Things have been really busy the past few months, but I’m finally getting back to working on the 3D printer.
I’ve been splicing various premade models designed for previous generations of the Monoprice Select Mini into one that allows for a better hotend fan and a part cooling fan.
Next up is properly setting up the wiring, manual switch, and integrating it into the wire bundle for the hot-end. Then recalibration.
I’ve been working on a few things crafting-wise lately, none of them too major at the moment.
Minis-Painting:
I decided to paint a few minis again. I’m hoping to print some minis for a game of Shadowrun soon, so while I’ve been tinkering with the printer I’ve been painting a few small things. A couple of minis that wouldn’t be out of place in-genre, and the Mound of Dead Bards from Gamers: Dorkness Rising. That last one took quite a bit of work.
3D printer tinkering:
I’ve so far never managed to get the Monoprice Select Mini Pro to print minis as nicely as the other printer. I’ve been tinkering with some settings, and also with the hardware. Main thing I want to change is to improve the part cooling. I’ve gone through a couple iterations of air vent modifications and swapping in a better fan, but I really think it needs 2 fans. Unfortunately all the designs I see that include 2 fans are for different configurations or models, so I’m having to do some model splicing and editing. The main issue I come across is that the bed levelling/z axis sensor is mounted in a specific location not present on earlier versions of the MPSM, so I’m having to carefully splice that mount in in a way that is compatible with the air mods. I’m also doing a bit of experimentation with freehand modifications with the 3d printing pen, but the improvement has not been significant.
Bag of holding repair/upgrade:
I’ve had a bag of holding from Thinkgeek for years. Used it in my college days until I overloaded it with weight and damaged it beyond my ability to repair at the time. I’ve carried it with me from place to place, and occasionally used it (with some safety pins over a busted zipper or two) for thematic appropriateness or something “close enough”.
Recently I came across the bag again, and decided to work on it, inspired by finally going to a tabletop RPG night again. So far I’ve repaired the failed zipper, reinforced all the zippers, and added pull tabs on the zippers to make them open and close more easily. I repaired the fabric where the stitching had been coming loose, such as on the zippers and the strap where it attaches. Since I’m gaming in a high traffic makerspace, and I recall these bags being somewhat common, I’ve been adding patches to make it more identifiable. Most recently I decided to finally test out tge chemical NeverWet, and a “magic bag” seemed an appropriate use of it
I’m currently waiting on the neverwet to dry. I was able to use an old pressure fit curtain rod to hole the bag between two posts for spraying. Thankfully the fumes died down after a while. They were pretty bad, and got throughout my home, even though I sprayed outside.
So, it’s been quite a while since I’ve posted, and at this rate I’m not sure I’ll be finishing a detailed description of the design and construction process anytime soon, so I’m gonna skip ahead a bit. If there’s enough interest in explaining the portions that I’ve skipped past (or I make myself take the time) I may circle back and explain.
Anyway, here we go. Through a lot of tinkering with 3D models, learning how to perform photogrammetry, and some hands on work, I finally finished the polterbuster, so now I have my completed costume.
I got the hand unit to hang off the side of the pack on the custom made mounting brackets with some magnets. Haven’t yet gotten the straps to handle the pack being unbalanced with the unit on it yet.
And now here’s everything put together. It was finished just in time for Halloween… when there was nowhere to go. Oh well, there’s always the next convention!
Maybe one day I’ll revisit this and get some decent sound effects added to it.
The clear compartment on the original vacuum contained a lot of white foam balls that got blown around by a fan when the toy was turned on. I decided that I was going to install lights here. The Poltergust from the games had one light… but this is an imprecise mashup piece. I decided to go with 4 lights, in homage to the proton packs, but green to match the aesthetic of Luigi’s Mansion.
Here’s what I was working with:
I removed the foam balls, and originally I was also going to chuck the fan in the trash too, but I hold onto parts like that just in case, and I’m glad I did. This will be important for later.
For the lights I ordered some 10mm green LEDs (I like my tech chonky, especially for a cartoon character), and the accompanying resistors.
I 3D modelled and printed a piece to hold them into place in the spacing I wanted, and after a lot of frustration and soldering I got them into the installed. I put a solderless connector on the end so that I can separate parts when troubleshooting and/or if I decided to mount the board on the other half of the casing. I really need to spend some more time practicing crimping those connections correctly.
To control the lights for a blinking pattern, I used a knockoff arduino nano and modified some simple code for it. Gotta watch out for those knockoff boards, though. I had to go find special drivers so I could use the board, which can be nervewracking. You never know what people are going to include in code from an obscure website.
To power the board, I decided to use a phone power bank, as I hoped it would be an easy solution… I should know better.
Hooking the board to the phone bank powered the lights…. for less than a minute. Then the power-saving features on the bank determined there wasn’t actually enough of a load on the circuit, and cut out. After a lot of fiddling with resistors, I added a bundle of 4 resistors in parallel to the circuit to add a little bit of load, which makes the power bank stay on.
Once I had verified that the power bank would stay on, I went to permanently install the components. I continued cannibalizing the cable I was using, and wired the resistor into the circuit. I also added the toy’s original power button into the circuit as it would be a conveniently accessible button already built into the casing.
Of course, the one segment of the wire I HADN’T messed with turned out to have a short. I had to cannibalize another wire end to bypass it. (Not shown, just annoying)
Then I had to figure out how to mount everything so that it would stay in place and function reliably. I went by my old standby for mounting things: hook-and-loop-backed command strips.
I used the modified USB cable to connect the two halves together for closing up. I also added an extension cable for charging the power bank without opening up the case, and a cut a hole for checking the display of the power bank to determine the charge status.
With that all wrapped up and functioning, here’s the light sequence. Pay no attention to the other changes for now.
I’ve been doing some training to improve my 3D modelling skills with Autodesk Fusion 360 (not sponsored, that’s just what I use) lately, and in one of the videos the instructor mentioned that he happened to like using a 3D mouse to rotate the models and help his workflow be just that bit easier.
Around the same time, someone happened to drop this link in the cyberdeck discord chat.
https://www.instructables.com/Space-Mouse-With-Arduino-Micro-Fully-Printable/
It’s the instructions for a homemade spaced mouse with 3d printed casing. It uses a small arduino board and some inputs (buttons and a clickable joystick) to make a space mouse.
Hmm… I can build a space mouse for a fraction of the cost of one I could buy? I get a new device for my workflow AND a project to practice a bit more with electronics and code?
Well, that’s a no-brainer. Project ahoy!
So, I started following these instructions, and built my space mouse.
Side note: every time I use the term space mouse I think of Mickey in a spacesuit. I don’t know if anyone else has that issue, but I just had to share.
I don’t think I’m gonna get too in depth covering it, as that feels redundant with the instructable above. I will point out a few things form my experience though.
This project made me very glad that the arduino boards came in a multi pack. I botched the first attempt HARD.
I put pins along the entire length of the soldering locations. Where are the wires supposed to solder to if you’ve put pins in all the holes? WHERE???
Turns out I wasn’t reading the instructions closely enough.
In addition, I was attempting to solder things while the board was embedded in the case. That didn’t work to well either. There was nowhere for the iron to really fit. While I was at it, it reminded me that I was using a very cheap soldering iron that I had only bought for something to do heatset inserts with, which wasn’t as precise or controllable as I needed.
So, with all that, I decided to make a few changes to my process and equipment.
1. I reprinted the case. I’d… kinda melted the board into the original case already and couldn’t get it out.
2. I carefully reread the instructions, found a diagram of the pinout for the board, and made myself a detailed diagram to work from before soldering ANYTHING.
I only needed a few of the pins that came with the board to make things to mount the board to the case, not the entire row.
3. I used thinner, more flexible wires. The originals were waaaaay too inflexible to fit where I needed them.
4. I did all my soldering away from the plastic, then carefully installed the parts in. No unintentionally melted plastic.
5. I improved my soldering equipment and process. I made liberal use of solder flux (I’d shied away from it in the past), added a fume extractor to avoid having to work with the door open, and got a much better soldering iron after conferring with other makers about their recommendations. It was my third iron after all, it was time to get something that would work well. Finally got myself a Hakko.
Also took a little bit of time to do some reading/viewing so I’d take better care of this iron and have better soldering results.
This was a series that helped me, by the way. Sometimes you gotta go back to basics… and realize your bad habits that you’ve gotta fix.
6. For connecting to the joystick, I used solderless connections to save my sanity, since it came with connector pins on the stick already.
Anyway, after all that on the hardware side, I finally got the thing built. All the wires soldered, the thing assembled and closed up.
I used black for the optional button extensions because I thought it’d be easier to use with contrast.
I ran into an issue with the thing slipping around on my glass desk, so I pulled out the rubber tape that I used for my arch lamp to make this nonskid as well.
Now it doesn’t move anywhere unless I deliberately pull it off the desk.
Once I got the thing assembled and plugged up to the computer, I loaded the code included with the instructable.
Buuuuut…
The code was written primarily for this thing to work with Autodesk Inventor not Autodesk Fusion 360. The shortcuts are different in those two pieces of software. It took a bit of reading (particularly looking up the shortcuts for Fusion, figuring out the Keyboard. h code, and the firmware itself), but I eventually got at least the ability to orbit (rotate the view around the models) and zoom to work.
… also kinda had to swap the X and Y axes in the firmware. I might have gotten those two pins swapped on install. It works now, though!
Once I’ve used it a bit more, and thought more on what commands I use frequently, I’ll have to take the time to reprogram the buttons to do something useful.
Anyway, I got a new toy, got some more experience on these kinds of electronics projects, and had a reason to finally upgrade my electronics setup to be something more useable. Win!
Note: Yes, I know that isn’t the common spelling for useable, but usable doesn’t look right to me, and IT’S A VALID SPELLING, DAMMIT.
[END OF LINE]
My 3D printer has never been great at printing large scale objects, so I had to go through addressing everything I could in order to prep for the VirtCon. Now that that is over, and I’ve implemented a couple more changes, I’ve decided to share what I did.
The biggest issue that I’d been having was adhesion over large areas and warping. I attacked this from a few different angles.
Always-On Cooling Fan:
My custom-added cooling fan that is so necessary for upper layers of parts can cause issues on that crucial first layer. I didn’t have a way to control the fan through software, so I added a hardware method. I cut into the wires for the fan, and added a power switch in the loop with a switch I already had, and 3d modelled and printed a casing for it.
I included screw holes in the 3D model in case I wanted to screw it down, but I decided to blu-tack it to the side of the Ikea table for now. It seems to hold well enough for now.
This has definitely helped, but the drawback is that I have to remember to leave it off at the beginning of the prints and flip it on a couple layers into the print, otherwise the print won’t adhere on the first layer or the upper layers won’t cool quickly enough and have a variety of issues.
Drafts:
Based on the placement of the printer near vents, and some online discussions, I thought that drafts of air could be a contributing factor. At first I tested it with foam-core board cut, taped, and hot-glued into a shape. This helped me figure out what form factor I wanted to use.
This material is flimsy, and I would think particularly prone to fire and/or melting. I decided to make a better box with some stronger materials (hardboard and metal hardware). This would double as practice with hardboard, which I had never used before but planned to use in the Flying V. I thought I could knock it out as a one day build.
HAHAHAHAHAHAHAHAHAHAHAHAHA
No. It was day 5 before I got it to the state I wanted it in. Partly because of finding out that I needed more parts and kept having to order them or go pick them up. LEARNING POINT: Take the time to do a better estimate what you need and then order double.
In keeping with something I was trying at the time, I kept trying something simpler to try out techniques before moving to more complex/difficult projects. I was building the box before building the guitar. Before building the box, I built the box that went on the side of the big box.
Anyway, after a lot of work, I built the big box.
The hook holds the box front door open when I need to see get in there for a while. The top door is for filament reel swaps. The remote is for the LED lighting strip I installed in there (I got fed up having to repeatedly get a flashlight out). The extension box on the left is to provide the space for the end of the printing tram and cabling. The side handles are for ease of moving the box to perform maintenance.
Since there was a box in the way, I had to move the camera inside the box to continue monitoring the prints.
Notes for future development on this box: I still intend (when the weather is better) to partially disassemble the box (mostly just take the lights out), and paint at least the inside (but probably also the outside) with a fire-resistant spray paint. This will help ease my paranoia of 3D printer fires while also minimizing fiber shedding on the inside of the box. Also need to fill the gaps at the corners with something.
Bed Surface:
Blue tape has been my go-to for years. It’s easy to set up, and works pretty well for my smaller prints. However, with larger prints the tape often works well enough to adhere to… at first. Then warping as the object gets taller tends to pull the object off the tape or even pull the tape up off the bed.
I tried bed-weld, but on this aluminum plate on my printer, it does not work well.
People recommended hairspray, but without a removable bed surface that would eventually kill my printer, as it would overspray onto the moving components that need to move with the best lubrication.
So, I finally tried one of the other options that had been recommended to me.
Purple glue stick! This has been amusing to play with, and seems to work. I did have to add a water spray bottle, a new scraper, some higher percentage isopropyl alcohol, and some microfiber cloths to my repertoire, but it made the project possible!
This seemed to make the biggest impact on the warping issue.
Calibration:
I was introduced to this website for 3D printer calibration:
https://teachingtechyt.github.io/calibration.html
I can’t recommend it enough! It guides you through a comprehensive calibration process, moving through common issues in a logical order.
It also helped me to be able to massively increase my 3D print speed. If I hadn’t run that process… I don’t know if I would have been able to print my parts in time. My printer still has the 12 hour timeout issue on Octoprint, and some of these parts did get up to 9+ hours even WITH the speeds turned up. I’m so glad the tests helped me realize I could crank up the speed for this project.
Other upgrades:
Today I’ve been doing some more work on the printer to incorporate some changes.
One thing that had been causing issues was that the zip ties on the x-axis belt were physically bumping into the frame.
I replaced these with some slightly smaller ones, making sure to turn the latch face up so it wasn’t bumping into the frame. I also taped the ends of the belt for insurance.
I also received and installed the flex-plate (Th3D’s EZFlex^2) that I received for winning VirtCon 2021. This part is still kinda on-going. I’m not familiar with this material, and having a bit of adhesion issues, but I’m gonna work through it. Having a magnetically adhering flexible build plate should help a LOT. There were times recently when I was worried that I was going to run into problems because I couldn’t get prints off the semi-permanently mounted build plate, but this one comes off AND flexes to pop parts off!
The installation was pretty simple, I just got a bit paranoid adhering the magnet to the build-plate in case I got bubbles trapped under it. Also, cleaning the aluminum plate to adhere it took a while. Had to make sure I got all that glue off!
I’ve had a few adhesion issues so far, but I figure that’s just normal stuff to work through with a new material I haven’t messed with before. Just got to learn how to treat it right.
Wrap-up:
The earlier upgrades DEFINITELY helped, as I was able to successfully print the parts of the Flying V, which were larger than any parts I had printed before. I’m still learning about this flexplate… but it should help out a lot in the long run if I can get it to work.
Early design phase:
Here’s some of the early stuff I went through, starting from making a sketch to figure out the angles to make the triangles in order to fit the components, and an early test to make sure the electronic components could work together.
Guitar Neck and Head:
The neck was easy to construct from components I had ordered. I 3D printed the head first as it could be designed and printed independently of the rest of the guitar while I took time designing the interdependent body sections. I originally intended to use a USB cable and an ethernet cable as fake “guitar strings”, but dealing with the cables would have added another layer of complexity that I didn’t need to deal with in the timeline of a competition project.
I modelled 2020 extrusion in CAD in order to use offsets to create interfacing components. This was included in various forms in the guitar head and the keyboard body. I also left spaces for t-slot nuts to attach to the neck. I put in screws in place of tuning pegs. Eventually I’m considering some changes such as thickening the head (to better support the monitor) and adding some decorative (or functional) fake tuning pegs, perhaps with SD cards in them.
Keyboard Sections:
I designed these sections next after the head, as the were the next comparatively simplest section, and I needed to get parts in production ASAP, using the print time of the simplest parts to design the next parts to print. This section connects directly to the guitar neck, and contains the ball joint hinge connecting to the monitor section. It also houses the keyboard and the second guitar strap button for wearing the guitar around.
Keyboard 1:
This section’s primary features are the connections for the latch, the plastic section designed to slide into the 2020 extrusion, and the added t-slot nut to allow to clamp solidly to the extrusion.
Keyboard 2:
This section is the beginnings of the slot to hold the keyboard in place. It also continues the trend of t-slot nuts to hold onto the rail.
Keyboard 3:
This section houses the magnets that hold the keyboard onto the guitar.
Keyboard 4:
This part holds the remainder of the keyboard, as well as the guitar strap button for wearing the guitar around. At some point I want to slightly redesign this so that in the upper right of the keyboard slot there is space for access to the keyboard’s power button and a wire connecting the keyboard physically to the Pi instead of relying on bluetooth.
Keyboard Joint:
This had to be designed closely in tandem with the monitor side of this joint. This section’s purpose was to be solidly mechanically connected to the keyboard half, and hold the ball joint connecting the two halves together. There’s a space for a screw on the bottom to attach the ball joint, and a space and end-stop for a bold sticking out of the side of the ball joint. An accident after assembly caused the bolt to break straight through the end stop. Conveniently, I had recently added a 3D printing pen to my arsenal, and rebuilt the stop by hand in the same plastic… just much denser than it had been when it was originally printed.
Keyboard Section Assembled:
Keyboard Attachment:
The magnets are used to attach the keyboard, which can be folded up to avoid being too obviously a computer. They keyboard then folds out to lay flush in the slot. I had to add tabs to get an easy grip on the keyboard to fold it back closed. Another item on my to-do list if I rebuild this is finger slots for grabbing the edges of the keyboard, for a cleaner build.
Note: They keyboard section could use a bit more weight to offset the weight of the monitor. The monitor tries to make the whole thing lean backwards a bit. For now, the guitar-neck kickstand partially compensates. Thickening the head of the guitar to match the thickness of the body would also help.
Monitor Sections:
This section was one of the last things I designed due to to complexity and the need to get stuff to the printer as fast as possible.
Monitor 1:
This section includes the hole for the charging port cable and mounting points for the latch that holds the guitar closed.
Monitor 2:
This section has the slot for the battery pack, one side of the mounting holes for the Raspberry Pi mount, a hole for the power switch, and part of the space and connections for the monitor. There is also an extended gap and mounting holes in the section near the 2020 extrusion, where I later installed a metal strap to rigidize the connection between this section and the hinge. Without that strap that connection would have been far too weak.
Monitor 3:
This section holds the majority of the monitor, and the ethernet port.
Monitor 4:
This section includes the holes for the audio jacks, the slot for the USB hub, and a recessed section to accommodate the cables coming out of the side of the monitor.
Monitor Hinge:
The hinge sections were crazy, trying to deal with the ball joint. This half also had to accommodate a metal strap connection to monitor 2, connection points for the Raspberry Pi mount, a hole for possible future expansion, and a slightly different sideways connection to monitor 3.
Monitor Assembled:
Ball Joint Design Tests:
This is the mechanical 3D printed assembly that I designed LAST, due to complexity. For one thing, I ended up having to get a new tool to figure out what the heck the threads were on the bolts on this thing.
I had designed the connections with adjacent sections as I was going along. This took a couple iterations to get close enough, but as previously mentioned, the bolt broke through the end stop, which I repaired with a 3D printing pen. I also had some difficulties with the screw going into the monitor joint. I appear to have made the hole for the heat set insert slightly too large and it slipped around. 3D printing pen to the rescue again! I covered the back of the screw and filled the area around the screw with plastic.
Full Plastic Assembly Layout:
Assembly required connecting all the appropriate plastic pieces together with bolts, then sliding the neck into the slot on the keyboard half and tightening the t-slot nuts. The picture below isn’t fully assembled, but shows about where things went. In particular, the neck is upside down relative to the position of the boards it’s supposed to connect to, since i needed to see the insides of the case for planning purposes.
Hardboard backing:
In order to save a lot more 3D printing time and hopefully add some rigidity, I designed this with a board backing. Scored and snapped hardboard.
I didn’t photograph this very well, as I was kinda in a hurry to finish at the time. I cut hardboard to fit the back, drilled holes, and wrapped the edges with electrical tape for a slightly neater look than the edges of scored and snapped hardboard. Down the line this needs to be painted or replaced.
Monitor/CPU Internals:
Here’s the setup overall. I added breakaway extension connectors for the battery and the power switch (I colorcoded them to prevent mishaps). The USB cables mean to go to the USB 3.1 connectors are wrapped in blue tape.
Some parts were hot glued into place, as I can remove hot glue with a bit of effort if I need to make modifications later. Also, it allowed me to get away with not having positive retention on the audio jacks, which would have been a pain to implement on the short timescale I had available.
Here’s the modification I made to the PiSugar2 Pro battery management board in order to add an external switch that I could get to more reasonably. It seemed to work at first, but for some reason it no longer seems to work, so for the moment I’ve bypassed this board and only try using it directly plugged into the Raspberry Pi.
Upgrade Port/Cover:
Here’s where I placed and covered a spot for a potential USB upgrade, intended for a cable connecting the keyboard physically to the Pi for power and more reliable data. The cover is my maker coin, modified. The door in the middle of the cover conceals a hole that the two halves of the coin are designed to wrap around a USB cable. The whole thing can be removed for a larger hole, big enough to run a USB cable through.
Whew. I’ve probably missed some stuff, and may need to go back and edit, but this should cover the vast majority of the build.
So, you may have noticed that I appeared to drop off the web for about a month. Well, I was busy trying to build a full-size cyberdeck in a month for VirtCon 2021. I took Adam Savage’s advice from his book Every Tool’s a Hammer and took advantage of the competition to set a deadline for myself in order to build a concept that’s been floating around my head for a bit.
Now I get to share the result with you.
I decided to continue with the transforming feature from the last one, and the musical instrument theme from the first one, so here’s the ΠTar Flying V Cyberdeck.
A 3d-printed deck based around the general outline of a Flying V electric guitar, folding from guitar mode to cratetop cyberdeck mode. The neck is made of 2020 Aluminum Extrusion (aluminum extrusion was one of the requirements of the competition).
Here’s how it transforms:
I haven’t yet documented this one as thoroughly as my last build, but I can share with you some of what I do have this time. I may be posting more about this down the line as I sort through my photos.
Here’s the in-universe “sales flyer” as if this were being sold at a small shop as a custom item. It includes the general specifications and basic operating instructions.
I’ve posted the STLs on thingiverse, here:
I’ve been clearing my tasks and workspace in prep for this competition, hosted by the cyberdeck.cafe group. The rules were announced last week here:
https://cyberdeck.cafe/mix/virtcon-2021-comp
The competition has two categories:
Competition A: Design a thematic t-shirt for the cyberdeck.cafe group to use for merch (additional requirements on the above link).
Prize: VOXELAB Proxima Monochrome Resin Printer
Wow. A resin printer for a prize? Pretty awesome. But that’s not where my talents or interests lie. I’m more interested in:
Competition B: Designing a portable cyberdeck (additional requirements on the above link).
Prize: An EZFlex build plate or your design printed on a large 3d printer.
I have about a month to design and present a new cyberdeck, this time using aluminum extrusion or piping. Conveniently, I’ve had the basics of a design that fits these requirements in the back of my head for a while, but insufficient motivation to build it until now. The prize is nice, but I’m really in it for the impetus and deadline to build another design.
I tend to get a bit… quiet about the details of my competition builds, so you’ll probably have to wait to see the design until I’ve submitted my entry to the competition. I can tell you that I’m definitely gonna be using some components and concepts familiar to anyone who has seen my previous builds.
Most of the parts are at the sanctum or will be arriving soon. Most of my concept seems like it should be straightforward (which I think means I’m not grasping something) except for one thing the whole design hinges on, which might get iffy.
Points where I have been or will be doing some learning:
Components new to me for this build so far:
Aluminum 2020 extrusion
Fan-based cooling on a Pi
PiSugar 2 Pro for a Raspberry Pi 4B
MDF board
Processes I have a bit of concern over:
Printing large objects without warping:
I’ve been running tests this past week trying to adjust for some warping issues that I’ve had with large objects. I’ve built a temporary enclosure to reduce issues with drafts causing unequal cooling (I’d post it… but I think the current version is a rickety potential fire hazard that I don’t want to condone for others). I’ve also changed some of my print settings to help with adhesion. These include checking the bed levelling (I still may need to redo this), increasing the first layer temperature, adding a large brim… and simply avoiding the area of the printer that seems to run into the worst problems. Side benefit: I’ve rebuilt the arch lamp, and didn’t have to use tape or glue.
Cutting MDF:
I’m currently planning to use MDF for part of the design to cut down on the parts count. I can eliminate about 12 3d printed parts from the design if I use some kind of sheet material, and MDF seemed appropriate. There are two methods I have access to at the moment that I plan to try: 1) cutting with a reciprocating saw or 2) attempting the “score and snap” method. I’m expecting some difficulties with this, but even if it takes me a few attempts and a bit to figure out, it should still be better than printing all those additional parts. It should be a simpler and stronger build this way.
As you’ve probably been able to tell from the dedicated page on the website, I finished the Pi-Tar months ago, but I never got around to explaining the final construction of it. 2020 was a hell of a year.
Looks like I left off after adding a power switch to the casing. I apologize in advance for lack of detail on certain aspects, I’m catching up on something from months and a few projects ago.
USB Hub:
In order to offload some power requirements for additional USB peripherals from drawing power through the Pi, as well as to make a more convenient location for plugging things in, I decided to add a powered USB hub. As an added bonus, this one came with an SD card slot and a microSD card slot.
This did add some complicating factors, though. I needed the wires to fit through some pre-existing holes in the casing AND the thing uses USB-C. I ended up having to buy some additional parts to make this work, which got rather weird. I had to find a USB-A to USB-C cable that would connect to the Pi on one end and fit into the existing hole into the casing. I made that part work, but it took some finagling with the wires.
I also had to find a USB-C to USB-C connector to connect that wire inside the case to the USB hub.
I also had to add a micro-USB power switch cable to connect the hub to the power bank (as mentioned in a previous post), and I attempted to modify the cable coming from the Pi to avoid drawing from or backfeeding to the Raspberry Pi (I can’t remember if I ever got that part to work without losing power, but that was the intent).
Once I figured out the wiring, I then had to figure out how to attach the thing to the exterior of the casing. Originally I was going to use the USB hub as the basis of some sort of pseudo-cartridge system with USB drives, but eventually cut it down to just being a conveniently accessible hub.
As part of this, I also decided that I needed to move the large audio port to the exterior of the case, and add an additional regular-sized audio jack. Moving the port freed up the existing hole in the casing to pass through all the wires for the interface hub. This meant I didn’t have to try to cut or drill a new whole in the interface between the two halves of the shell.
The 3D modelling and physically attaching that in was troublesome, but electronically it was simple. I just added an audio splitter cable to the end, putting one female end exterior to the case and the other connected to the adapter I had already been using.
I went through a LOT of iterations with the interface module (one of the names I’ve been workshopping, may be subject to change), and eventually settled on a two-part assembly that screwed onto the casing through existing holes. It took a LOT of measuring and iterations to get it to fit reasonably, both with the electronic components and the actual casing. I designed it in two parts, with the larger portion (containing the hub itself) screwing directly onto the casing, and the smaller portion (containing the audio jacks) sliding onto the larger part, and then screwing into both the casing and the larger interface section. The smaller section also served the purpose of covering the hole that the wires went through.
I quickly learned in that process to only print as much of a model as I actually needed to test the fit of parts, in order to reduce turnaround time and materials wastage. I also found out that parts moving in different directions can lead to weird shenanigans, like installing one part causing another to become unplugged.
Once I got that figured out and painted up (along with the new power switch), I finally assembled it for the last time. That was… interesting.
Final Assembly:
There are a lot of parts of this assembly that have to be done in a specific (and weird) order or else it physically cannot be assembled. Connecting the Raspberry Pi, it’s case, and the interface between those wires and the shell is a very delicate process of going back and forth and making sure that you don’t crush ribbon cables while also carefully routing wires before and after attaching the cable interface (the bit with the universal greeblie).
Connecting the top and bottom halves is also fraught with issues (having to carefully move wires to lay properly while closing the shells), and you have to do that before you can even start on attaching the interface module. I know I’ve missed many steps in documenting this process, including some that anyone crazy enough to attempt recreating this might want, but I’ve only got so much time and patience at the moment. What madman decided to design from a pre-existing case this way?
Oh, wait, that was me. Ahem.
If you have further questions, please let me know what you would like to know more about and I can see about adding it.
Grand Tour:
Note: the stuff on the end of the grip handle and anything in pewter color is purely decorative and non-functional.
This has been a long project and a valuable learning experience. I learned more about Raspberry Pi (both from a hardware and software perspective), spraypainting, 3D modelling, 3D printing, electronics work (soldering), managing the details of a project, and working with professionals when I needed parts that I couldn’t yet make myself. I’ve even made new hobby contacts in the process who have helped me pick up more skills and helped out on other projects such as the Warp Core Lamp and encouraged me to make the Pioneer Falchion as another project.
I call this project “complete,” but as with a lot of other makers, this is more of a “project made it to baseline.” I’ve got some improvements I’d like to make (better power supply, attaching a headmounted display, making the Pi swappable as new models are released), but I’ve at least reached the initial goals I made before too much scope creep got in the way.
You may see more of the Pi-Tar (and possibly a sequel?) if/when I make upgrades to it.
Technomancer's Sanctum 