odd issue with print, jumping from place to place

[Aurora900]

So I would search the forums first... but I'm just not sure what to search for this...

I'm having an issue on one model I'm trying to print. It was pretty much flawless until the last layer where the printer started making some very (seemingly) unnecessary motions. Instead of printing one contiguous line, it started printing little bits of the last layer here and there and coming back to finish them after doing other parts.

[img]http://staticexperiment.com/rostock/DSCN0243s.jpg[/img]

You can see in the image here, where instead of printing one line around that triangle shaped feature along the edge it started doing it in rows and skipping around from one side of the print to the next. Because of this, that feature is completely messed up instead of being smooth. What I'm trying to figure out is what caused this and how I can prevent it. Is it a slicer issue? A model issue? I don't know where to start.

I'm using Repeteir host, slic3r, and a Rostock Max v2. The model came from thingiverse here: http://www.thingiverse.com/thing:372451 and slic3r told me that it wasn't manifold or whatever so I ran it through the netfabb repair thing it said to do. I'm also wondering if maybe the repair caused this?

Also what on earth is the deal with this manifold business and how does it actually affect the print? And if that's a problem for slicing, then why do so many models have this issue? (Nearly everything I downloaded today gave that warning) (I tried searching for this too but didn't find much)

Any help would be much appreciated!

Thanks,
Danielle

[Polygonhell]

Manifold just means the shapes are closed and do not overlap.
You can't just make the circle and stick the triangles on/in it, you have to merge the geometry into a single piece.
A lot of models built with Sketchup tend to be none manifold, slicers do unpredictable things with the models.

[Aurora900]

Oh wait, I see why it did it... Once again I blatantly miss the obvious... It's because that part is supposed to be ridged, but every time the print head leaves it makes a mess... so is there any way to get the edges of small details like that to stay in place better?

[bot]

X/Y resolution is generally poor with a large nozzle. Try shrinking the extrusion width setting, as well as the layer height setting. Also, try scaling the model UP, at least in the Z direction. Looks like a stargate wheel?

[Aurora900]

X/Y resolution is generally poor with a large nozzle. Try shrinking the extrusion width setting, as well as the layer height setting. Also, try scaling the model UP, at least in the Z direction. Looks like a stargate wheel?

Yeah its a stargate... I'll try those things on the next print. Also, what actually is the highest resolution we can go for the layer height? I never saw that spec listed anywhere for the rostock. I know the orion does 50micron layers, but they don't have it listed for the rostock...

[bot]

I saw 100 micron layers advertised for the rostock somewhere (the MAX), but any FDM printer can be tuned to do below 100, though you'd likely encounter diminishing returns below 100 microns with the stock 500 micron nozzle.

[Polygonhell]

Slic3rs default extrusion width calculation is utter crap in my opinion, if you really want to use Slic3r, I would at least manually set the extrusion widths to your nozzle size or slightly larger (0.55 for a 0.5 nozzle).
Or you could just use Cura or KissSlicer, both will give you better prints.

[Jimustanguitar]

Looks like a stargate wheel?

Yeah its a stargate...

Better get it right. Who knows where you would have ended up!



For layer height, my buddy Phil has gone down to .07 mm layers (70 microns, right?). The Z positioning accuracy is the same as the XY accuracy, it takes all 3 motors to move in a straight line in any direction. For each motor, the stock setup is accurate(ish) to .0125mm per microstep (thanks BDJohns!). There isn't a 'hard' steps per e value to quote for motion though because of all of the trigonometry involved with Delta machine movement.

Try it and see! Tell us how fine you can go down on your layers. I suspect that you'd run out of resolution on the direct drive EZStruder before you'd toe the line on positioning accuracy. With a gear reduction extruder, you may run into the limit of the plastic and its properties first. Hard to say.

[Aurora900]

Looks like a stargate wheel?

Yeah its a stargate...

Better get it right. Who knows where you would have ended up!



For layer height, my buddy Phil has gone down to .07 mm layers (70 microns, right?). The Z positioning accuracy is the same as the XY accuracy, it takes all 3 motors to move in a straight line in any direction. For each motor, the stock setup is accurate(ish) to .0125mm per microstep (thanks BDJohns!). There isn't a 'hard' steps per e value to quote for motion though because of all of the trigonometry involved with Delta machine movement.

Try it and see! Tell us how fine you can go down on your layers. I suspect that you'd run out of resolution on the direct drive EZStruder before you'd toe the line on positioning accuracy. With a gear reduction extruder, you may run into the limit of the plastic and its properties first. Hard to say.

Good to know. I'd love to play with super small layer heights, but I'm finding that even with 100micron layers I'd be waiting at least a day for a print to finish haha. So I'm probably sticking to mostly 200 unless I need crazy detail.

I guess what I need right now is a crash course on what parameters affect the print in what ways and how to combine them to get the best results for each situation... it's pretty daunting and I'm slowly getting the hang of this stuff, but there's a lot I still don't understand. This is going to turn into a hell of a hobby

What's the best place to look to learn about that kinda stuff?