I wasn't able to find any info on this on a quick search, sorry if someone's addressed this issue before.
I have a v1.2, that was just overhauled, squared, shimmed, etc. etc. After scrutinizing the first calibration print after re-assembly, an interesting defect caught my eye for the very first time: Full lateral layer shifting on the columns (using the older calibration file) towards the F+ Direction! (towards the front of the machine)
I then looked back at as many saved calibration prints that I could, from approximately 6 months after the printer's first print, up until now,(about two years worth) and lo' and behold, the defect was also present on every one, regardless of resin or machine settings! Albeit, to different extents. It seems to be less pronounced on 30x-y-z grids, but is still present. I'm fairly sure I've noticed it before, but shrugged it off, as parts were printing. However, I've never seen the kind of high quality results that many of the forum members have posted from their 1.2's. In any case, I'd like to solve this puzzle, as it would be nice to have the printer perform well in parts that require large surface areas, or to see if it enhances support structure attachment.
In recent prints, the shift seems to start at the transition between the full projection area "base" layers of the print, and at 50um X,Y,Z, tapers off over the first 10 layers of the columns. The shifting appears to "settle," and features are then printed approximately straight for the rest of the build. It's uncertain if the base layers are also not printing right on top of one another, as it's difficult to determine where the edges of each base layer stop. Visually, the shifts appear to be uniform across the entire layer, but I haven't actually measured it yet. It happens with the new "square" pattern calibration files as well. Here are a few pictures to help describe it:
Any idea what's causing this? Is it a normal artifact due to the design of the printer, when printing this particular transition from a large surface area to a small one?
I flipped the linear motion guide rail to the opposite side during the re-build, so if it was a defect there, it would have to be in the bearing truck itself, and I'm not sure a bad bearing or three would repeatedly be in the same position each time. That being said, the LM guide does have a noticeable amount of play present (always has). I can actually feel the arm rocking up and down a good bit when doing a static test at the end of the arm by wiggling it gently.
Perhaps there are a few other problems with my machine that I haven't found yet.
This tear-down, I did notice that the horizontal frame member (black plate which hosts the electronics, stepper, guide rails and such) was pretty warped. In the end, I had to adjust the the front end of the plate downwards by about 2 degrees in order to ensure that motion of the Z-axis remained perpendicular to the red slide plate (measured with a quality Dial Test Indicator). In retrospect, I should go back and actually measure the ball-screw to LM guide rail spacing, to see if they're actually parallel. At first glance, they do appear to be.
Looking forward to hearing your thoughts,