I agree with Chris. I want the CAD model to represent the actual part and not some fudge factor version of the part. If you don't want to tweak the dimensions of the CAD model and you don't want to try to tweak the printer settings which can be tedious and difficult, you can easily adjust the scaling factors in the slicer, independently for X, Y and Z. For a complicated part with many features, that may have you chasing your tail, fixing one feature and wrecking another, but if you have a simple part with only one precision feature, such as a bearing block with loose clearance holes for mounting screws and a precision hole where a bearing is pressed in place, you can print a copy, measure the bearing pocket, and adjust the scaling factor for each of the three axes and slice another part to print, all without altering the CAD file.
I don't think tramming is a thing with FDM 3D printers. The tool length and tool diameter are on the order of the nozzle diameter so tramming errors should be fairly insignificant. The extruded filament may slip a bit to the side if dispensed along a vector that's slightly misaligned to the Z axis but I think most of those small errors would cancel, with only a very slight distortion to the circumference of a vertical hole. If you level the bed to the nozzle with three or more points, the bed will be perpendicular to the Z axis so you should see no large scale errors across the part's surface as you'd have if, for example, you were fly cutting on a milling machine that wasn't trammed. On 8/4/20 7:52 PM, Gene Heskett wrote: > On Tuesday 04 August 2020 16:13:52 Bari wrote: > > I finally got around to measureing the 5x5 cube I did this morning. > X=5.31mm,10.40,15.45,20.58,25.55 > Y=5.25,10.35,15.40,20.4925.62 > Z=brim 0.70,5,61,10.54,15.57,20.43,25.39 > > The top 5x5 has a fat spot near its bottom that reads almost 5.9. Only in > X. I was not in the room at the time. brim is too thick. Raise glass. > And scale. > > So I need to subtract a bias from xy to even get in the ball park. > And add a teeny bit to Z. > > And this was after I had taken it apart to see in I could improve the > tramming which I did a considerable bit of, Z moves a lot easier both > ways now. MUCH less binding between the 3 point trolleys now. > But I can't find anything that could be miss-construed as bias. Home > offsets, yes but no fixed subtraction to compensate for bias. Cura post > processor job? Is there such a critter? > > Thanks all; > > Cheers, Gene Heskett _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
