Last time I posted (just over a year ago … ugh), it wasn’t about the AD Project, but about an enclosure for my printer.
This post will summarize what’s been going on with that and a few miscellaneous projects I’ve engaged since last time. I’ll make a post later on that catches up with the AD Project.
Printer Stuff
The enclosure has been built and in operation for the last year-ish, but I didn’t use my own design for the extensions that attach the top table to the bottom, as I’d originally planned. Instead, shortly after I printed my first test part, I discovered zuspiel’s Tinker-Friendly Lack Enclosure. This design allows you to quickly remove the top half, if needed, which seemed like a good idea. (Edit: Mikolas Zuza posted Prusa Research’s interpretation of the Lack enclosure only three days before this post was made. Like much of what they do, it offers several excellent refinements. I am of course referring to the second enclosure at that link.)
Because I’m a cheapskate poor, my enclosure doesn’t have pretty acrylic panels like others often have. The enclosure uses white foam board (the kind kids use in school for science fairs and such), which makes it look somewhat like a low budget version of the renders in A Safe Space for Printers. The foam board also theoretically blocks light from the printer if I’m trying to sleep. I say theoretically, because the only light in there right now is a pair of LEDs on the print head. But I do plan to add some RGB lights to the interior of the enclosure. I just monitor my prints via webcam.
The hinges on the door were originally “temporarily” made of simple blue painters tape, but I finally got sick of them coming loose and replaced them with proper brass hinges last month. Tape also once formed a loop to serve as the door latch (held closed by a pair of tweezers through the top) and is still what holds the other three sides on.
I have and was using Webberen’s Filament Spool Holder/Nest for awhile, which worked great (esp. after I put bearing retainers on each axle), but I built a filament dry box a few weeks back. Complete with indicating orange dessicant. The amount of room behind the hole that allows filament into the top of the enclosure is a few inches shorter than the dry box really needs (means the filament has to curve sharply down from where it exits the dry box, which increases resistance a bit, even with PTFE tubing)
In the process of building the enclosure, I learned about and performed the MOSFET mod to eliminate a design flaw in the printer that could’ve caused damage to its control board, or even a fire. I still need to sort out strain relief for the heater bed wires behind the build plate.
If you recall from my last post (doubtful), I mentioned the desire to add more Z height to my printer. A quirk of my printer is that when I installed the Z braces, I cut the threaded rods too long, such that they extended an inch or two over the top of the frame.
In order to do the z height mod or really even put the thing in the enclosure, I had to cut the threaded rods back to the correct length, and that meant removing them. Bit of a pain, but it got done. Once the enclosure was ready, the printer was reassembled and calibrated, and some apparent clogging issues were quickly sorted.
Some time later, trouble struck. I had printed some flexible tires and PLA wheels for Artie. I soon noticed that they seemed to wobble when they rolled. Weird, but maybe a one-off issue.
Due to the limited space under Artie and the fragility of printed parts, I was also considering using a Hobson’s Coupler instead of gears to bend the motive power around a 90 degree angle. I printed some tall, round parts in order to put together a demo of the mechanism, and eventually I realized that they wobbled as well. In fact all of my prints were skewed. Specifically, they printed leaning toward the negative Y axis as they got taller.
Uh-oh. This would make it difficult to do any sort of drive system for Artie. This problem plagued me for months — a lot longer than it should have — and caused work on Artie to stall. I periodically racked my brain for possible causes: belt loosening, pulleys slipping, etc. Nothing seemed to help, so I pretty much stopped printing functional parts for awhile.
I asked about the issue in the printing department of one of the Atlanta Microcenters while returning from Dragon Con. They had no suggestions, but my brain suddenly kicked into gear, and the answer smacked me in the face: the Z-axis gantry (which was held rigid by the Z braces that I’d had to remove and cut) wasn’t square to the rest of the printer. It leaned slightly forward. This should’ve been obvious, given the symptoms. The reason that never occurred to me is that I had squared the gantry when reassembling the printer, and I hadn’t noticed the skewing problem right away. I couldn’t make the connection that the problem had started after removing the braces.
I am not what is called a civilized man, Professor…
Speaking of Dragon Con, you may recall from the Tale of the Third Robot (back in “Rogue Done”) that the Alternate History Track had planned a Splendid Teapot Race for 2016 that had to be cancelled due to lack of entries (but we still did exhibition races). 2017 was therefore slated to become the first official, non-cancelled Splendid Teapot Race at Dragon Con.
Rather than sticking with the Teapot Pepperpot Mk I (or building Mk II), I elected to go a different route. A more steampunk route. And what could be more quintessentially steampunk to today’s crowd than Jules Verne’s most famous creation?
Thus construction began on Nautealus (emphasized Nau-TEA-lus) … the world’s first tea-powered, land-traveling submarine. The ship’s motto, naturally, is Mobilis in Mobiltea.
Essentially, I took an R/C truck from Walmart and replaced the body with a chibi version of Harper Goff’s classic Nautilus design, with an integrated teapot. I modeled the hull in CAD so that it would fit over the R/C chassis. I didn’t really get a chance to properly design the front mounting point, so it was sort of haphazardly super-glued to the chassis there and had to be re-glued before the race.
Because of the size of the ship and the previously-mentioned skewing issue, each section of hull (forward, center, and aft) was split down the center line and each half was oriented on the printer such that the skew would be consistent and unnoticeable. Basically the outboard sides of the ship are slightly lower than they should be, but this is masked by the natural downward taper of the upper surfaces of the hull as you move outboard. The only place the skew is really apparent is the lid of the teapot, and then only if it’s viewed straight down the center line, or set on a flat surface.
The Nautealus was printed in Inland PLA, mostly black. The bridge was given a tiny golden ship’s wheel and was lit by a gently pulsing green LED. The dome shaped windows on the bridge were made by cutting the back off of a pair of googly eyes and removing the innards. Just as I did with the Teapot Pepperpot Mk I, I gave the ship an integrated Arduino Nano and Catalex MP3 board that selected random sound effects and dialogue from the movie (the mostly spherical speaker was hidden inside the teapot), and also powered the LED.
It was a bit awkward to turn everything on and off — the speaker, R/C system, and Arduino are all independent of one another. The Arduino and its battery (one of my Crayola-styled external phone batteries) must be accessed through the salon windows, which are simple press-fit plugs.
After priming the hull, I painted it silver, added panel lines, and then employed Modern Masters Metal Effects rust paint. This is oxidizing paint … it contains real iron particles and an oxidizing agent that causes it to rust at an accelerated rate. After the paint rusted, the hull was gloss coated, giving it a “wet” look (otherwise it looks very orange and dry).
I learned that painting technique from this video by Bob Martin (RCSubGuy), one of the folks that builds actual R/C models of the Nautilus. I didn’t have my airbrush out of storage at the time, so I had to modify the technique to do without that. My hull didn’t turn out as nice as Bob’s paint jobs, but he’s done lots of them and this was my very first attempt.
In the Splendid Teapot Race, the Nautealus took second place. The prize was a grab bag of tea and tea-related snacks and accessories.
After the race, the Nautealus excelled at injudicious use of her signature maneuver … a high speed charge with the piercing ram on the bow. No, I didn’t mistake the other contestants for warships and slavers, I just helped destroy the race course’s obstacles (which were made of foam board and/or cardboard). Though I did hit one a obstacle bit too hard, which broke loose the super-glue holding the bow down. Apart from that, the ship survived perfectly unscathed.
This isn’t the droid you’re looking for
In Episode VIII: The Last Redeye, I talked about working on a Treadwell droid.
I didn’t take Artie to Dragon Con 2017, and I missed having him. So I spent a lot of time hanging around the droids at the convention, especially the BB-8 units. If you aren’t aware, the holoprojector lenses used on those droids are typically made of dyed resin, cast in jewelry cabochon molds. It occurred to me that I could print a master for Treadwell’s eyes, make a silicone mold from it, and then cast lenses in clear resin. With enough sanding and polishing, the resin copies should serve just as well as glass lenses.
So that’s what I did. I haven’t perfected the process, but I’ve gotten a couple of passable attempts out of it. You can see examples and more detail starting around post 22 over at the Treadwell build log, which gets updated almost as infrequently as this web log does.
The Spooky Side
Right after working on those lenses, I had to switch my brain into Halloween mode, as usual. My big project for 2017 was getting my PVC fence going.
The full plan for the fence consists of nine 8′ sections … two along the left driveway, five along the sidewalk, and then two along the right driveway (we have a driveway on either end of the house with the front yard between them). The four sections that are along the driveways also have to be inclined at something like a 5-10 degree angle to match the yard). Each section has three horizontal beams (one at the bottom, two at the top). There are I believe nine posts per section, with a larger post between, and even larger posts in each corner.
Most people buy plastic finials to top off each PVC pipe (vertical post), but not me. The reason is that, with the purchased finials, you are fitting a round peg in a square hole, and they are not intended to fit each other. If the base of the finial is large enough to accept the PVC pipe, then the finial looks too large in proportion to the post. If the base is too small, then the ends of every pipe would have to be squared off to fit. In total, there are 81 finials (9 posts per section times 9 sections of fence), and constructing the thing at all was already a lot of manual labor, so I preferred to just print something that is sized to fit exactly onto the post.
Still, 81 is a lot of finials to print. It takes my printer more than 6 hours to print 3 finials at the appropriate size, so that’s over a week of non-stop printing. It also costs a little more than buying them (depending on your filament brand and electricity costs), but the effort saved and the improved aesthetics were worth it to me.
I also designed some finials that you can’t buy, to go on top of the larger pipes between the sections of fence. These are shaped like bats. They have red reflectors in the eyes to give a hint of eeriness when cars drive past. LEDs would probably work better, but would require power.
Phase I, which I completed for 2017, was to assemble the five sections bordering the sidewalk. I’ve put off doing the other sides until Phase II, because five sections was plenty to get done in only a few weeks, and because I still needed to sort out how to construct the side sections so they matched the incline of the yard. I’m thinking that where the vertical PVC pipes pass through the wooden horizontal beams, the holes need to be more like slots with the pipe centered and free to rotate to different angles. This turns that entire section of fence into a sort of overly complicated parallelogram.
The fence sections were painted with a couple of coats of black latex paint (some were primed, but I quickly tired of doing that), and the finials were put on. Then I broke out my old pal, the Modern Masters rust paint. BOOM! Rusty wrought iron fencing, perfect for any haunted mansion.
I want to build a static prop of Sam from Trick R’ Treat at some point, so I started gathering stuff last year. While I was doing that, I used the same printed master/silicone mold/resin cast process I’d used for Treadwell lenses to make Sam’s Jack O Lantern lollipop. Results were mixed — mostly because my mold was dodgy and I didn’t use enough dye — but it’s passable under low light.
dyne3d 
