I created a sign for my son and his family. During the preview everything looked great. When I finished the 3d finish cut I found a line across the sign about 2/3 the way up. It looks like the cut went slightly deeper and to the right a small amount for maybe two passes. See Pictures So I thought it could be software related. I rebuilt the sign creating all new components, and reran it. Same thing, just not quite as bad as the first. Line is the same place as the first. I did have the blank is a different place on the bed. This to me kind of eliminated a machine problem. Has anyone experienced this before? Does anyone have any ideas on the cause. Using: Aspire Gerry's 2010 PP Mach3 .0625 Tapered Ball Nose
unlikely to be software, more likely to be hardware, however, you can test... create a much simpler design and run it. must be about the same size. do the rough and finish passes separately so you can see if the rough pass does the same thing as the finish pass. reduce your acceleration by 10% and see if that changes anything. you will of course have already tried using slower feeds, but be careful, there is such a thing as 'too slow'. did your cellphone ring while you were cutting? they can cause some interesting interference (-: however, a software fault is possible since it only affects one or two passes, if a motor lost steps then it would persist with the errors after the loss. ergo, you need to cross check the Gcode very carefully in a simulator. I use the Planet-CNC control software (without hardware) just to visualize the toolpaths, and then also put it through OpenSCAM. The advantage of plant-cnc is that it actually error checks the code as it reads it, the advantage of OpenSCAM is that it shows the remaining material. both allow you to step along the code and see what is happenning 'line by line'.
Hello George, First, awesome working on this carving! I thought from the first look it was a missed step, but if it happened twice at around the same spot I would say that it may be in your linear bearings or maybe debris in the track of the wheel? Not sure what linear system your machine is using. It still could be a missed step even if your driver has a thermal shutdown for a split second as well, but I would say its more or less something along the rail or bearings. Maybe you could disengage your drive system on that axis and move it freely to see if there is a problem along the rail. Hope this helps Mark EDIT whoops looks like David was typing at the same time
David, Thanks for the feedback. The more I investigate it I think it is a problem with a particular component in the design. The basic design includes the same base, border and banner. I then can change out the names and main component (Ram, fish, bird etc.). I had previously done four of these signs using a fish. I then rebuilt it with the ram. The ones with the fish cut just fine. When I built the one including the ram, I remember having a little problem seeing the banner correctly. I then cut the sign and found what you see in the pictures. I then redid the sign rebuilding all the components EXCEPT the banner - and when I cut it the same issue was there. Exactly the same place and same size as the first sign. It was also placed at the very top of what would be the banner by itself. That to be points to a problem with the banner components. Does that make sense to you?
Mark, thanks. In my reply to David, I am believing it is a problem with a component in the design. I am using open rail (you can see that I had the same problem with wheels as Gerald) with R&G. Since I had just disassembled the gantry to install steel wheels, I am confident it is not bearings. The gantry runs good without the motor engaged. I just cant see how it could be a missed step in exactly the same place and for the same amount two time in a row. But granted I am new to all this so am certainly open to be corrected. Thanks for you input.
Hey George, this is a tough one. If you think it is the ram then try a small sign with only the ram and see what happens. It's a possibility that the component geometry is causing code that is making your Z stepper miss a step then gain a step. This may or may not be in conjunction with motor accel. settings. If the problem persist then adjust your accelerations as Dave said. If it still appears then revisit the component and try to reduce the emboss if possible. Seems far fetched, but worth a look at. Keep us posted. Joe
I think I confirmed that this was a software problem. Took the sign and stripped it down to just the base, banner and border from the existing sign. Ran it and had the same flaw. Next rebuilt the sign using the same but all new component - base, banner, border. Ran it again and no flaw. So the banner had to have had some glitched introduced into it at some point that caused it. Now to tear my machine down and replace the open rail with steel rail.
