Hello. I have an openbuilds lead 1010 machine (bought from makerparts.ca if that matters). I use UGS as gcode engine, black box as controller, cambam as cam and inkscape as vector graphics software. The inkscape+cambam are reliable on my other machine (diy with some noname controller run with mach3) and I dont think ugs is a cuplrit here but let me explain what is my issue: Almost all cuts I do are about 0.5mm too vide. All are straight, all are 0.5mm too big. In both axes. The error does not scale much. At least as far I can measure (I work with plywood so its sometimes tricky to measure things with better yhan 0.1mm precision). Till now it was not a big deal as I did not need that big of a precision on this machine but its time to tackle this problem. So my questions are: 1. I tried to cancel any backlash on lead screws, and I dont feel any play with my hands. Is there a well tested way to set the lead screws which is 100% good? 2. Is this machine reliable to 0.1mm at all? Probably that should be first question but I beliewe it is, asking to be double sure 3. This 0.5mm looks suspiciously specific. Is there a setting in ugs for, i dont know, cancelling backlash and I have it set? Looked for such thing, not found. 4. Does anyone else has such problem? A bit more background: Yes, I checked dimensions in inkscape, cambam, mill size. All looks ok I did calibration. 6mm dimension turns to be horizontal 6.5mm, horizontal 87.312mm turns to be 87.58mm, vertical 26.458mm turns into 26.91mm etc... Edges are not wobbly. checking with calipers edge they are perfect. Edit: Checked the dimensions in cut viewer pro and they are as they should be. so the flow between inkscape and gcode seems to be ok. Either ugs is doing something funky or the machine/blackbox... The inkscape and gcodes added to this post if anyone is curious. Maybe there is some wonky setting added to header of gcode? Or ugs does some creative interpretation? Any help appreciated
Not likely to be a black box or UGS problem. How did you calibrate your machine? Do you measure the actual diameter of the bit before creating the G-code?
I did the calibration in UGS. There is a wizard for that. Yes, I measured the bit. Its the same as in cambam tool definition. If the dimensions would be too small I would think that its a matter of play/backlash but the bits are "normal" so they dont bite into material so the backlash if existing would make dimensions smaller. Also I played a bit and cut out 1mm strips from plywood. They are straight and smooth. No wobble on any axis. I did that by just defining a shape (rounded rectangle) and set the bit to 3.175 for first cut and 1.5 for the second. It cut a rectangle plus this thin strip. Its all fine and even, so I dont think its a machine problem, but Im willing to dig deeper.
Endmill ground 0.5mm over diameter? Caliper the endmill / cut a straight slot and measure the kerf - and use the exact value in your CAM setup.
Calipers show 1.44 to 1.56 depending on the location of the measurement. Its plywood and calipers dont measure deep. Made a photo of the cuts, mills, ruler. All looks ok. But dimensions are around 0.5mm off.
6mm movement for calibration is much too short. use the longest travel you can, with an accurate scale and a V bit to 'point' at the scale. check that your bit is in fact 6mm. it might be 6.35mm. measure it with a micrometer and use that for the bit size in the CAM
Calibration was done over the span of 600mm. The bit is 1.5mm. The slot was supposed to be 6mm. Is 6.5mm. Similar difference is on bigger dimensions. Its not calibration issue.
Good point. Did not checked that before. After checking all looks ok. Seems that I assembled it decently. Tramming looks good all is square, the cutouts are also square.
C Can you cut a rectangle, say 100mm x 100mm, then measure the width and height of the finished piece? Thanks.
OK, assuming your machine is calibrated and has no slop in the axes (an absolute must).... Measuring the diameter of the bit is the first step - they are often undersized relative to the stated diameter. Draw a simple part as a test for the bit to cut into a scrap piece of material. Enter the bit diameter into your cam program, create the toolpath and cut the part. Now measure the cut part. If it's still undersize, enter a slightly smaller bit diameter into your cam program (e.g. 2.85mm instead of 2.95mm) then re-calculate the toolpath and cut again. Do this again if necessary to get it correct. I do this every time with a new bit and record the "tested" bit size accordingly, ready for the next time I use it.
The tool diameter is fine. That was the first thing I checked. No idea, I compensated for this in tool diameter and for my use its ok. I tend to think it is caused by the flimsiness of the screw driving the gantry.
