GregValiant 1,455
I have a good rant on single wall "calibration" cubes around here someplace. The bottom line is that if you calibrate for a single wall model you will be able to print perfect single wall models. On all other models the flow will be wrong. How many actual single wall models do you print?
Calibrate your E-steps. Scale a calibration cube to 100 x 100 x 1mm tall and print it. The skins are the best indicator for flow and let your own eyeballs see what's going on. You can tweak the flow using the LCD as it prints. It will be darn close to correct at 100% flow. I calibrated my Ender for PLA and I have found that PETG is better at 105%.
(BTW I picked up a microscope at a garage sale. It's really good for inspecting prints but a magnifying glass ain't bad.)
Regarding cooling.
As you've no doubt found, PETG is prone to warping. Certain shapes are worse than others. If you have a section that is long, narrow, and fairly tall then you know it's going to lift. I often design elephant ears into my prints to help hold them down and I always use a fair coating of hair spray as an adhesion promoter because I've found it's a must for PETG on the glass bed. There is a plugin in the Cura MarketPlace called "Tab Anti-Warping" that you can use on corners of your print to help keep them from lifting.
The bed is hot and so there are often problems with curling near the bed. Once you get up higher you can start to use some fan. At what layer and how much fan depends on your own system. I designed a single 5015 system for my Ender and it's pretty good so about 40% works well for PETG. I also wrote myself a post processor to add cooling profiles to gcodes. I check the preview in Cura and then build a profile that turns the fan on and off at specific layer heights. That works quite well but it wouldn't work for just the bow of the Benchy.
This print is a shark fin in PETG. It's long, narrow, and tall. Even with the tabs it pulled their middle area up almost a millimeter. (It's paused here as to slide a 4mm x 225mm steel rod down a hole on the inside to keep it from warping. I mounted these on the roof of our SUV so we could find it in parking lots. The sun is tough in Florida.)
Here you can see how hard it was pulling. I used a simple cooling profile on this as only the tip at the top really needed it.
The cooling profile thing is part of this modestly named app - Greg's SD Print Tool which you only get if you throw away those single wall calibration cubes.
I went into another rant and almost forgot one of your questions.
When Cura does a slice there can be line segments in the range of microns. Each of those line segments takes a line of gcode to print it. That's the resolution we're talking about rather than the overall resolution of the model (which is a function of the STL, 3mf, OBJ, etc. file). So the Maximum Resolution needs to be tuned for your printer so it can run smoothly around circles and curves while keeping up with all the calculations it needs to make. The longer a line is the more time the printer/planner has to calculate the next move. At some point we need to remember that we're talking about molten plastic here so whether an extrusion is .00001mm long or .1mm long are you going to notice a difference? You will if it stutters and leaves blobs while it's trying to process all those .00001 long lines.
Edited by GregValiant
Recommended Posts
YALE70 1
The maximum resolution is set at 0.25 right now. Out of curiosity, are there any cases where you'd want this to be set low? Maybe for a high detail print? Or is one value typically good for most applications?
Regarding the Benchy; I did initially have the corner preference setting on smart hiding for a previous Benchy and it generated that nasty Z-seam that ran about halfway up the bow (see pic) along with the kind of haphazard placement of the seam along the stack. Hide seam made the one I posted above a lot better looking visually like you said, since the seam is now along one of the back corners but I still got that warping in the bottom of the bow. I will say, having run temp towers on this specific PETG; it loves to curl on thin overhangs regardless of what temp I print it at. I'm still a bit perplexed on what's considered ideal cooling for this stuff. Everybody seems to have a different take: no fan, a little fan, or full fans. Of course it all depends on the exact filament and fan setup, but what's considered generally correct? Is 40% fans for PETG too much? Too little?
The flow calibration was done per Teaching Tech's guide with the vase cube, and because I wanted to improve dimensional accuracy. However, like cooling PETG, I see a lot of mixed opinions on if messing with flow is a necessary/good thing or not. I had it initially set to about ~72% (I was getting roughly ~0.52mm on a 0.4mm line width setting) but I was having issues with my walls not sticking to each other, so I gradually bumped it up to 80% to try and alleviate that. The E-steps were calibrated since I replaced the hotend though this time around I did it with the nozzle removed and it was only slightly off from what I had it set to previously. That said, I probably shouldn't be doing it with the whole Sharpie + ruler thing if I'm going for utmost accuracy.
Link to post
Share on other sites