I too have been experiencing issues with printing at the outer radii of the bed. I read several posts on this forum and decided to do a little data gathering and post the results on this forum.
Some baseline definitions, so we aren’t operating on assumptions:
The “X” tower is located to the left and slightly behind the LCD panel, and is at the 210 degree position
the “Y” tower is located to the right and slightly behind the LCD panel, and is at the 330 degree position
the “Z” tower is centered and on the opposite side of the borosilicate glass build plate, at the 90 degree position.
The tower positions correlate to the tower adjustment parameters of Alpha 90, Alpha 210 and Alpha 330.
Attached are a number of graphs depicting the elevation of my nozzle above the build plate at 15 degree intervals from 0 – 345 degrees. When the fuscia-colored line is further out from the center of the circular graph the nozzle is higher than it should be; when closer in the nozzle is lower than it should be. The circular axes represent nozzle height and are labeled from -0.3mm to 0.3mm in 0.1mm steps; desired height on the graphs is 0.0mm.
Each graph includes a data table similar to:
Rostock 100 Alpha 210.50_90_330
Max=0.15
Min=-0.16
Delta=0.31
Jump = 0.12
• “Rostock 100” indicates that the readings were taken 100 mm from build plate center.
• The tower rotation parameter that has been changed from default is always listed first; in this case the 210 parameter has been changed to 210.5 while 90 and 330 remain at their default values.
• “Max=0.15” indicates that the nozzle, in its travels around the bed, rose a maximum of .15mm above ideal elevation.
• “Min=-0.16” means the nozzle dipped 0.16mm somewhere around the circumference.
• “Delta=0.31” is the algebraic sum of “Max” and “Min” and shows the total deviation of the nozzle, from its lowest elevation above the build plate to its highest.
• “Jump” number indicates the greatest change in elevation that occurs in one 15 degree position change. In the above data set, the most the nozzle changed elevation between two adjacent readings was 0.12mm. I tried to choose a tower rotation setting that gave the best combination of roughly circular shape, lowest Delta and lowest jump. These values seem to be mutually exclusive, i.e. settings with low Delta have higher Jump values, etc.
The are 8 graphs([1] Baseline, Alphas are at default values, [2] 210.4_90_330: the setting I will use for the foreseeable future; [3-8] graphs generated when varying each of the three default values by 0.5 degrees; all following the above format but with different Alpha rotations dialed in. I thought this visual depiction might be a good way to grasp the effect of rotating the towers a half-degree or so.
The process used to gather data:
• Careful setting of nozzle height at 90 (Z tower), 210 (X) and 330 (Y) towers using the paper thickness method as described in my build manual, including adjustment of Horizontal Radius as needed, per the manual.
• Hot end at 240 C, bed at 60 C; at temp for 5 minutes minimum before taking any measurements. A digital dial indicator was used to obtain elevation measurements.
I have attached an Excel worksheet showing position calculations (X and Y values) for every 15 degrees around the build plate; radius = 100mm.
I developed a script to move the nozzle around the build plate and pause for 6 seconds (to allow time for me to record the indication) and then move to the next position. I have included scripts to allow the recording of these measurements at 50mm and 100 mm radius from the center of the build plate.
I have included graphs showing the baseline (how my printer performed with Alpha settings at 90, 210 and 330) and for several tower rotations. For my printer, best performance occurred at Alpha settings of 90, 210.4 and 330 (graph: Rostock 100 Alpha 210.4_90_330).
I hope this data and the attached tools (Excel worksheet and scripts) help to provide some insight into Tower Rotation behavior. See “Tower Rotation #2 (Revised)” for remaining files.
Also, added a pic of my dial indicator mounted on the effector.