Dev log 06 (06/09/26)

Worked on finalizing the PCB for manufacture, turns out I have more problems than I thought so lets go through them:

Removed the charging option for ports:

• USB port on laptop can not provide the power needed to enable charging for all 4 ports.
• If I wanted charging made available I will have to connect the USB-C ports CC pins to the MCUs PD pins (yes the STM32G0B1KBTR MCU I picked can handle direct connection with diff pairs and PD)
• But even if PD connected (which allows the port to draw more power) the USB port on the laptop still can not provide enough power to since most laptop ports provide 500mA –> 900mA while charging option requires 1.5A.

Added test points:

• Bro its £70 (Manufacturing and shipping) therefore if something goes wrong I would like to give myself a higher/easier chance of finding what is wrong with the board.
• Test points enable me to physically debug the board with a volt meter etc, aaaaa frick i just remembered, I need to add GND test points as references… crap would have been bad if I forgot to add those cause that would make my test points redundant…

Fixing DRC errors:

• Man so many DRC errors since I was moving stuff around some vias accidentally joined up with traces running below them, but all those are fixed now, should be fine.

Dev log 05 (05/09/26)

Made minor changes based on feedback I got, I added tear drops and a 0.1uF filtering capacitor for the USB 2.0 VBUS to handle high frequency noise and flyback.

Note:

School has started now so I have way less time than I orginally thought so yeah.

Dev log 04 (30/08/26)

Finished PCB, I lost around 5 hours and i have 11 hours recorded but not loaded up in stardance for some reason I have attached the link here: https://lapse.hackclub.com/timelapse/W-kbFj_EBe9T

Some changes I made / what I have done:

• Changed the clock IC into a crystal oscillator cause it saves costs.
• Used JLC PCB impedance calculator to calculate the correct width and space between traces for 90 ohms of impedance for the USB 2.0
• Switched to 16pin USB-C port cause it is more readily available than 14 pin.
• Fixed some minor errors with the schematic such as forgotten/incorrect connections

Info on the PCB:

4 layers:

• layer 1: Main layer for components and differential signal routing with GND fills
• layer 2: Main GND layer
• layer 3: Power layer, main routing for VBUS and 3V3 fills)
• layer 4: Mainly for routing OCS, EN and RGB_DATA signals

What I will do next:

• Work on case design.
• Considering adding test points on the PCB to help debug possible failures.
• Find a way to solve the time not being added to stardance when its logged in lapse….

DEV LOG 03 (20/08/26)

Schematic development

• Worked on USB hub IC required connections.
• Switched out load switches from dual channel ones to single channel one (TPS22919DCK) cause i wanted battery charging to be available for my USBs so the load switches need to be rated for 1.5A cont 5.5v max and active high (cause easier and more reliable) I know there are dual channel versions of what i need but they they arent in stock in JLCPCB so yeah….

Personal note:

I orginally wanted a direct connection via SMBus between the USB hub and MCU but i couldnt figure out cause the data pins are multifunction pins with the config pins and both have different requirements in terms of pull up and pull down resistors so I got confused so i decided just to do default config instead to keep it simple.

DEV LOG 02 (15/08/26)

Bruhhhh component selection sucks….

Load Switch selection:

I decided to change the load switch to TPS 2052B (DGN Package cuz idk… i mean there was a reason but i forgot it… lol) cause it offers dual channel (space for 2 USB instead of one) and the B variant is just better in terms of performance.

USB controller IC:

For the USB controller I decided to go with the USB2514B-I/M2 (SQFN Package since JLCPCB offered it and it was the cheapest thing i could find the footprint for and its the new version) since it can handle 4 USBs which is just what i need for the basic USB-C hub im designing.

Crystal Oscillator for the USB controller IC:

Based of the USB hub specs it needs a crystal oscillator that has:

• 24MHz Freq tolerance 350 ppm
• <100ps of jitter
• 1.2 V CMOS Logic

For this component selection was so so so so so so annoying (again for me idk abt u) like I definitely need to study this and research this more cause I wanted to find an “active crystal oscillator” cause it saves me the hassle of calculating the capacitive load that I need to work out the values for the load capacitors or smth like i need to research that, but anyways ill come back to those parts later…

In the end I found an active crystal oscillator “1491-24003-BTBEGA” with the help of ai lol………….. cause its cheap and its offered by JLCPCB in their PCB assem, the annoying part about this was that you had to create the footprint and symbol your self… took me like 30mins cuz i had no clue how to create symbols and footprints in Ki - cad…. yeah that sucked.

Anyways thanks for reading I suppose? (whoever reads this ofc lmao)

DEV LOG 01 (13/08/26)

Right guys, its finally time to start working on this project, been busy the past few days studying fundamentals of PCB design (AC return current, differential pairs for USB and components such as clock oscillators and MCU required connections, LDOs or Load switches and some basic theory like how energy is transferred in electricity (electromagnetic wave, electric wave + magnetic wave and such)

Plan:

• Design a basic USB-C hub get the basics down.
• Move onto advanced designs such as ethernet, HDMI and audio (will need to upgrade MCU to cortext M4)
• Create a fully custom USB-C hub with all the features that I can possibly add, such as a OLED screen or audio speakers or framework style swappable ports or literally plug in the framework ports (if i can design a board that can do that, I mean its just a USB-C right….? I hope….)

What did I do today?

Component selection… damn I really didnt expect it to be this annoying (FOR ME) since theres thousands to pick from so I ai to recommend some trust worthy well known brands to pick from and give me some requirements for the component.

I only selected a load switch today the TPS 2051B by Texas instruments since it offered:

• 3 pins for VBUS switching for the USB.
• Undervoltage lockout
• 500mA continuous current (Required for USBs)
• Soft start (limits inrush current to prevent large current surges)

Sidenote:

I think I went into too much on the little details on PCB design but I guess the physics stuff I learnt helps like how a capacitors works and stuff will help next year in my A-level studies ig. *tho i swear i kinda forgotten most of the stuff…. ig ill review it later ofc.

Dev Log 16 (27/07/26)

What have I done?

• Worked on Git hub stuff like read me and uploaded some files.
• Mainly learnt how Git works like cloning repos and stuff (man that was so annoying learning how use Github is was so confusing *for me).
• Link to my Github for the macro pad.

What is there left to do?

• Get firmware done.
• Get PCB sent to manufacture for assembly

Dev log 02 (27/0726)

#What have i done?

• Learn the basics of web page creation created notes on the following:
• CSS basics covered
• HTML basic elements covered e.g div, header, class, href….

What will i do next?

Continue to develop this website and my skills like i want a better UI than this so ill check out to make it look more nicer.
#Features I want to add:

• Blogging system to blog on my projects.
• Place where I can upload my project portfolios.

Dev Log 01 (22/07/26)

Research stage and learning process

• Currently getting the basics down

What did i do?

• Learnt basics of HTML like how its the structure of the website and tells the browser what each thing is and where it goes, e.g that piece of text is a heading and goes there.
• Understood that HTML is made up of Tags that open and close and the functions of tags like or and headings like –> and such, also learnt some tages are self closing.

What will i do next?

• I will continue to study the basics and figure out how every bit of code works in html as i understand from previous experiences that you will need a solid understanding of the basics to move on to other stuff.

Note: I use ai to explain and help me study.

Ring Gear and Handle
I spent some time making the carrier pins for each layer of planets and the ring gear. My gears are set up in a 64/16/32 ratio (Ring, Planet, Sun). I’m going to layer 3 of these stages together to achieve a total reduction ratio of 8:1
The carrier pins are set up on my sun gear. I have designed small raised sections to prevent excessive plastic-on-plastic rubbing, these raised sections also give lubericant a place to sit, reducing overall friction which is critcal for a backdriven gearbox like this.
From experience, a little bit of friction at the very end stages will result in a very large effect, often making the system completely undrivable. This is because the 8:1 total reduction ratio means for example 3N of friction in the last stage will translate to 24N of additional force needed to start spinning
The ring gear is split to allow the herringbone gears to actually be assembled. (I’m not letting my lack of DFA getting the better of me again!)
Next I will work on a handle and gearbox lock mechanism, if i have time i may start on overall housing design and/or propellers

Dev Log 15 (21/07/26)

What have I done:

• Finished Aesthetics, added a Voronoi patterned geometry on the OLED cover, i experimented with adding a full Voronoi pattern across the top faces but that just looked messy and based on the feedback i gotten from friend and other people i have decided that its better to only have the OLED cover with the pattern.
• Added TPU feet to the casing as based on my experience with 3d printing, 3d printed objects tend to slide really easily on smooth surfaces like tables so yeah… will be printed separately from the case and glued on.

What will I do next?

• Well I guess its time to start to refine the BOM and Github stuff and get the PCB to be manufactured…. thats going to be a pain….
• Start work on firmware, need to research what to do and how to do it.

Thats my devblog today, thanks for reading.

Dev Log 14 (20/07/26)

Will post full dev log later, but short summary:

• Completed snap lock system just need to work on aesthetics and then im done.

Dev Log 13 (20/07/26)

Sooooo its finally starting to come into shape, think im done with the case? nahhh its a still a freaking ugly box, just wait to see what im going to do… (check my pfp for hints).

What have I done?

• Completed the cover for the OLED driver.
• Designed a slot system to secure the OLED to the case. (Totally did not overthink this…)
• Designed Rotary encoder knobs that i will 3D print, added the texturing to it for more grip.

What I still need to do to complete the case?

• Need to add in the snap lock system (ughh i can feel that this part is going to be annoying, probably need to 3D print a prototype to test before I can get it right)

Anyways thats my dev blog for today, thanks for reading.

Dev Log 12 (19/07/26)

What are the main things I done:

• Worked on OLED placement (so annoying due to the sizing requirements cause i really wanted this transparent OLED that looks really cool but it came with a driver board so i had create mounting space for it, ik i could have integrated it into the board itself but that will cause more failure points and for a macropad its not worth it so yeah)
• Edited PCB OLED driver board mounting holes to accommodate the OLED placement (another really frustrating part cause i had to revert and redo stuff…)

What Is there still to do for the case?

• Complete OLED mounting on the case.
• Add the snap lock system

Dev Log 11 (18/07/26)
Alr yeah took a 10 day break but back at it again, had to make changes to my PCB cause i didnt know about settings design constrains so some of my clearances were wrong…. butttt I work on the casing today and got the hardest part of the casing done, man i was stuck on this part forever…. i swear the longest part of designing smth is thinking of ideas over and over again… *screaming internally, like i spent so so so so long trying to figure out how to add in a plate to support the MX switches while using no screws and allowing room for the OLED and its driver board… Turns out i was OVERTHINKING…. and the solution was so simple, just create a hole for it and design a cool looking lid to cover it.
–––––––––– What did I learn? —————––
Dont overthink stuff.
———————What will i do next time?–––––––
Work on the case, should be done within a couple of hours, and its not going to a plain box…

Will post full dev log later on today, just need to log time

Dev log 10 (10/07/26)
I lost 3hrs of time….. :(((((
but oh well, at least im progressing, so heres what i did.
———————— What have I done ————————
I have started work and completed the main parts of the lower half of the case, i decided to add a 5 degree slant to the case for ergonomics and made a slot for the USB slot, my goal for the case is to not use screws so i designed a type of insert for the PCB mounting holes as shown in the picture.
————————What have I learnt—————— I noticed some boards have a plate for their MX switches to support them and I have done research of snap lock profiles and how they work so i can implement that into my case.

Dev Log 9 (08/07/26)
———————Quick summary–––––––––––
Finished 3D model of the PCB, didnt really learn much today apart from the orientation of a RGB led (SK smth mini-e) is defined by a chamfered notch. Will move onto designing the case now.

Dev log 8 (07/07/26)
––––––––––PCB Tracing Complete–––––––––––
PCB tracing complete, in my opinion its the best part of PCB design but yeah. I know I could have routed the 5V line to the right of the board but by the time I realized that I would have needed to redo large sections of the board, so I will keep that in mind if I ever want to make a version 2.
———————– What I learnt ——————
You can create a trace at any angle from a vias on a pcb, keep sensitive signal lines away from 5V and 3.3V lines, Minimize the interruptions of the ground plane to allow an uninterrupted ground path etc… 5V trace thickness : 0.5mm 3.3V trace thickness 0.3mm signal trace thickness : 0.25mm, when routing power to components it is ideal to split the power in parallel than in series as it allows the voltage across branches to be the same (basic physics).
———————–What will I do next? ——————
I will move on to casing design once I ensured my PCB can function, I can ensure that by asking around communities like discord or slack.

Ill post full dev log later on today