Thanks for the responses and ideas. It looks like the problem was indeed a wobbly bed.
In the last 48 hours I created a smaller box to test things out. Instead of a 120mm W x 120mm L x 120mm H box with a full bottom, I created a 40mm x 40mm x 120mm box with a rectangular ring at the bottom instead of a full floor. This was to reduce the print time and material waste. This test box included nubs on three sides and holes on three sides. The plain surface side was positioned to be toward the back of the printer.
I first printed out this test box in black PLA and you could see surface roughness like previous prints. Like before the roughness was affected by both the nub and the holes on adjacent surfaces. Interestingly, there was a small rough patch on one of the surfaces with the nub that was an “echo” of a hole on the surface opposite from the plain surface. Rwide had suggested moving the object to the front of the print surface, but by reducing the footprint of the object to 40mm - and thus moving the back surface forward by 40mm - I think shows that this is not the issue. In the picture below the bottom of the print is on the right, the top on the left:
I did investigate the two links that xchrisd sent along related to the see-saw print surface and spent a couple of hours dismantling my printer and making adjustments to the Y linear rails. Determining the looseness of the slider was difficult, but I made the assumption that mine was loose since I had observed the print surface rocking back and forth during repeated Y movements. After reassembling the printer, I reran the test box (now in purple) and the result was that there was no roughness on the plain surface - or any surface. There is subtle evidence of some “echoing” of the nub and holes, but only on the surface where those items are located, not on adjacent surfaces. Below is a photo of the Before and After runs of the test box. Note that both of these were run with the Luban setting of “Normal Print” and not “Smooth Surface” (which adds significantly to run time).
Again, thanks to all for your suggestions and comments. Looks like it was just a wobbly print bed.
While I think I’ve solved the problem for now, I would like to find a metric - for the future - that would allow a more definitive way of measuring “how much wobble is too much” without relying on pure visual observation. If anyone knows of such a methodology, please pass it along.
A lot of us added linear rails to support the bed. They’re not expensive and it does work. A few 3d printed parts and some screws. You don’t need to bolt it down; just set the rails on the frame.
I found the linear rails you mentioned. One of the comments in the forum posting suggested that this was really only appropriate for CNC due to the pressures involved there. Can I assume that you - and others - have found that this minimizes the see-saw action when 3D printing?
You could be giving yourself a false positive here by changing the material and thinking the issue has been solved, where the issue could have been the material all along.
How??- I am only guessing, but if there was a slight oversize in the filament thickness, meaning drawing it through was harder at times when the print head was furthest away from the spool but easier when back near the gantry.
It’s only a far-out-there thought, but another test in the original black and measuring the filament thickness might be worth a look.
The primary reason I changed filament was that I have used up a lot of that filament when printing up the original attempt (20 hours) and my “test box” (5 1/2 hours).
You might be right that it’s a filament issue. So what I’ll do is try making another test using a completely different filament and see if it works ok. If so, then I’ll print another test in black.
Hi Rick.
I’d try turning the model 45 degrees on the build plate also if you ever get this again, to see if it happens on two sides or fades in as the plate moves forward/backwards.
New gcode always helps, too, I find.
Hi @rgmcleod52, did you ever do any further testing or was your issue resolved by the rebuild/maintenance you did? I am experiencing a similar issue, and like @Spanners alluded to, my hypothesis is that it is related to the Seam and potentially limitations of the smallest step the SM extruders can handle, rather than the filament or a potential wobble. I have done a more detailed write up of what I suspect which can be seen here if you are interested - Any advice on strange issue? - #8 by Waterboy, however this is only a theory and when I stumbled on your thread I thought to check in if it was resolved noting that it happened approx. a month ago.
I admit that I haven’t printed out a tall object like I was reporting on in the last month. But when I did make adjustments to the linear actuators the problem did go away. And I did not make any adjustments to the Z Seam. I could visually see the wobble prior to making the adjustments and then the smaller test pieces I tried did not show any of the roughness I previously experienced.
I have read about the linear rails that some users have attached to minimize the wobble problem (as documented here in Thingiverse: https://www.thingiverse.com/thing:4860729), but have not yet purchased these rails. That will probably be my next change.
I read about a new “Bracing Kit”, but it looks like it’s more suited for the new CNC module and didn’t look like it would minimize the wobble issue.
