What does then G33 L0 return? There you can quite good see how good leveling worked and where the problem lies. I always use the escher delta wizard with 6 parameter to calibrate. Gives very good results on my printer.
Same forbidden picture. Guess link was ok, just image is private (see in incognito browser for testing).
It is very time consuming to calibrate a delta by hand. Alone adjusting the geometry using escher delta wizard requires you the measure many heights and it must be precise. Cartesian printer by hand is ok and best also time consuming. But for deltas the costs for a z probe are very well spend. Deltas are never correct with respect of geometry resulting in z going up and down depending on xy position. Also I did not see your image I guess that is your problem here.
Yes, it took me haft a day to make the position below 3 axis and the x0 y0 z0 at the same high. But the position opposite of z axis is higher than the other. z ......... ............... .........0........ ................... x ......+...... y ...++... it is something like this where the plus one is the one higher than the other
Yes, that is some error you typically correct by modifying the tower angles slightly. If they are not 100% the height differs a bit depending on position. The escher tool would say you how much but requires z heights at different positions. If I hear half day for that calibration I see that 50 euro for a z probe are well spend since I know you sometimes need to recalibrate if you modified something. In your case move x and y tower towards z tower. Maybe 0.1 degree. Tiny changes can make already a big difference since we want to correct only 0,0x mm normally. I mean it already looks quite good for manually calibrated.
There is also the option to use distortion correction to help with first layer to better stick. While this normally also requires a z probe you can set values also manually.
Yes! I did it. Thank you very much. It is exactly because of the x and y tower. I used Escher as you mentioned earlier to calculate. The result was soooo good!
Last trick to solve remaining difference is use a higher first layer like 0.3mm - it will forgive better the differences. But looks very good for manually levelled.
Comments
Have you run
G32 S2
G33
What does then G33 L0 return? There you can quite good see how good leveling worked and where the problem lies. I always use the escher delta wizard with 6 parameter to calibrate. Gives very good results on my printer.
Here is the link to the picture: https://lh3.googleusercontent.com/gTLejidyUWxIA8zxD8Vksh9bKygrNmIKxADojcZCqarZi46RexVuJPmp5OEPKobuXVG79turRWEVsL8EWq4RzqObseLNUoOmdSXQb20LEEfzWIOJ_Mct6sTl-IQGrZOw8oG7P5LQL-bHFtXKhlvbI2HHudP8MK5hf-iK6NdPc8obElxkHHs2Ii7fDclpDvbwG23lpbOx8l4iFYjGqX0EJUJj95o5zEKIKKoRFff6sMj6CdfcRweQItDqAIcMvXblN_9X4RZEH0RBYnKhv7l5JwMbJG8LQ4P5XtBe5kJt0-5gEKa8uNIDZAS-Z1DqoaKCgSB-Qkespl4_scz2guigGVYS_QyidNoysHUlsxW0liMG1BSO06uwvXeE8nB6lA8-Gb8YIN7JpN6cJoUnDiB-CWvhD42zhppFqACnvoo7b5I9MNFqwXEJuoQi7VhpCLdpj9S95lZXtpFHkKQLXt5yiEDfflXESW88ub50TVThY0AbrI31NK87q_XZe5i3ZABw2TavURWwL6LjjPWUG7E7mMHhtrOHYEjvldcq-LJ-l-AYkiSoXitEx1SvL8kEMhRQKMBt7VV83sV8ikwO_P1QO_eO3yI4DaRtV4VxSQ5FhlVIywCrhYVYlIi3yS288T438C2X9U_5Go23EY5mNqy4OWgkuldmblM9W4d_JdiyEdKX24quFb1_HtXDru2chw=w1263-h947-no?authuser=0
It is very time consuming to calibrate a delta by hand. Alone adjusting the geometry using escher delta wizard requires you the measure many heights and it must be precise. Cartesian printer by hand is ok and best also time consuming. But for deltas the costs for a z probe are very well spend. Deltas are never correct with respect of geometry resulting in z going up and down depending on xy position. Also I did not see your image I guess that is your problem here.
Yes, it took me haft a day to make the position below 3 axis and the x0 y0 z0 at the same high. But the position opposite of z axis is higher than the other.
z
.........
...............
.........0........
...................
x ......+...... y
...++...
it is something like this where the plus one is the one higher than the other
There is also the option to use distortion correction to help with first layer to better stick. While this normally also requires a z probe you can set values also manually.
I used Escher as you mentioned earlier to calculate. The result was soooo good!