Prusa XL belt tuning: ridges at top of prints until backing off tension
I just reassembled my 2T XL after a debacle with the whining and screaming toolheads. Prusa sent over new linear rails and new pulleys. I installed all of them and reinstalled the belts. If anyone is willing, I would like to compare notes on the belt tuning, because the results surprised me.
First thing I did was make sure the X rail was square on the gantry, then I ran the belt Hz tool and printed one of the Prusa rockets. It had layer shifts and ghosting toward the top of the print. I tightened both belts a half turn and printed another rocket. It had fewer issues, but still issues toward the top. Then I printed just the detailed top section by itself, low on the bed, and that came out nearly perfect. At that point I figured there must be something going on with height, and maybe the belts were not tight enough, since all the other calibrations are good and shorter models print fine.
After that I restarted the whole process with square gantry, Hz tool, and the homing tower. I kept tightening to chase out the ridges, but as the tower got taller, more ridges would show up near the top. This meant starting over. Each of full tower takes about 6 hours, so I abandoned a few, each time restarting the whole process. One of the other towers actually did have a full layer shift, like in the Prusa guide.
So I tried again from scratch, but this time I kept my Hz on the lower end. Instead of continuing to tighten, I stopped and tried the opposite direction by loosening both belts a quarter turn. The ridges went away, and for the first time I got a homing tower with no ridges at all, from close to the bottom all the way to the very top.
My belts now read about 75 Hz on the right, which is a little low, and about 80 Hz on the left, which is supposedly just right. I am measuring with the Prusa Hz tool on my phone and it has been consistent for me. Throughout this whole process, even when the belts were too tight and causing ridges on the homing tower, the right belt always reads about 5 Hz less than the left belt if the gantry is square. I hope that does not mean my gantry is tweaked, or that my belts do not match in their properties.
I am posting this because the instructions only mention tightening to remove ridges. They only mention loosening for large layer shifts, which I was not seeing at all. In my case, overtensioning seemed to be the cause of the ridges. At this point I would like to tune it further, but honestly I am a little afraid to touch it. After seeing so many reports of hit or miss behavior on this machine, I am tempted to just be happy with whatever quality comes next and start printing. At any rate, if you have any experience with belt tuning on the XL, or any other tuning for that matter, especially when taller prints give problems, I would love to hear it.
Here are some guides that I had read. This one suggested I hold the gantry tight and square while first turning the screws, that seemed to help quite a bit: https://investegate.de/prusa-xl-belt-tension-alignment-calibration-guide/
Prusa guide of course: https://help.prusa3d.com/article/adjusting-belt-tension-xl_401793
Ellis helped me with my expectations on print quality, and just generally helped me understand things about these printers: https://ellis3dp.com/Print-Tuning-Guide/articles/setting_expectations.html
I am using firmware 6.2.6 since 6.4.0 brings back the screaming toolhead issue.
