Quick Backlash Calibration Method

thingiverse

You should probably check out version 2 instead: http://www.thingiverse.com/thing:2040624 This is a set of prints that can quickly tell you your printers backlash settings. I designed it the M3D Micro in mind but there is no reason why it should not work on other cartesian printers as well. This is by far the quickest method for calibrating backlash I have seen as it produces empirical results in one go instead of using trial and error. If you own an M3D Micro I recommend using the gcode files I have prepared. If you own another printer you may have to slice the files yourself. Disclaimer: For my own printer I am fully satisfied with the precision of this method. Others may find that it only brings them in the right ballpark for further tweaking with other methods. And others again may find that results produced with this method are too far off to be useful. Make sure your spool-location does not affect the movement of the print head at all (This last paraphrase probably only applies to the M3D Micro as it is very sensitive) Print Settings Printer: M3d Micro Rafts: No Supports: No Resolution: 0.350 mm Infill: None Notes: The files need specific slicing settings as most slicers tend to ignore features that are as flat and slim as these. For this reason I recommend using the gcode files that have been modified and tested with both the official M3D software and M3D Fio/Octoprint. Turn 90 degrees in the slicer when making the Y-files if you choose to slice an STL file yourself. Post-Printing How to measure Very important notice: In order for this print to work you print bed must be very precisely calibrated. Otherwise the print will come off the bed. My own preferred calibration method for the print bed is this: http://www.thingiverse.com/groups/m3d-the-micro-3d-printer/topic:5062 But print bed calibration is very individual from one type of printer to another, so calibrate the way you prefer. Adjusting the Y-axis: In your printer software set Y-backlash to 0 before printing. Print one of the gcode files that has a Y in the name. Most M3D Micro printers will need the QuickBacklash_Y_0.70-1.69 file. If all of the samples are curvy try the QuickBacklash_Y_0.00-0.99 instead. Watch the print start and remove the extra filament that is extruded in the beginning. It is just to prime the nozzle. The included files will make 100 samples. You have to study these samples very very carefully. You have to find the sample with the highest possible value that does not curve. We are talking about a curve that only moves 0.01 mm to the side. It sounds impossible to spot. But it is actually doable. I recommend having a ruler in one hand and a flashlight in the other. I never take out my build platform. But if I did, I would probably use a magnifying glass. When you have located the correct sample, calculate the backlash value using the chart/picture above. The 0.7 file will begin at 0.70 mm in the corner marked with a blob. The 0.0 file will start at 0.00 mm I think it is a good idea to make three prints and take the average value spotted for each. If one of them is way off disregard it and make another. If in doubt choose the lower value. Enter the value you found in your printer software. Adjusting the X-axis: Similar to the process for the Y-axis. But most M3D Micro printers will need the QuickBacklash_X_0.00-0.99 file. If all the samples are flat try the QuickBacklash_X_0.70-1.69 file instead. How I Designed This Designed in OpenSCAD. Sliced in Slic3r and finally the gcode was tweaked in Notepad to work with the Micro's native software.