Vertical Banding Revisited

[Right to Make]

Did you specify the flavor of the smoothie you were getting her?

[mhackney]

This has been my experience with Duet as well. BUT you must also consider your STL file. A curve in an STL is also made up of short line segments (from it's inclosing triangle) and consequently, the # of triangles you use to generate the STL has a big influence on the length of these segments. Many STLs you download will have pretty poor resolution in that respect. For my work, I use the most triangles RhinoCAD will allow me to generate.

[marcom]

I am experiencing similar vertical banding on my prints, which seems to be directly dependent on the segmentation in Repetier (Changing the parameter DELTA_SEGMENTS_PER_SECOND directly changes the frequency of the pattern - I'll post some pics).

I am curious if upgrading to the Duet with the dc42 firmware is sufficient to eliminate this altogether.

FWIW, it seems like Duets are out of stock both at Filastruder and Think3dPrint3d so it looks like I'll have to wait, unless someone has any alternative sources.

Thanks!

[bot]

I haven't done a conversion of a Rostock max, so I can't say for sure, but my gut tells me yes. It can also be helped by going to 16 tooth pulleys and 0.9d steppers.

Geneb recently did a conversion of a max. I wonder if he can comment on segmentation moire.

[marcom]

FWIW, here are my pictures of the thing I was printing under different conditions (an iphone holder).

I was slicing in MatterControl and Simplify3D. The banding gets tighter and more pronounced as you go higher, likely because the print slowed down as the layers got smaller. I also did a print on my Prusa i3 for comparison (less even layers but no banding).

The most annoying thing about the banding is that it is print-speed dependent, which makes the back non-uniform in prints with varying speed. I suppose I could keep the print speed the same to see if it gives me a more uniform pattern (at the expense of too-short layer times at the top or a slower overall print) - just wish I didn't have to do that.

Side view

Back view

I figured I would add this data point in case it is helpful. I suppose I'll bite the bullet and get the Duet, if that's what solves this issue.

[marcom]

I haven't done a conversion of a Rostock max, so I can't say for sure, but my gut tells me yes. It can also be helped by going to 16 tooth pulleys and 0.9d steppers.

I suppose that's my question - if the dc42 firmware on the Duet doesn't segment, will I _ALSO_ have to go to the 16 tooth pulleys and the 0.9d steppers, or is the controller upgrade enough to eliminate this particular issue?

I am just trying to determine bang/buck, and given that I just finished building the machine, I don't want to take it all apart and replace all the key components if I don't have to (at least not yet ).

[bot]

If you plan on doing it anyway, do it first without switching the motors and pulleys. You'll know pretty well right away if it disappears, and if it does, you'll be the guy who brings this data point to light! Win win! I didn't have the patience to do any kinds of comparisons with my duet. It worked so well off the bat in my custom machine that I've just been printing since.

[marcom]

If you plan on doing it anyway, do it first without switching the motors and pulleys. You'll know pretty well right away if it disappears, and if it does, you'll be the guy who brings this data point to light! Win win!

Haha - great point! Now if only they could get some Duets back in stock. I've put an order in at Filastruder and will update when I receive it and carry out the test. Thanks!

[dsiegfried]

By having a regular pattern like this it causes the vertical lines, because there is some overlap between the infill and the outer walls. Even if you have 3 or more outer walls, those lines still show through.
Try randomizing your infill angle at every layer. It was suggested to me to set your infill to rotate by a prime number like 31 degrees every layer. It helps keep artifacts from showing up on the sides of your print.

[DavidF]

OK, so what happens if you combine the .9 steppers, 16t pulleys and the duet?? Im ready to make a move on something.....My friend just sent me prints off his massportal an they have totally killed anything I have ever seen printed. So now I need an upgrade!!

[marcom]

Well, this arrived today:

So here is my "before" picture:

And I'll report once I get the Duet installed and configured. Hope it's not a mess - I might be posting on the Duet board for help!

[geneb]

YOOOOU CAAAAN DOOOOOO EEEEEEEET!!!!!!11ONEONEONE.



g.

[marcom]

And now the moment we've all been waiting for (or at least I have)...

Duet installed, calibrated, and...

Drum roll...

Left: Rambo w Repetier, Right: Duet w DC42

Same GCODE file. No question - the banding is entirely due to the delta segmentation. I can die happy now.

This is a MAX V2 with stock EZ Struder, stock hotend with 0.35mm nozzle, stock motors (with Astrosyn dampers).
Printed on black PLA.

And now a picture of my printer looking like it's barfing a controller board:

I'm going to do some additional calibration and cleanup, and if there is interest I can post my config.g file.

[3D-Print]

Awesome job....... and amazing the difference. When I get up and running with my duet I will reprint this as well.

[Eaglezsoar]

We can always use a new config.g file to add to our collection of different setups.

[bot]

Great work! Now it's finally settled. DC42 is our saviour.

[KAS]

Looks good.


what's this rumor that a new Duet version is coming out?

[gchristopher]

And now the moment we've all been waiting for (or at least I have)...
....
Same GCODE file. No question - the banding is entirely due to the delta segmentation. I can die happy now.

Wow... That's compelling.

[DavidF]

Just throwing this up there. Rostock V1 only mod is the tricklaser cf arms. I have not re-calibrated in over 2 years even after installing the new arms, just reset Z height a couple of times....Oh and I hot glued my effector plate together LOL (with good reason) This post if for my own musings as I plan on upgrading to the duet as well.....[img]http://i1102.photobucket.com/albums/g446/davidflowers1/Mobile%20Uploads/20160305_213214.jpg[/img]

[bob64]

Wow. I guess it's time for me to upgrade to a duet board. Pity we can't use the old screen/knob anymore - that knob was handy for calibrating z-height at the start of a print.

[geneb]

Bob, if you install mhacney's FSR mod at the same time you do the Duet upgrade, you'll never need to set your z height again.

g.

[bob64]

Yeah, except my bed constantly expands or contracts. 40 minutes into a print the bed has dropped another .2mm or so. It feels like it takes forever to reach thermal equilibrium.... and I think the stock firmware auto-off's the heaters after like 30 minutes?

[mhackney]

Bob, I print well over 6hrs a day on each of several deltas. They all have different bed heaters and bed systems but they all have FSRs and Duet controllers. The Rostock with the snowflake heats up fast (24V power supply on an SSR). I think you'll find when you remove the unnecessary constraints on the bed attachment, that the expansion/contraction buckling will all but go away. The issue is that the stock bed uses 6 screws to securely fastest the bed to the top of the printer. That is over constrained. The FSRs allow the bed to float on 3 points and any expansion and contraction is relieved without buckling.

[Jimustanguitar]

Yeah, except my bed constantly expands or contracts. 40 minutes into a print the bed has dropped another .2mm or so. It feels like it takes forever to reach thermal equilibrium.... and I think the stock firmware auto-off's the heaters after like 30 minutes?

You're not experiencing this oddity, are you?
http://repables.com/r/726/

[bob64]

Yeah, except my bed constantly expands or contracts. 40 minutes into a print the bed has dropped another .2mm or so. It feels like it takes forever to reach thermal equilibrium.... and I think the stock firmware auto-off's the heaters after like 30 minutes?

You're not experiencing this oddity, are you?
http://repables.com/r/726/

Doubt it, when performing the 24v mod, I basically plugged that hole full of wires to carry the extra current.

I guess I'll try removing some screws from the bed to try and alleviate some of the buckling.

Edit: Turns out my entire top base panel has warped along with the snowflake. !@#