First Layer Inconsistencies
Hello, I've been printing a bunch of parts today in several batches of the exact same gcode, no changes in settings environmental conditions etc... Yet, I've noticed significant difference of first layer quality between batches. My guess is it depends on how well the initial bed probing goes, which then influences the entire print - sometimes it's spot on, sometimes it's not... at all.
- printer: MK4
- nozzle: 0.80mm
- filament: Prusament PETG
- print setting: "0.40mm QUALITY" with very few changes (disabled bridges, reduced elephant's foot compensation)
- filament setting: "Prusament PETG" with reduced nozzle temperatures by 20C as I was getting a lot of stringing with the defaults
Here's a picture of parts from two consecutive batches:
I've been looking through slicer options, printer settings... but I haven't found anything that'd help with this. Is there something akin to MK3S+' Z-adjust on the MK4? Any tips would be greatly appreciated! Thanks!
RE: First Layer Inconsistencies
If the two batches use the same g-code file, then there cannot be a setting in PrusaSlicer causing this.
Then it has to be something on the printer itself. Most likely a hardware-thing
You already mentioned variations in Z-offset caused by Mesh-leveling. In fact the Z-offset is then on the edge, so a small variation gives big effects-> Redo Z-offset
What about:
- enclosure/room temperature? Differences can give different outcome.
- different spools of filament? You might have to redo the Z-offset, pressure advance and/or flow rate settings.
- Moisture in the air?
- Cleaning of the bed?
We will do what we have always done. We will find hope in the impossible.
RE: First Layer Inconsistencies
With the 0.8mm nozzle there is a higher chance of oozing that could affect the bed leveling. I would look to see if there is anything on the tip of the nozzle before the bed leveling starts, and wipe it off if there is.
RE: First Layer Inconsistencies
Thank you for the reply! I'll have a look at your suggestions, mainly pressure advance and flow rate, as I haven't dug into those so far. I like your point about the same gcode producing different results effectively eliminating the cause from being in the slicer settings. Sure simplifies the search! 🙂
I tried re-running Z calibration earlier, however, the issue would occur still in roughly 1 out of 5 batches. About the same prior re-calibration.
I do have an enclosure, but haven't used it for these prints. I don't think the temperature changed in a significant way between batches (large room, open window), although I didn't make measurements to verify this. I suppose in the future I'll try running these in the enclosure to rule out temperature swings as the cause.
All batches printed from the same spool, which, prior to printing, was in a vacuum sealed bag with a bunch of desiccant packets.
I clean the bed after each batch in the sink with dish soap and the soft side of a standard kitchen sponge. I always dry it thoroughly with a wipe that doesn't leave fibers and finally let it sit for ~15 mins to let any remaining moisture evaporate. Not using a glue stick for these prints, they adhere well enough even with the suboptimal first layer.
RE:
I've done some test prints today to verify the ideas you guys have given me. I placed 5 test prints on the bed like the number 5 on a die ⚄ to rule out any out-of-square axis problems, thankfully the results were consistent across the buildplate. For some tests, I simulated bad leveling by placing a few sheets of paper on the bed prior to probing, later removing them for printing, introducing an offset.
- oozing/leftover filament on the nozzle: this turns out to be the likely culprit, or at least a contributing factor. I was cleaning the nozzle before, but I haven't been super thorough. When I intentionally left filament on the nozzle, the first layer suffers in quality more often, but certainly not always - the nozzle cleaning routine at the start does get rid of the excess in most cases.
- flow rate: increasing this did help close the gaps in first layer better, but the part comes out very messy - there's more stringing and globs of plastic get stuck on sharper corners.
What I found helped the most:
- Double, triple and quadruple check the nozzle is clean. Then check it again.
- For the 0.8mm nozzle, in slicer [Print Settings > Advanced > First Layer] was defaulted to 1, resulting in a very rough fill pattern. Reducing this to 0.8 helped a lot.
- Unwind ~1 loop of filament before starting the print to avoid the spool potentially pulling on the extruder, throwing off probe measurements.
Here's a picture of parts printed today - first without any adjustments, the other 5 applying the above:
Still noticing very minor gaps around the hexagonal hole, but I suppose that's beyond the precision achievable with the large nozzle. I do have a 0.40mm one laying around if I ever need the extra precision. For now, speedy prints it is! 🙂
Thanks for all the suggestions, they helped a ton! I'll post again if I figure out any additional adjustments to improve the first layer.