts won’t stick, even with 2 layers of blue tape.

    I’ve also tried temperatures from 190-210, printing speeds from 30-75mm, and two different batches of filament. Cooler and slower has some improvement, but it’s not perfect.

    Deleted/reinstalled Cura, and tried it on a 2nd machine (both are Win7 64bit). Same result.

    I also note that objects are undersize – the 20mm calibration cube is ~19mm on the upper layers, and an M3 screw / nut won’t fit in the holes per the design:

    http://www.thingiverse.com/thing:24809

    Is it possible somehow I’m not getting enough plastic extruded past the first few layers somehow?

  7. Sander, my Ultimaker (order #100003698) also had some limit switches with stripped threads. I'm certain it came like this from the package, as after the first stripped when I barely tightened it down, I checked the remaining carefully and found the issue. I was able to fix it myself with an M4 tap and screws, but not every owner would have such things readily at hand.

    You may want to consider changing to a limit switch made of a more durable plastic or ideally one with metal insert for the screw thread.

    I might also suggest changing to a slightly longer screw here, as the 10mm or 12mm+washer on the Z limit switch does not fully extend through the part, limiting the thread engagement length. This increases the potential for the thread stripping out. You may also consider a flanged button head for this application, it should provide a more secure mount. An M3x12 flanged button head possibly?

×
×
  • Create New...