Hi all, I am looking for some help with a problem that has me stymied. I am running with aspire 8.5 which has given me no trouble(that I didn’t cause myself) for the past 5 or 6 years. I recently put together a new cnc which I am running with the openbuilds blackbox as well as their control software. The problem showed up when I ran an old program that was fine on the previous cnc. It was a vcarve design set to a max. depth of .1”/about 2.5mm the final cut was 3.8mm deep and the z axis didn’t retract at the narrow points but instead ploughed into them at full depth. I used the GRBL pp that was in the aspire list. The x and y axis were the correct size. all axis were calibrated to within .02mm and the jog function was right on however when I used the command line G0 Z10 it moved 20mm. I had to use a Z- command between any two Z+ commands Basically I’m stumped at this point. Any help would be appreciated Andy
Z-axis steps per mm Calibrated? Or could be that you have Z axis Acceleration or Max Rate set too high causing stalling during retracts. Stalling during the move can affect your calibration (it stalled at 10mm, but steps per mm are closer to 20? Check the mathematical steps per mm calculation to see if you are in the right ballpark to confirm)
Thanks for the suggestions, I have calibrated the z axis, it’s within .02mm at abt. 209 steps/mm I cut back the max rate and acceleration but I will try cutting back further. I tried using the sharp cnc commander and found that the g0z10 commands worked consistently in both z+ and z- but the program still cut too deep by about 1.3mm (on a 1mm flat depth) and didn’t retract. I got close to the right depth by setting z0 1.3mm above the surface. I also tried a flat depth of 1mm in .25mm increments first cut was over a mm the other 3 were right around .25. I also tried a profile and pocket cut x,y and z were all on the money. I’ll try a vcarve later today with a decreased max rate and acceleration
Slowed the max feed to 250 from 500 and acceleration to 50 from 100 with no joy at all. I’ll take the z gantry apart next to make sure every thing is tight and square, the motion is a little jerky at times. Any other suggestions appreciated
Grip the endmill and push it up and down, see if you detect movement in the Z axis, or even relative to other axes. Or, does the "jerky" match the symptoms in section 1 of docs:blackbox:faq-identify-motor-coils [OpenBuilds Documentation] Mechanically, check shaft coupler setscrews, stop collars and leadnuts.
From the description it sounds more like a weird G-code error than a mechanical issue- G90/91 or unexpected G54 origin or something- though it never hurts to quadruple check the mechanicals, keep dialing that stuff regardless, can't be too square and tight.
Well, I’m a much happier camper today although perhaps a bit sheepish, my problem appears to have been primarily mechanical. When I took the z gantry apart I found some slop in both the leadscrew nut block and the top bearing. Fixed that and now both flat depth and retract at narrow spots is as it should be. Still don’t know why, when I send commands directly as in G0Z10 I get some fairly random results with the openbuilds software but when I use Sharp Commander software, I get what I expect. Not so much a problem as curiosity at the moment as I have never needed to program a job Thanks everyone for the help, Andy
G0 happens at Max Rate, so you are probably still stalling because you haven't got properly tuned Max Rate/Acceleration: See gnea/grbl Some hosts override that by doing G1s at a low speed, but thats bad practice, as some jobs may send G0 moves causing stalling Our Jog buttons use $J actually: See gnea/grbl