@3dprinting Everything fits perfectly. The top half is printing now.
(Yes, that's a 3D printed stand-in for the controller board. Adafruit published an STL. It has 0402 components, and they came out more 0402-shaped than blob-shaped. But the connectors cracked when I removed the supports.)
I'm running out of ways to procrastinate crimping the cables. I do not enjoy crimping.
@3dprinting The next challenge was attaching leads to the LED strips. I glued this 144/meter RGBW strip into this aluminum channel. I'd printed end caps with embedded magnets (the aquamarine parts). I'd drilled holes through the table to run the leads down to the power supply (previous post). But the solder tabs on the strip are tiny, and I'd made no provision for strain relief.
This is what I came up with. I've assembled one, and it seems to be okay.
@3dprinting The cable tie will be inside the hole in the table when this is in place. There's room, because this assembly is smaller than the connector on the cable end.
I printed tall jaws for the StickVise to hold the aluminum channel, of course.
@3dprinting This strip will run vertically up the left side of the cabinet. Another will run up the right side and will be just the same. I don't know what I'm going to do for the strip across the ceiling, because there's no room on either end for anything. I'd assumed I'd run the leads straight up somehow, didn't really start thinking about how to attach the leads until yesterday. I've got 1.5 or 2 mm of space on each end, I think.
@3dprinting I never posted photos of the LED power supply after wiring. Here are some photos of the LED power supply after wiring.
Everything unplugs or unscrews except the wires on the power inlet module.
@3dprinting And here it is with the lid on. It looks remarkably unremarkable, just another power supply thing.
@3dprinting The LED controller and light bars are installed into the enclosure. FInally. Hooray!
I am very pleased with the way they came out, though I worry that the top bar is fragile. There's no strain relief on the cable; it's just hanging on by its solder joints.
I might tack the wires up to the underside of the table later, but for now, they're staying in place just by their own stiffness.
@3dprinting Here's a test pattern, viewed from the back. From the front and sides, you can't see the LEDs directly.
@3dprinting That test pattern uses golden ratio colors to generate apparently random but pleasing color sequences. It's a pretty cool technique which I just re-learned about. This 'blog post describes it pretty well.
https://martin.ankerl.com/2009/12/09/how-to-create-random-colors-programmatically/
Creating random colors is actually more difficult than it seems. The randomness itself is easy, but aesthetically pleasing randomness is more difficult. For a little project at work I needed to automatically generate multiple background colors with the following properties: Text over the colored background should be easily readable Colors should be very distinct The number of required colors is not initially known NaΓ―ve Approach The first and simplest approach is to create random colors by simply using a random number between [0, 256[ for the R, G, B values. I have created a little Ruby script to generate sample...
@3dprinting The LEDs have firmware now. I made a short video that puts them through their paces -- they have modes that match the LEDs on the MK4's front panel, but with more flair. I have no idea how to sync up with the printer's state yet. I'm assuming Octoprint will make that possible somehow.
@3dprinting Yesterday I _finally_ put the printer into the enclosure. After a few hours debugging a badly assembled power connector, it's up and running.
This enclosure is far from finished, but I was having trouble with air currents in my office ruining prints.
@3dprinting I decided it was time to redesign the googly eyes on the printer enclosure. (See upthread at June 19th.)
a) they don't wobble as intended
b) they stick out and get in the way of the door handles
c) they rotate freely in the round holes and the eyebrows are always at weird angles
d) I dropped one of the doors and broke an eye stalk.
𧡠60(?)/β
This time, the googly is a multi piece assembly that screws tightly onto the door. No wobbling, no protrusion, no rotation.
Note the place to insert an 8mm Allen wrench to hold the eye in alignment during assembly. The eye is held onto the orange bit with double sided Scotch tape.
𧡠61/β
@3dprinting I'm adding an (Un)original Prusa Drybox to my printer enclosure's lid, along with auto-rewinders. (Thanks to @GavinCampbell for the tip.)
The drybox uses lots of heat set inserts, so I completed a side quest to build an insert press. Now I'm back.
𧡠62/β
@3dprinting I'm trying to resist the urge to mod the drybox, but I'm putting the drybox on the printer's lid, and the lid tilts past 90Β°. So I need a way to keep the rails from falling out.
The original is in the first photo; my version is the second.
𧡠63/β
The autorewinder adapter for the (Un)original Prusa Drybox did not fit. The sleeves that surround the rail are too thick, and they keep the bottom cover from going on.
This isn't an issue for other users; they just let the rails ride higher in their slots. But I screwed my rails down.
So I added a shim under each rail. (It's teal in the photo.) It's temporary; I'll eventually redraw the sleeves so they fit.
𧡠64/β
@3dprinting Actually, the sleeve thickness is a problem for other people. The rails and therefore the filament spools, up to 5kg, are resting on the bottom panel, putting high stress on the screw holding it up.
𧡠65/β
@3dprinting The (Un)original Prusa Drybox is installed. I made feet, and the front feet engage little cleats. They're not exactly French cleats, but they're francophone.
𧡠66/β
@3dprinting The cleats hold the drybox on even when the lid is tilted.
But when I tilted the lid all the way up, just past vertical, the box fell off. So I shouldn't do that. I'd need bigger cleats or a different angle of engagement.
𧡠67/β
@3dprinting Here are some renders of the foot and cleat. The back feet don't have cleats; they just rest on the lid.
𧡠68/β
@3dprinting Now that I have an MMU, I am thinking about how to route the PTFE tubes from the drybox to the MMU. My original plan, before I saw how far back the MMU sits, was to route it through the lid. Something like the green line here.
But that wouldn't work. The cable is kinked way too tightly.
𧡠69/β
@3dprinting Prusa's solution looks like this. Five holes in the back window.
I don't like that because the window is no longer removable without unthreading the tubes. My intention is to make it possible to install/remove the windows while printing or remove the printer and drybox, separately, without tools.
But the routing is much cleaner.
(Photo by OndΕej StΕΓteskΓ½)
𧡠70/β
@3dprinting So I'm pondering exactly how to do it. If you have ideas, please let me know.
𧡠71/β
@3dprinting @Siff Maybe something like this? (Very rough draft.) The orange part would slide into the slot in the back panel and be glued in place. The green part would snap in, held by magnets, after the tubes were in place in the grooves at the bottom.
𧡠72/β
@3dprinting @Siff @esden Here's my idea for passing tubing into the drybox. Once again, the goal is fast easy disassembly so I can remove the printer to the workbench.
(sound on; it's narrated.)
𧡠73/β
@3dprinting @Siff @esden Here is the new back beam for the drybox, assembled. The tube holes are not parallel -- they will splay out inside the box and group together where they go through the enclosure wall.
I made the beam in three pieces. I used dowel pins as well as screws and inserts. The dowels aren't for strength so much as for straightness.
𧡠74/β
@3dprinting @Siff @esden I think I could post just this subassembly to Printables as an (Un)original Prusa Drybox remix. It seems generally useful for people with 5 filaments. (Prusa MMU and 5 head XL, for example.)
It's a minor annoyance that the (Un)original Prusa Drybox was made for six spools. Nobody can use six spools in the Prusa world.
𧡠75/β
@3dprinting @Siff @esden The dowels have a small effect on straightness. The parts were pretty good without them.
𧡠76/β
@3dprinting I've finally started filling the Gridfinity storage in the top of my enclosure. Here are some custom bins for the nozzle change tools and spare nozzles. Thanks to @Ronguest for the nozzle tube model.
I put my bins up on Printables, if you're interested.
https://www.printables.com/model/1070804-gridfinity-bin-for-prusa-mk4-nozzle-replacement-to
https://www.printables.com/model/1071020-gridfinity-bins-for-nextruder-nozzles
𧡠77/β
@3dprinting I continue to work slowly on my printer enclosure. Today I glued magnets into the Gridfinity base plates and modules. I printed new base plates in two colors a few days ago. These will be installed under the enclosure's lid as shown in 𧡠77.
𧡠78/β
@3dprinting Here's the current state of my printer enclosure's filament routing. It's good, but it's not quite right yet. I'm going to keep working on it.
The drybox is on top of the enclosure. Filament feeds into tubes at the front. The spools are on auto-rewinders, so there's no buffer for retracted filament.
https://www.printables.com/model/702217-mmu3-mmu2s-5-spools-auto-rewinder-adapter-for-unor
Here's a filament change.
𧡠79/β
@3dprinting The tubes exit the drybox in the back through the clips I showed upthread in 𧡠74. From there, they're free and floppy until they enter the enclosure through the back wall and immediately into the back of the MMU3.
𧡠80/β
@3dprinting Upthread in 𧡠69 through 72, I discussed options for passing through the back wall. @esden and @Siff advised.
The tubes go through a roller that lets them rotate together. The roller is clamped between the black surround and the black/orange panel.
For disassembly, the panel pops off, the roller lifts out and the whole back panel can be pulled away. The roller splits into two halves after the bearings are pulled off.
𧡠81/β
@3dprinting Here's how the roller comes apart. (There is voiceover in the audio.)
𧡠82/β
The good:
+ Filament is easy to load. The tube inlets are right up front, no need to reach behind the printer or feed through a buffer.
+ It works fine whether the enclosure lid is closed or tilted up at a 45Β° angle.
+ Everything disassembles pretty easily with no tools required.
𧡠83/β
The bad:
- Full spools have enough angular momentum on retractions that sometimes they pull the tube inlets off. Then the filament wraps around the axle.
- The tubes in the back are free and floppy and not parallel. (aesthetic)
- The back panel is flimsy. It's corrugated plastic, and it bends a lot when removed.
𧡠84/β
@3dprinting Tonight I finally put the printer enclosure where it goes, a mere ten months after I started designing it. It's no longer blocking the closet doors. The old Prusa MK3 enclosure is on the right and is inferior in every way.
I really need some LEDs in the drybox.
That's all the further the lid will open now. But the filament tubes look tidy with the lid up.
𧡠85/β
@3dprinting Upthread in 𧡠84 I said that sometimes the filament tube inlets come off. I've addressed that issue by creating bendable clips that hold the inlets in place. Tool-free, of course.
These photos also show the six-to-five adapter. The (Un)original Prusa Drybox is designed for six spools. The MMU3 uses five. So I made an adapter. (It's a refinement of a similar adapter seen here. https://www.printables.com/model/702217-mmu3-mmu2s-5-spools-auto-rewinder-adapter-for-unor/files)
𧡠86/β
@3dprinting I lifted the drybox up on stilts. While it was feeding filament to the running printer. Now there's 62mm of clearance under the box.
Upthread in 𧡠66 I showed how the drybox is removable. It still is, it's just closer to the stratosphere.
𧡠87/β
@3dprinting Now I have storage for various print bed sheets.
This is the Expandable Steel Sheet Holder by Whity, upside down because I'm left handed.
https://www.printables.com/model/51462
And the printer plays on.
𧡠88/β
@3dprinting 49 days ago, my printer's hotend fell out. I wrote about it here.
This week I finally repaired it -- travel and procrastination... The repair was straightforward; the hard part was cleaning off my workbench so I had room to work on it. π
https://chaos.social/@kbob/116063218020236097
𧡠89/β
@3dprinting While I was there, I decided to replace one of the auto-rewinder spools that had cracked in the last few months. And then I decided to redesign the spool holders. I'd always thought it should be easier to insert a spool.
So I came up with this design. The tall post behind the spool makes it so you can't (easily) push the spool past the slot. Just push it up against the post, then drop it into the slot.
𧡠90/β
@3dprinting The original took too long to assemble, with 20 screws. This one snaps together.
I made a compliant bushing to hold it onto the aluminum rails inside the drybox. I'm not sure that was an improvement. It does hold snugly, though.
𧡠91/β
@3dprinting Finally, I upgraded my MMU3's firmware. Prusa released MMU3 performance improvements last week; that's a nice bonus for a printer that's 4 generations old.
𧡠92/β
The MMU3 has been our workhorse for single-nozzle multi-material printing, and before it hands the baton to the upcoming INDX system, we wanted to give it one last significant upgrade. We usually design our hardware with some headroom, so it can...
@castaway PrusaSlicer generates an M115 G-code that tells the printer to check that its firmware is at least X.Y.Z. The recommended firmware level is in the printer configuration.
And then the printer firmware checks that the MMU FW version is compatible. That's not configurable or overridable -- they changed the comm protocol with this rev.
@castaway To answer your first question, I measured a filament change at about 5 second faster with the new firmware: 54 seconds instead of 59. Not much, but it adds up with a few hundred changes.
Prusa quotes a 9 second speedup; they're either measuring differently or measuring a different printer.
@castaway I don't think it times out. So yeah, plan to upgrade firmware when you update configurations.
The configuration updates seem to improve print quality a little each time, so I track them.
Kernel Bob 
Jess Robinson