MakerGear Forum

Hi,

I've been struggling a bit with uneven extrusions, especially on small tall objects. I've attached a couple of photos that show the problem, but keep in mind that these are taken through a 20x microscope, so it doesn't look quite so bad in real life. The first photo is at 0.15mm layer width, the second at 0.25.

I have a MakerGear M2 from 2012/2013. Most parts are original, but I've replaced the original idlers with the new improved ones. I've tensioned the belts, tightened the platform spider screws and adjusted z height bolt, adjusted filament extruder tension, verified wall thickness and adjusted the slicer's multiplier, made sure the hotend isn't loose, switched to an unopened roll of PLA, replaced the y motor, updated firmware, tried different temperatures, done a cold pull to clean out the old hotend, ... I think that sums it up. The uneven extrusion doesn't seem to be specific to either the X or Y direction, and I don't see a pattern in the Z direction either.

So at this point, my suspect is the hotend itself. I have a new hotend ready, but I wanted to try to fix this issue first. Is there any theory whereby an old worn hotend might cause something like this? Or should I be looking elsewhere for a fix?

Other hypotheses to note:
1) my extruder motor gets very hot. What digipot setting should I be using? Maybe it's causing the extruder mount to get soft... I've lowered it a smidgen, no help so far.
2) I haven't tried a _brand new_ roll of PLA. I opened a new sealed package with silica gel in it, but that package has been sitting here for a while already. Could old or humid PLA be causing this?

ketil wrote: ↑

Fri May 21, 2021 4:36 pm

especially on small tall objects

Those towers need much more cooling to let each layer solidify before the nozzle makes another pass. Building several copies will help, probably more than any additional air flow from a fan, although you'll likely need a brim around the bases to improve adhesion.

Slowing down so each layer takes a specified minimum time sometimes helps, but for very small objects it just hovers a smokin' hot brass nozzle directly over the thing: no way to cool off, no matter how much air is involved.

Examples of having it come out right;

https://softsolder.com/2015/04/03/epson ... ot-repair/

https://softsolder.com/2011/05/27/thing ... erfection/

Mad props for taking care of all the other gotchas!

Hey Ed,

Thanks for your input once more, getting your skilled eye on an issue is much appreciated!

However, I don't think cooling is the issue either. I'm already going fairly slow, but I've tried printing this piece again, now 4 at the same time. This time I also rotated the four parts at 45, 135, -135 and -45 degrees about the Z axis, for the heck of it. I've tried at both 0.25 and 0.15mm layer height, and the result looks the same as printing just one. The part itself looks a bit like a standing ladder, so when I printed 4 of them, there are 8 towers being printed. The stl: https://build.openflexure.org/openflexu ... g_test.stl

I'm attaching a new photo of the 0.15mm version, it looks very similar to my earlier photo. There's clearly room for improvement regarding the precision.

ketil wrote: ↑

Tue May 25, 2021 5:51 pm

already going fairly slow

Going slow is at least part of the problem: the nozzle spends too much time over each tower and the layers get all melty.

There being nothing like a good new problem to take one's mind off all one's old problems, here's a version in PETG: That took about 15 min with my usual 0.25 mm layers and has no obvious problems, other than crappy bridging that's only to be expected with no cooling. If it weren't a test piece, I'd flip it 90° to lie flat and ignore the horrible stairsteps on the topside ramps.

Thanks for testing! The part you've printed looks acceptable for the purpose of a flexure, by my understanding. It should be printed this direction to make sure it works as intended in the openflexure microscope. Some sagging is acceptable.

I've used this part to exemplify the uneven walls I get in my prints, but my experience is that I have some trouble with larger prints as well. When printing parts that are meant to fit together, they are usually too tight, or stuck, and I need to use designs with greater tolerance.

I rummaged around in an old box with gear, and came across a spare unused makergear hotend from 2012, with a 0.25mm nozzle! I switched, and I'm getting exactly the same issues, so I can hereby confirm that the hotend clearly wasn't the cause, so I'll have to continue debugging this. I'm thinking now that there must be just a tiny bit of play in the nozzle. Those jagged walls are driving me insane

Have you cleaned and lubricated your rails? Also, check your belt tension.