Upgrading to ARM and 0.9 Steppers - Vertical Banding Issue

[xnaron]

I've been trying to find a cure for this vertical banding issue as pictured here:

[img]http://imageshack.com/a/img30/8264/t4l1.JPG[/img]
[img]http://imageshack.com/a/img39/7821/fim.JPG[/img]

It's fairly faint but you can feel it with your finger.

My Setup:
I am running a RostockMax with my Mag Mod Effector. I am using a 3.5mm Jhead with a geared stepper similar to what makergear used on their printers. I use this extruder and jhead on my other cartesian printers without issue. I am running Rambo for the controller and Repetier 0.91 for the firmware. This issue has occurred in all versions of the repetier firmware I have tried. I am using the last release of KISS as the slicer.

What have I tried to diagnose the cause:
- I have printed the object at a 30 deg angle to the X axis and the print looks the same.
- printed from the SD card.
- I have tried KISS and CURA slicers with the same results.
- I have tried different print speeds. It appears that the banding becomes more frequent when the speed is lowered. I lowered it to 30mm/s and have run it up to 100mm/s.
- I have rechecked and adjusted all of the DELTA geometry settings in the firmware.
- I have tried tensioning the belts and loosening them with no effect to the banding.
- I have tried different acceleration and jerk setting.

Has anyone ever seen banding like this before? Anyone have any ideas? I have been printing for a long time now on a lot of different printers and I am running out of things to try here. I have been thinking about getting an Arduino Due with a Ramps-FD and run the ARM version of repetier.

thanks,
Brendin

[Batteau62]

I've noticed this before. I don't have an answer. My theory is something in the g-code. To be as repeatable as they seem, something has to be exerting control. To get thicker then thinner extrusion, something has to be varying the z height, Yes? So, as the head travels, a command says "stay at this height" If an error happens i.e. the head raises or lowers. The command tries to re-establish the correct height. So a wave form is created. And each succesive layer repeats according to the individual mechanics of the printer Just my two cents

[Jimustanguitar]

First off, those are really clean looking prints. I'd be pleased to consistently get that level of quality.

Interesting that there's a diagonal pattern as well as the vertical pattern.



I have no clue what to look at, but I agree that it looks like a software thing instead of a mechanical thing. Here are my stabs in the dark:


If you've got mag arms, I have a hard time believing that there's slop or play in them... is there any chance that there's grit or a flat spot in one of the joints? Does removing the platform and rotating it 1/3 make any difference?

Are you doing external perimeters first or last? Try slicing it the other way as a test.

Are all of your prints this way? Have you tried a different STL file? How about repairing that one with netfabb? Maybe that's the way the triangles are in the source file and you're actually getting an accurate print.

[Batteau62]

No that I can't seem to stop thinking about this It occurs to me that it might be "die swell" http://en.wikipedia.org/wiki/Die_swell" onclick="window.open(this.href);return false; I have a Filastruder, and this phenomenon has a lot of bearing on filament diameter.

[DavidF]

I just had a print with the same thing in it, funny thing is it was the same color filament as yours. I doubt that has anything to do with it, but it is a unique problem... seems more like the way it slices than a problem....

[DavidF]

its a pretty big part, but here is the same thing...
[img]http://i1102.photobucket.com/albums/g446/davidflowers1/Mobile%20Uploads/utf-8BSU1BRzA1NzcuanBn.jpg[/img]

[mhackney]

Any chance that this is an artifact from the model's STL and maybe slicing? Try enlarging the item 1.5x in the KISS when you slice it and see if that changes the frequency of the bands or not. If it does, and they are 1.5x further apart, then we have somewhere to look. If not, then we can probably safely rule out STL or slicing artifacts.

[xnaron]

thanks for the replies!

Yes my printer is mechanically sound and I do not have any play. My layer alignment is pretty much perfect. I am very happy with the prints. Details and pics here http://forum.seemecnc.com/viewtopic.php?f=54&t=1704" onclick="window.open(this.href);return false;

[img]http://imageshack.com/a/img23/7470/dsc01261s.jpg[/img]

Going from memory the problem may be more pronounced with my 0.35mm nozzle vs the 0.5mm nozzle.

I should mention that I have tried printing this out on my cartesian printers and it is very very faint and you have to look really hard to detect it. For some reason it is a lot more pronounced on the delta. You can feel it on the delta vs not on the cartesian.

DavidF yes I get the same thing on arcs and circles as you pictured.

The stl file is pretty simple. I am attaching it here. I print it in Vase mode with 2 loops to make it hollow. Looking at the resulting slice in KISS and gcode in Repetier host it doesn't look like there is any funny business going on.

[neurascenic]

Xnaron,

I am throwing this out there from a totally naive perspective...

Can you print a tube about the same size and height?
I am wondering if there is a slight error (rounding?? )in translating the Cartesian to Delta (Polar?).
Maybe one of those artifacts will go away in the round.

[xnaron]

Xnaron,

I am throwing this out there from a totally naive perspective...

Can you print a tube about the same size and height?
I am wondering if there is a slight error (rounding?? )in translating the Cartesian to Delta (Polar?).
Maybe one of those artifacts will go away in the round.

I should have mentioned that I have printed a cylinder and the banding is present there as well.

[DavidF]

so mine is just something to do with the shape of the part?? Here is the orientation it was printed, yes I see that it is duplicating the shape of the arcs...different issue??
[img]http://i1102.photobucket.com/albums/g446/davidflowers1/Mobile%20Uploads/utf-8BSU1BRzA1NzguanBn.jpg[/img]

[mhackney]

I ran that STL through KISS and looked at the gcode in a gcode viewer to see if anything popped out. This viewer is nice in that you can trace the print one line of code at a time. In this case, each side is a printed as a single gcode command and not broken up into individual segments like I thought it might be. However, I do think the scaling test would be worth checking - gives us more data.

[xnaron]

I ran that STL through KISS and looked at the gcode in a gcode viewer to see if anything popped out. This viewer is nice in that you can trace the print one line of code at a time. In this case, each side is a printed as a single gcode command and not broken up into individual segments like I thought it might be. However, I do think the scaling test would be worth checking - gives us more data.

I agree. I will try it.

[mhackney]

I'll make a prediction now that I've looked at the gcode from the STL. "The vertical lines will have the same frequency".

I now don't think this is an artifact of the software - at least up to the firmware. Tell us your slicing parameters and I'll slice in KISS and try it too.

thanks,
Michael

[xnaron]

I'll make a prediction now that I've looked at the gcode from the STL. "The vertical lines will have the same frequency".

I now don't think this is an artifact of the software - at least up to the firmware. Tell us your slicing parameters and I'll slice in KISS and try it too.

thanks,
Michael

0.4 skin thikness
2 loops
1 stacked/spares...
0 inset

infille : Vase

Extrusion width 0.355
Infill Width 0.355
layer thickness 0.2
loops go from inside to perimete checked
wipe not checked
de-string checked
1.0 seam hiding
Jitter 0

speed 50mm/s for all and set to Fast

[xnaron]

here is a pic of one at 1x and 1.5x

[img]http://imageshack.com/a/img546/6164/drto.jpg[/img]

[mhackney]

That sure looks like the spacing has not increased but there are more lines.

[geneb]

Brendin, if you're still using those magnetic arms, could you try putting a little bit of lithium grease into the rod ends? Not so much to reduce friction, but to reduce micro-vibrations coming from the stepper motors.

I see similar diagonal striping on Blue MAX - it uses TL's carbon fiber arms.

g.

[neurascenic]

Another naive possibility.

Slight Temperature Variations. Though, I would admittedly expect something more random.

[mhackney]

This is quick test. I used a clear PLA. What filament are you using?

[img]http://mhackney.zenfolio.com/img/s5/v13 ... 8747-3.jpg[/img]

Although you can "see" vertical bands, I can't feel these and I don't see any surface irregularities under a 10x loop. I'm using an E3D hot end with .4mm nozzle. Not sure if this is the same thing you are seeing since you can feel yours. But it is interesting that almost all of the translucent PLA colors I print show this type of pattern and none of the ABS opaque filaments do. I wonder if it is some sort of optical effect I'm seeing that's different than yours.

[xnaron]

This is quick test. I used a clear PLA. What filament are you using?

[img]http://mhackney.zenfolio.com/img/s5/v13 ... 8747-3.jpg[/img]

Although you can "see" vertical bands, I can't feel these and I don't see any surface irregularities under a 10x loop. I'm using an E3D hot end with .4mm nozzle. Not sure if this is the same thing you are seeing since you can feel yours. But it is interesting that almost all of the translucent PLA colors I print show this type of pattern and none of the ABS opaque filaments do. I wonder if it is some sort of optical effect I'm seeing that's different than yours.

That looks exactly like what I am seeing. I can barely feel them. I have seen it with silver reprapper.com and silver ultimachine 1.7mm filaments. Are you running a 0.35mm nozzle? Which firmware are you running?

[mhackney]

.4mm nozzle on E3D. This is repeater .80

[xnaron]

.4mm nozzle on E3D. This is repeater .80

Yes I have found it through all versions or repetier from 0.8 to 0.91. I want to try Marlin again but need to build a working config as the one I have is really old. I am pretty sure it did it on Marlin but don't want to say for sure until I confirm again.

[DavidF]

Only thing I can think of, try rotating the part 60 degrees and reprinting, might help in ruling out the towers and arms.

[xnaron]

Only thing I can think of, try rotating the part 60 degrees and reprinting, might help in ruling out the towers and arms.

I've tried that. The banding is the same.