I'm throwing this out that as a long shot, as I'm not sure if anyone can help me, but maybe there's someone out there who had a similar issue and might have successfully dealt with this before.
I'm dealing with this:
Granted, if I take a picture of the same benchy at a "flattering" angle it looks fine but the prints in real life don't look great.
Setup and attempts to fix the problem:
- CR-10S.. -ish, modded to a point where I don't think calling it a CR-10S is right.
- SKR 1.3 board with TMC2208 drivers in UART mode, StealthChop enabled.
- Tried changing the current from 600 to 900 for all motors. The only difference was motors running cool/hot and slightly different motor noises, print quality remained exactly the same.
- Now running Klipper but flashed back to Marlin for testing at one point. Print quality is improved with Klipper, but the same core issue persists on both firmwares, leading me to believe this is probably hardware (mechanically) related. Tried running Klipper with input shaping and pressure advance both off and on, no difference in terms of the issue I'm having.
- Sliced with Simplify3D but also tried with PrusaSlicer using the MK3S quality profile with just a few changes like startup G-code and bed size, fairly similar results with both, mainly this issue is present on both slicers
- BLTouch, though the bed is manually leveled to the best of my abilities anyway.
- The first layer is beautiful and smooth.
- E-steps calibrated and they remain consistent throughout multiple repeated tests both on Klipper and Marlin.
- X and Y belt tensioners (metal), played around with barely enough tension to keep the belt from slipping up to a highish note when plucking them - literally no difference in print quality.
- Replaced most of the V-slot rollers, yet I didn't see any signs of wear/damage on the old ones
- All rollers are engaging with the extrusions but aren't too tight, i.e. they can be manually rotated in place with a bit of force but won't rotate freely in place.
- Squared up the frame to the best of my abilities
- Leadscrews are aligned dead straight to the motors in the X- and Y-axis.
- Tried 3 different couplers - rigid (stock), flexible (the aluminum "spiral" ones), and the ones with the hard rubber in the middle. The ones with the rubber made a slight improvement but didn't fix the issue. Pictures above were taken with those couplers installed.
- Anti-backlash T nuts, tried loosening/tightening them - no difference. Leadscrews don't bind when the T-nuts are tightened though.
- I had a closed-loop belt syncing the leadscrews together. I took them off and maybe there was a slight improvement, but the issue is still there. Photos were taken with the belt off.
I'm currently using a DD setup with a pancake motor, BMG extruder (Trianglelabs V2), V6 all-metal (Trianglelabs). Tried switching back to Bowden - fairly similar results in terms of the problem I'm having, the quality itself is a bit different compared to DD, DD being more. "tight" I guess would be the word in terms of quality, but the layer stacking is still similar. Tried using a metal "stock" style extruder on Bowden - the overall quality was just trash compared to DD and the issue still remained. Tried using a PTFE lined hotend instead of all metal on the DD setup - surprisingly the print quality overall was worse with it compared to the metal throat, so that leads me to believe the Trianglelabs throat is of decent quality and isn't the source of my problems like with cheap metal throats.
Designed my own DD mount in hopes that not having the stepper motor hanging far away from the extrusion would perhaps result in less hotend "wobble" during faster moves. Didn't help at all, although I do like it and it's more rigid compared to another popular one I was using before.
Tried multiple slicer settings, these are the ones I'm using now for reference
- 500 mm/s acceleration and ~40 mm/s base print speed (conservative speeds for testing)
- 1 mm retraction (tried 0.8 mm that results in slight stringing up to 1.5 mm that results in under extrusion)
- 40 mm/s retraction (tried 20 mm/s to 90 mm/s - no difference)
- No Z-hop
- Wipe 2 mm (tried on and off - no difference)
- Extrusion multiplier at 0.98, tried from 0.9 to 1.1. There was a slight reduction in the layer stacking issue with the multiplier at ~0.93-0.95, but it results in under extrusion in the top layers. Tried to combat that by having 6 top layers but there still were gaps. Everything over 1.05 results in over extrusion to a point where the top surfaces are just scared.
- Coasting off with Klipper due to using pressure advance, though with Marlin and coasting enabled the same issue persists.
- Printing PLA at 205 °C. Tried 190-215 °C - no difference.
- Tried printing with the heated bed off and a layer of hairspray for adhesion and as normal at 60 °C. The quality is somewhat better without the heated bed for the first few layers as expected, but overall same issue persists.
I'm a bit lost on what to try next, I think I've tried every combination I could think of in terms of settings, tightness, etc. My only thought is maybe it's in some way related to the leadscrews as they are the only thing I haven't changed. Although I did flip them over AND swap sides yet that made no appreciable difference which makes me second guess myself if my leadscrew theory is correct.
Maybe I'm missing something obvious.
I'll be delivering a crap ton of benchies to an orphanage or something, hopefully, the kids will have some fun with the stacks and stacks of benchies I've been printing for the past week.
Here's an XYZ cube:
I'm not seeing any diagonal lines in any lighting.
