Delta math adjustments

I checked inner dimensions and outer dimensions of this cheap printer that I have. I need to increase the size of the X,Y output, but have greater increase in size towards the center of the build plate than the outside of the build plate. Any suggestions on what numbers to adjust and in which directions?

The error you’re seeing definitely sounds like a result of incorrect values for your delta radius and/or arm length (either of which should also cause your layers not to be planar), but you shouldn’t go adjusting these blindly and measuring the parts you print. You need to measure your machine to find out the correct values to plug into the software.

I would start with the arm length. Delta radius is trivial to calculate at the CAD stage, and should never be incorrect. I suppose there could be tolerance-stacking issues that throw it off in the build, but that would be exceptionally shoddy workmanship. Delta arms, on the other hand, are notoriously difficult to manufacture to a reliable length specification, and really need to be measured per-machine.

1 Like

I doubt I could measure it any more accurately than the rough measurements that are in place. It is not a build. The printer is known to not have the most accurate measurements and assembly but it was cheap. I imagine I might be able to measure within ±1 mm. That is about as good as I figure I could get.

If you have a measurable difference in the error of the size of your parts from the center to the sides, the numbers in the firmware might be further off than that.

Of course, you should also check the steps/mm of the linear axes, and check those axes for backlash, which are other possible sources of error.

I am closer. I went from M665 L120.80 R62.7 to M665 L120.17 R63.1486. Less than 1mm in either direction for distances this small make much more difference. It looks like my steps/mm is the printer default and not the numbers that the wiki folk determined were actually right. That may explain things being a bit small. I will try the steps/mm and then the radius and length changes which are less than 1mm difference each.

After that change, I changed the steps/mm and so things are slightly larger now. Dimensions seem more accurate. It is a bit more steps/mm than the community around the particular printer decided was right. Oh well there. I increased overlap between the skins and walls so they merge properly. I think it was too much flow vs size before. Now it is just right or slightly lacking. I am not sure. I will test more later. At any rate, the increase in printing quality is way up now.