Embedded Files
"Ohhh!!! so this is where all of my time has gone to..." - I'm reminded everytime I update this.
>Welcome! Here you'll find high school sophomore year cardboard gaming steering wheel to college freshmen year full 3D printer build, and everything else in between.
>Projects with ✨ next to them is my favorite from each year, that recieved my "Project of The Year" award
Access the offical GitHub repository for the printer by clicking on the banner!
calm sunday afternoon photo!
KOROON P3S✨092525
>I find it hard to put this into words, but what was once a absolute "wouldn't it be crazy if??" idea...
>It is now right in front of my eyes, some how even better than i had imagined (usually it's the opposite), all from basically nothing
>The 8 month long journey has been rough, but now that it's done, I can give myself a pat on the back for not giving up
>Bolting things together at 3am, solidworks modeling under the influence of sleepiness and watching 3dp youtube while suppose to be doing homework, this project had it all
>Kingroon + Voron = Koroon??
Project Goals
>the goal? convert a $90 kingroon kp3s to a fully enclosed, high speed corexy printer
>there is the kingroon kp3s, and to be fair it is a perfectly fine 3d printer, but it's not the kind that i'm obsessing over
>a small, fast and enclosed desktop fdm 3d printer almost don't exist, but i really want one because i just think it's neat
>also the capability to print warp prone material is also would be a plus
"Donor" Kingroon KP3S
>Need to open door and remove top panel for printing pla, due to risk of heat creep, a flaw very similar to other enclosed printer. Although I haven't actually test if it's actually going to suffer from heat creep, since I used everything I could to prevent that dual 3010 + bi-metal heatbreak and thermal paste to heatsink
>Corexy motion, 9000mm/s^2 max accel, 350mm/s max speed (limited by mcu)
>Fully enclosed build chamber, gets up to 50*C
>Direct drive
>"Volcano+" hotend up to 48mm^3/s volumetric flowrate w/ hs-pla, 40mm^3/s w/hs-petg. up to 280*C
>Gdstime 8K RPM turbo 5015 part cooling fan
>Heated bed up to 100*C, spring steel pei sheet, manual bed leveling
>180 x 180 x 170 mm build volume
>Massive 7in touchscreen powered by klipperscreen
>Magnetic door
Side filament holder>
Electronic bay active cooling>
>At last deployed in a lab for engineering prototype production
"R&D" prints
>calihorn, calihorn, calihorn, calihorn, calihorn, calihorn, calihorn, calihorn, calihorn, benchy, calihorn, calihorn, voron cube, calihorn, calihorn, calihorn, calihorn, ut tower, calihorn, calihorn, benchy, calihorn...
>yes i love calihorn and orange filaments
>these are only the "solid" and complete prints, most of the other prints are failed prints, calibration pattern or towers
>throw back to may 2025, when a calihorn was being printed on prototype printer as another calihorn watches on...
>Right, but i do have a reason of printing so many calihorns:
>I think that most prints on this printer is going to be blocky, square engineering prototypes, and i think calihorn kind of fit that description, but sprinkle in some detail and extreme overhang test.
Prints samples
>it all started here, this is its first ever print
>corexy layer shift (unsecured pulley) and extrusion woes
>a set of first prints, all pla, and all have extrusion issue in some form
>it's getting there??
>3 "generation" of benchies a month apart
>Anyways, here are print samples with the current published profiles,
>My goal is to have decent print quality with fast speeds and cost-effective filaments, since again, this is a printer geared towards engineering prototype production
>These photos are under lighting conditions that highlights the impefections for me to identify and fix if I make future printers
eSun PLA Basic @220C, 0.24mm layer height, 2 walls, 15% grid infill, no support
>UT Tower (left): about 1 hr
>TIW Calihorn (closeup): about 29 mins
>Voron Cube (left): about 26 mins
>3D Benchy (closeup): about 34 mins
Hatchbox Terracotta PLA @220C, 0.24mm layer height, 2 walls, 15% grid infill, no support
>TIW Calihorn (closeup): about 29 mins
>3D Benchy (closeup): about 34 mins
Kingroon HS-PETG @255C, 0.24mm layer height, 2 walls, 15% grid infill, no support
>tiw calihorn (closeup): about 26 mins
3d benchy (closeup) : about 32 mins
>black glossy PETG under light shows printing artifacts such as ringing, inconsistent layer stacking, and VFA
Kingroon HS-PETG @255C, 0.20mm layer height, 2 walls, 15% grid infill, no support
A light-up Minecraft lantern Might be my favorite thing i have printed, like ever. I don't print many fun prints like this. Shoutout to the designer on Makerworld: https://makerworld.com/en/models/2023786-minecraft-lantern-3in1-light-edition#profileId-2181560
Silver and Brown Kingroon HS-PETG, pause fliament switch to get dual color lantern body, Light filter printed out of the Orange used for demotrio above
Generic ABS @270C, 0.24mm layer height, 2 walls, 10% gyroid infill, reverse on even, brim on, Dried @65C, 12hr
>I do not know quality or origin of this roll of ABS (although it was definitely moist), so i thought it'll be the perfect test for the generic profile, since this seems like the most "generic" filament out of any filaments
>Actually, these prints are printed at extremely high speeds, volumetric flow was at 30mm^3/s. Although ABS usually have no trouble flowing at such speeds, that is borderline pushing the boundary for printing ABS, because warping is a massive issue, and so is layer adheresion strength, which is drastically compromised
>One thing that stood out to me is ABS' ability to cool and thus led to excellent overhang and bridging, substantialy better than PLA and PETG dispite being a viscous filament at 270C
>Also I already knew this from the printer parts, but it always suprise me how light ABS really is
>In the end, it seems like aside from the bad VOCs, ABS is the best performing filament on this 3D printer, which I am very proud of, since before this, I was never able to print ABS.
>Having a printer I made specifically for engineering prototypes be able to print "engineering filament" like ABS this well, I'll take that win.
My favorites, I call them DemoTrio, as you can tell every test prints usually feature these 3. They are usually printed to test worst-case batch printing performace, which took about 1h11m
>TIW Calihorn (closeup): about 26 mins if printed individually
>Now, for some RepRap vibes
>This is the A motor mount for Koroon P3S, printed on Koroon P3S! Printer replication begins! (jk) Took about 1h3m
3d benchy (closeup) : about 32 mins if printed by itself
>Voron Cube (closeup): 25 mins if printed individually
>Koroon P3S Toolhead carriage - printed on Koroon P3S
took about 52 mins
>Voron Stealth Burner main body - a detailed part, the toolhead cover for the Stealth Burner, common used on Voron 2.4
>Print took about 1h30m and exhibits a lot of ringing, some shrinkage lines, but excellent bridging
some more misc samples
>generic PLA profile print
>but the notable thing here is the VFA, which i unfortunately couldn't fully resolve, but it's a problem that effects a lot of modern corexy 3d printers
>my best guess is that this is belt-induced, since the pitch of the VFA matches the gt2 belt pitch
>generic PETG profile
>first "real" print for a functional part i need in my lab
>other engineering parts it printed are mostly part for the printer itself
>generic PETG profile
>a bit difficult to fully show the print without the correct lighting
>and finally, ABS printed, just like the original plan...
>printing with a roll of random mystery abs that i picked up from the floor?? unknown age, unknown quality
>an important last step for sure, i have always wanted to print abs filament properly
pre?-post processing
>finished generic ABS benchy
>unknown age and quality but definitely known moisture level apparently
>also, i dropped it and now the chimney is gone...
>this print took about 37 mins
>My analysis on why TTC happnened, took weeks of debugging, dozens failed lids, hundreds of KWH of electricity
>It's got a bit expensive, so I started debugging using no filament and no heating, creating a no filament profile and tied up the extruder lever so no extrusions would take place
>Ran through common causes like over heating Raspi or MCU, not those, my was properly cooled
A quick reflection on the project:
>I have to say that I stepped outside of my comfort zone many times in this project, especially when Klipper "Timer Too Close" pops up suddenly, and also some time "MCU Lost Communication"
>Summary on Timer Too Close: I am not 100% certain that it is caused by the GD32F303 MCU's step rate limit but here are the evidence: https://www.klipper3d.org/Benchmarks.html
>The reason that I came to this conclusion was through continous klippy log monitioring shoutout: to the klipper forum GOAT: https://sineos.github.io/#dashboard
>I obsereved that everytime the printer does a fast 500mm/s move at 9000 mm/s^2 accel, the "bytes_retransmit - mcu" number goes up.
>So simple prints no error, complex print error? No, I have ran never ever into this issue when printing the hundreds of calihorns and benchy, fairly travel heavy prints.
>The real cause, I would think is those prints are small, thus they never induce the max travel speed 500mm/s, MCU is still able to keep up. Evidently, the first time this error shows up is in this box lid massive print that took up the whole bed
>As can be observed, at the beginning, there was no bytes that failed to transmit, then a few max speed travel moves from corner to corner, then to the raised borders, heavy travel between corners and kilpper shuts down quickly after since it can no longer communicate on time with MCU
>Solution? Lower the travel speed to 300mm/s, though that's a blow to my goal of the printer going toe to toe with top flagships, but in the end it's still with in that "fast enough" time range
>I dread this screen
>Monitoring Debian task manager suggest nothing wrong with host
>Timer Too Close solved! Yay..... WHAT
>"mcu unable to connect" Out of all times that I would have given up, this was it, this was the most common error on Klippers, I have never encounter it other than first time connection woes in the prototype printer
>School starting in a week, I want to deploy it soon, but I had no energy left after finishing my full time research role in the summer, the printer was 3 am project. I also got really sick, so there is that, I would not recommend debugging a 3d printer while fully covered in hives and having a fever!
>More no filament prints later, no problem? right?
>No, there is a fundametal different no heat printing and heated. The PWM control for bed and hotend heaters most likely cause large amount of EM noise, and look at that, the main USB cable SBC to MCU, the only communication between the "brains" and the "muscles" is the weakest link in the entire system, being routed right along heater wires and power supply wires
>Also I know these power wires are "noisy' because they made light in my house flicker from the PWM switching.
>I was using the crappiest USB cable I could find from misc cable bin, so that probably didn't help, despite all USB cable being shielded at some level, I now get why industrial controls don't mess around with proper sheilding
>Then another potential issue that might be causing bad signal integrity, is the touch screen pulling too much current from the Pi USB port, so i elminates that by wiring the screen directly to 5V buck converter
>After applying all these fixes the printer seems to have minimal issue with respect to electronics and software in the time it's been printing in the lab. I tried to throw random prints, push it hard when printing ABS on to it, and seem to do fine so far now.
Inital wiring set up, black USB cable was around the DC 24V main power, white heater wires for bed and hotend.
>New wiring set up, with Red/Black Braided, and higher quality cable.
>Screen wired directly to buck converter
>Granted that's probably too many ferrite choke, not really needed, but they only come in packs of 10.
>I obviously have no clue about anything FDM printing before all of this, so i would like to say a massive THANK YOU! ♥ to the Voron Design team for the incredible open-source Voron printers that i studied and took heavy inspiration from
>of course, THANK YOU! ♥ to the developers of Klipper, Moonraker, Mainsail, Klipperscreen and Orca slicer (and all slicers in the Slic3r family), as well as as the 3d printing community: youtubers and forum members everywhere.
>and, thank you to the Texas Inventionworks, they made it possible to make cool stuff like this for me and other students
the label on the printer
cool cinematic render haha
Where did the logo come from?
>The inspration of the logo comes from a very fitting part of the printer actually
>The corner piece of the frame cover just happens to be a nice geometric shape that i liked and is unique since the part itself was unique!
>Colored in the sides of the part and clean up the geometry inside, then add an interesting offset
"Top Tier R&D Facility and Equipment Transportation"
bery goof
koroon p3s - details
Materials: r&d costed a fortune
Robotic Arm Toolchanger 042525
>When i think about robotic arm, i don't think of 3d printing rightaway
>But the more i think about it, it does kinda makes sense, since it can be used for the chapter in 3d printing, which is conformal, or non-planar and a 6-dof robotic arm is definitely good for that
>The main thing i developed here is a simple twist-lock system to secure a low-load (like a fdm toolhead) end effector
>For more detail please check the development log above for the techical report, or the final (award-winning) presentation below
1 lb battle bot for M E 302
BatBot2 042425
> I finally tried battle bot because of a class
>I never liked battle bots since they take too much resources and sometimes ending up being destroyed (this one didn't because the kit motors are no where near powerful enough)
>i anticipated that most teams was going to use flipping (because of the ineffective weapons) so i made robot operational upside down as well
>"battle scars" were just weight saving holes drilled last day before it's due
>precision screw driver for scale (i didn't have a banana)
>this is the understated masterpiece
>every little bit of volume in the bot was used to control "wire entropy" it barely closes with help of bolts when finished (i guess this a control volume system in a thermodynamics sense)
>it is a dense little chonker, very dense compared to its opponent, weighting in at 0.995 lb/1 lb
main offensive weapon was the flipper (basically a wedge bot)
heavily armored body and wheel
secondary spinner weapon (did hurt my finger but does not no any damage)
batbot - iterations
concept i came up with early in the semester, which ended up looking closer to the final bot than the first "actual" bot
the first bot, really overweight because it's too big
>first robot in CAD
>impossible to manufacture, overweight and no one had time to improve on anything since everyone in the team had heavy course work
final battle bot, batbot2 was designed and built by me in 2 weeks before it's due
batbot2 - details
Materials: too much
Bevo Metro 042925
>This thing is basically a toy car but i kinda like it
>It does some cool line following and obstact detection which my friend did the coding magic on, shout out to him here: https://haicil.github.io/allen.ai/index.html
>I think i did fine for designing, making and coding it over one weekend (nights included), it doesn't look anywhere close to the render, but the vibe is there
I'm coming for Cap Metro and their Gillig buses, haha, just kidding, Thank you Cap Metro 801 bus drivers for taking me to school everyday
concept render
code flashing
bevo metro does line following and stops at stations (tape rolls)
Ramshorn Gallery 120224
>As part of the ramshorn scholars program, students are asked to create a visual piece to reflect on their first semester at ut austin
>Along with a lot of amazing creative art pieces, photo compilation video was the most popular way that students used to tell their story
>I used the oppotunity to start learning more about stepper motors and linear motion systems while creating something striking while the photos compliation format
belt driven "y" axis
8 picture frames in total
lead screw driven "z" axis to lift the pictures
controlled with arduino uno and uln2004 driver
on show at the 2024 ramshorn community gallery
ramshorn gallery - details
Materials: 3mm warped piece of plywood (allegedly) stolen (it was laying around on the ground) from tiw, pla 3d prints, gt2 belt, arduino and drivers, printed tape laminated photo, and 4 expo markers (out of ink) (allegedly) stolen from sanger learning center
Date: 2nd December 2024
Wotorobo 072624
>Robot that is cabpable of picking up a full 32oz bottle from the ground
>First "in house" robot fully made with my own tools
>Was intented as a educational project to introduce my brother to robotics since i had a great time at school.
made use of lego parts for ease of assembly
the wheels are probably the biggest thing i have printed
yet another 2 floor robot
peak wire manangement
wotorobo - details
Materials: too much
Date: summer 2024
P.H.E.A.R Nightingale✨120123
Robot for high school robotics team, won regional competition, does have a tendency to act up if looked at it the wrong way
(left) cardboard scale model concept, many ideas did not end up making it but still cool
(right) the final iteration that went to state competition, rendered in Blender for marketing
oh did i mention i made a lego set box for the occassion ? :)
nightingale - details
Materials: too much
Date: competition november to december 2023
Achargius 052024
Yes this is a robot! Well kinda
Throw your phone onto it, it charges your phone with a wire, fast and efficient, none of the inefficient wireless charging stuff
push phone over to align with the wall then push down onto connector
somehow works despite being made of 80% cardboard
used lego for some reason, which is a terrible idea and lego can not withstand the impulse of pluging in a charging cable
chargius - details (with engineering documentation)
Materials: too much
Date: high school senior design project spring 2024
scroll----------------------->
Ostacle "Avoider" 062024
Robot that intented to avoid obstacles except it kinda doesn't
Another robot in the sibling education series
Beeps at the wall like a mildly bad driver
obstacle avoider - details
Materials: too much
Date: summer 2024
Kinsolving problems✨111624
"KSPESP" 1st place in UT IEEE RAS Robotathon "Rasterchef"
read more in 2024 poty write-up above
Compact footprint helps the robot be more agile in the maze
Profoundly Prestigious, Pretty Powerful, Perfectly Printed, Physically Precise, Patent Pending, Proprietary Ping Pong Pickup Prototype (Perhaps) P^P
line follower works when that batteries aren't low
Custom team flag
kinsolving problems - details
Materials: too much
Date: november 2024
I.D.C.I 102724
id communication interface designed n' built over a weekend make-a-thon (by ut student engineering council)
aimed at improving communication within hospital for healthcare professionals, combating alarm fatigue problem
2nd place out of 24 teams, and won monetary prize
"Strike Sense" 011025
>my little brother's science fair project
>he had the great idea combined with his passion for martial arts, but with no knowledge of how to make it come to life
>taught him almost everything he needs to make the glove (extending all the way back to the "sibling education series" with when we were building wotorobo and obtacle avoider
>this included somethings that are new to both him and me, like using esp-now protocol to wirelessly transmit the score.
>he took 2nd place in embeded systems regional science fair (garsef) and 2nd place at state science fair (txsef) as well
initial mock up of everything on the glove
after cable management and securing everything in place
"strike sense" - details
Materials: too much
Date: october 2024
chessbot (UT RAS Demobot)
https://ut-ieee-ras.notion.site/chessbot?v=13cebc1c571480a2ae30000c7825d283
>basically a chess-playing robot, but with corexy motion for some reason
>this is a big team project was fun to work on and hangout with a bunch of people!
>i designed almost all the componets of corexy system: motor mounts, idlers,the "gantry", v-wheel carrige, laser cut and built the box, ran the corexy belt, designed and printed chess pieces
wheel
>yeah that's it
>spent a weekend on this
scroll----------------------->
the monitor in its new paint job, unfortunatly it does work anymore :( but i do have a bigger panel ready to be repurposed
15" Monitor 110422
repurposed display panel from a old laptop to make a usable monitor
backplate painfully fabricated using handsaw and drill on hardboard
triangular stands made of cardboard
i broke it while trying to renew its looks with paint i bended the flex cable too much and the pins snapped off, it served me well for 2 years, but i'm still sad
15" monitor - details
Materials: hardboard, standoff, cardboard, display panel, display driver board, psu
Date: november 2022
everything surrounding the keyboard is cardboard
Luoikeeb Giga78c 031023
not really a real project, but i did modifiy this keyboard to:
added usb-c connector
swap out to linear switch (was brown switch before ew..) which took hours of soldering
color onto the leds
brand new custom chassis out of ... cardboard! (yes including the "copper" weight) and fake volume knob that is just the delete key
Exellent Blender Animation that my brother did
luoikeeb 75 - details
Materials: original keeb and cardboard, usb c f, and paint
Date: 2023 spring break
Really stretching it.
literally... both of these made use of stretchy bands
Power Button 052922
the motherboard from the laptop that i harvested the display from still works as a pc
the only way to turn it on is sorting something at the original power button connector
since i don't know the the circuit layout, i found that if i use the flex cable to "poke" the connector the pc will turn on
made a mechanism to repeat the motion using of cardboard and hardboard.
did not ended up working since the laptop does not display output anymore
power button - details
Materials: hardboard, cardboard, elastic band, paper
Date: may 2022
scroll----------------------->
everything surrounding the keyboard is cardboard
Steering Wheel✨ 010922
>One of the most random thing i did
>I wanted to control a gamepad joy stick with rotational movement
>Rack and pinion is the mechanism to do that but i did not have anything to make it properly, so i cut every tooth on the gears and follower by hand with a knife...
>The end product does work through, it does control the joystick with steering-like movement
>As a plus i wrote about it on my college essay.
steering wheel - details
Materials: hardboard, cardboard, elastic band, paper
Date: jan 2022
Report abuse