Large Scale Central

CR30 3DPrintMill

Dan Gilchrist said:

Steve said:

Considering the use of blank laser printer water-slide decal paper. Anyone have experience

with that media?

I print my own water slide decals with an old HP LaserJet and decal film. I also have a white toner cartridge so I can do white letters, etc. Finding suitable quality graphics (or creating my own) can be a challenge as I am not a Graphic artist.



What brands of decal film have you had good success with?

For that matter, are there any brands you would suggest avoiding?

Hi everyone. I have gone down a hole printing 3d printer accessories and tricking out
my printers software wise.

Both now have ball-bearing spool rollers, Pi 4B Octoprint controllers, and various other accessories.
Got around to completing the CR10 heat cabinet properly:
Picture 2021-10-19 12-18-48

I mentioned OctoPrint in my last post. For anyone unfamiliar with it I highly
recommend you give it a test run:

It has an amazing set of plugins for customization:

As it runs via your browser you can control /monitor your prints from any browser-enabled
device on you local network - I sometimes use my ipad in the living room while watching tv…

It can control a webcam, allowing you to observe the print, or take a time-lapse video of it:

Since it runs on a dedicated Pi you can turn off your main computer once you’ve downloaded the gcode
without disturbing the print process.

Installing is easy, grab the SD card image here and flash to a card. Attach an HDMI monitor and USB keyboard and powerup. It runs a setup wizard for you. Then run raspi-config to setup ssh, etc.

In post #6 I described printing a 1 piece 24" ramp, and mentioned that the Kadee
coupler hook was catching the ramp bottom. Today I spent some time analyzing
it and realized it was because it was a Kadee 906 G-scale (1:22.5) coupler,
not meant for One-scale (1:32) cars. A while back I did an experiment with a
half-dozen different models of Kadee couplers to determine the best for use
with the MTH cars. I settled on either a G-scale 906 or a One-scale 1906. The 1906
look better (proper scale) but seemed less reliable on bumpy track due to the smaller
contact area. The 906 looks “too big”, but was more reliable. Anyways, I put a 1906
on the battery car and and the hook clears the ramp with room to spare!

If I can figure out how to mount the 906 at its proper height it would also clear the
ramp (I think…). This would require some cutting on the existing MTH cars, but
obviously would not be an issue with the cars I am going to print.

SO - one more thing to ponder…

Have been sidetracked doing “getting ready for a long, cold Winter” stuff. Cut/split/stacked six cords of wood, equipment serviced and back in barn, etc. etc. etc.

Started back on the cars now. I can spend the Winter feeding logs to the fire while I print Pullmans :smile:

Trying to dial in the eSun PLA+ filament. First half-dozen tries were dismal:
Picture 2021-11-04 11-43-59
Picture 2021-11-04 13-42-05

But late last nite I got a very nice test print:
Picture 2021-11-04 22-02-27

This is a 4" section of the middle window panel. Going to try the top & bottom panels next and see how they paint/assemble.

So while I’m waiting for the 3rd (top) panel to finish printing I thought I should review the slicer programs for the CrealityBelt.

Kiri:Moto is considered the best, but is cloud-based so out for me…
IdeaMaker was suggested by my vendor, used it some.
BlackBelt Cura wouldn’t work for me.
CrealityBelt Cura for windoz.

I had multiple failures with IdeaMaker, shown in the last post. Unfortunately Creality hasn’t ported CrealityBelt to linux yet, so I’m stuck with running the windoz version under Wine - it crashes every now and then. But when it stays up it creates an .stl that prints pretty well. I’ve tried running it on my windoz laptop, but it doesn’t size properly on the attached DisplayPort monitor, being almost unusable. When run on the laptop screen its scaled too small for my old eyes to see. I might try another monitor to see if it works any better…

Here are the 3 pieces, top (lettering), middle (window) and bottom (will eventually be ribbed) panels after touching up the edges:
Picture 2021-11-05 13-06-18

And after applying several coats of Rustoleum gray primer:
Picture 2021-11-05 14-13-16

I’ll lightly surface sand with 1000 grit in a few hours and then topcoat with red and orange.

So I didn’t do enough surface prep:
Picture 2021-11-06 14-03-07

My plan B is to print the side as 1 piece, here (on the right) with heavy primer:
Picture 2021-11-06 14-16-37

I’ll sand with an orbital sander till the lines are completely gone, then I expect, another light coat of primer.

Some insight, please…

I realized I could use a sheet of flat plastic stock for the Pullman bottoms, only printing the truck mount posts and then attaching them to the flat stock. These could be printed on the flatbed printer (CR10), eliminating more than 25% of the belt printer work - a BIG savings!

Browsing on Amazon I found possible selections of ABS, PVC and Acrylic sheets. Keeping in mind that I will be gluing printed PLA+ sides and ends to it, which would be most appropriate?

I could also use something entirely different such as aluminum/metal or even wood. When converting the deadrail boxcar I discovered a sheet of metal sandwiched between the boxcar floor and the bottom body. The plastic portions would have been more than rigid enough, especially with the boxcar body screwed on. I concluded it was for weight - to keep the boxcar from bouncing off the tracks.
The Pullmans have no such metal plate, but being 2-1/2 times longer they are already heavier. Anyone have any insight on this? If its for weight perhaps I should use metal sheet for the bottom?

The Daylight build portion of this thread is now here

I added a linear rail to the extruder rail:

In this photo the rail is installed and the old wheeled carriage is shown lying on the bed.
No discernible difference in performance, I suspect the rail will remain in spec much
longer than the wheels would have, can already see some wear on their bearing surfaces.

I added a pair of upgraded motors, new on the left, stock on the right:

They are larger, so they supposedly run cooler. But more importantly they have support bearings on the top end of the shaft. You could literally see a tiny amount of bend on the stock shaft - NOT GOOD -

At first I had the belts replaced incorrectly and it couldn’t home. An email to RepKord with a pic of the setup and Alan responded within 15-20 minutes with what he suspected was the problem - it was.
You can’t beat service like that!

Been back at trying to dial in my eSun PLA+ In post #25 I reviewed the slicers I was using. I’ve been revisiting that issue as I’ve never gotten a great print with eSun PLA+

The wine/windoz version of Cura (top) and linux Blackbelt Cura (bottom) both have issues with the Z axis edges:

So I fired up the linux version of IdeaMaker (middle print) and did a sample of where I left off before abandoning it:

Note the irregular/rough top surface. I could never steer it out…

Then I imported the NAK3D printer profile for the CR30 from the IdeaMaker site, along with their generic PLA filament profile. I also imported an eSun PLA+ specific profile from the IdeaMaker site. These prints are both with the NAK3D printer profile. The top print is with the generic NAK3D PLA profile, the bottom with the eSun specific profile:

Note that the Z-axis edge are now almost perfect :smile:

The top surface is a little rougher, will tackle that next. There is also a very slight hump where the window frame ends (middle right):

I’m hoping a raft will remove this. :thinking:

So the raft didn’t help:

The panel separated from the raft - disaster!

Tried it again without raft, additional top layer, and 40% infill. After sanding:

A look at the inside. 4 top layers and 40% infill:

It is quite strong:

I’m bending the thin part above the window opening.

I was able to bend it about 90 degrees before it broke:

Now that I can reproduce clean prints its time to deal with finish work.
I got a bottle of:

First I printed a fresh panel:

Then a quick coat of this resin:

Then a quick pass with 220 grit sandpaper on an orbital sander:

Two coats of the Rustoleum orange under the USB microscope:
The new part is on the left, the old sample (primer/sand/paint) on the left.

Here is a view with a normal camera:

With the naked eye you really can’t see any lines :smile:

If I were to take my time and sand with 400 then 800, then primer & sand, and then airbrush with genuine SP Daylight Orange I suspect it will look quite nice :smile: :smile:

Next I should create a panel with ribs. I won’t be able to sand those, so I’m hoping I can refine the resin process to make those look OK…

My first attempt at a ribbed panel:

I’ve since modified it to have the alternating large/small rib:

That one is in the printer now…

The test on the left had large/small ribs and was treated with 3D ABS resin for smoothing.
The test in the middle had large/small ribs and was treated with XTC30.
The test on the right had large ribs only and was treated with 3D ABS resin.

The gloss of the Rustoleum makes comparison tough, especially in webcam pictures.
The small ribs are almost beyond the ability of this printer, but refinement of the smoothing process should make that panel look better than the “large rib only” version.

Tomorrow I will work on smoothing with filling primer to compare results.

To that end I added the “Continuous Print” plugin to OctoPrint to automate things:

It can now print them faster than I can ruin them :anguished:

Looks VERY nice Steve - as you have learned, 3D is not nearly as easy as it looks - nor as it should be…

I tried printing the ribbed panel test at a right angle. It looks MUCH better :smile:

The test on the left is the original method, the right is the right angle method:

They are placed on top of an MTH Daylight Pullman for reference. Note how the
FDM lines are almost completely absent with the right angle method.

So here’s the fly in the ointment: printed in this orientation I’m limited to about 8.25" length.
Which means I would have to print and join 4 sections for a 28" long Pullman. I could still
print the top and window panels full length as they have no ribs and can be sanded to eliminate
the lines.

How should I design the joints to keep them from showing?

  • butt joint?
  • slight gap on face to allow use of bondo or other filler?
  • ???

Why not make the joint overlapping?