12/20/2025 12:06 PM - Reviewed Flipper Zero hardware and defined v1 design scope

https://docs.flipper.net/zero/development/hardware/tech-specs
https://docs.flipper.net/zero/development/hardware/schematic

I spent time reviewing the Flipper Zero hardware documentation to understand how the device is structured and which components are used. From this, I created a reference table of the main hardware blocks to better understand the overall system rather than copy it directly.

Based on this research, I defined a simplified scope for the first version of the project, including USB-only power, using a BLE module/dev board, a single Sub-1 GHz antenna, and a physical reset button. This helped make the project more realistic to build and easier to debug before moving on to schematic design.

This research made it clear that the Flipper Zero is far more complex than it appears, particularly in terms of power and RF design. To avoid overcomplicating the first version, I deliberately reduced the scope so I could focus on understanding the core systems rather than trying to replicate every feature.

12/20/2025 12:12 PM - Selecting Components for the v1 Prototype

Based on the Flipper Zero reference design and the simplifications defined in Step 1, I selected the specific components and modules to use for the first working version of the project.

For the NFC subsystem, I chose to use the X-NUCLEO-NFC08A1 expansion board rather than placing the NFC IC directly on the PCB. This board includes the ST25R3916B NFC reader along with the required matching network and antenna interface, significantly reducing RF design complexity for the first revision.

For the main MCU and BLE functionality, I selected the STM32WB5MMGH6TR module. Using a pre-certified BLE module avoids antenna tuning issues and allows me to focus more on firmware development and system integration during early testing.

For Sub-1 GHz communication, I decided to use an already-built CC1101 module instead of the bare transceiver IC. This simplifies the RF layout and makes the Sub-1 GHz functionality much easier to bring up, at the cost of some size and optimisation.

These were the main points of concern where I dedicated alot of time to finding appropriate parts and reading through datasheets.

Given the overall complexity of the project, using modular, pre-built components was a deliberate way to reduce the number of unknowns in the first prototype. RF systems, in particular, are difficult to debug when multiple things fail at once, so isolating them into proven modules made issues easier to identify and fix. This modular approach allowed me to focus on system integration and learning how the subsystems interact, rather than risking the entire project on several highly complex custom designs at the same time.

12/20/2025 10 PM - Schematic Design & Iteration

For this step, I worked on creating the full schematic for the board in KiCad. This ended up taking most of the day, mainly because of the number of peripherals involved and the complexity of the STM32WB power pins.

I started by placing the STM32WB5MMGH6TR module and figuring out the required power connections. A lot of time went into checking the datasheet and reference designs to make sure pins like VDD, VDDA, VREF+, VBAT, and VDDUSB were connected correctly and had the right decoupling. I definitely made a few mistakes early on, especially around power rails and net labels, but caught and fixed them as I went.

• VDD = core digital power
• VDDA/VREF+ = analogue/RF stability
• VDDUSB = USB interface reliability

Once the MCU power was sorted, I cleaned up the power architecture for the rest of the board. USB-C is used for power, with onboard regulators generating 3.3 V for the MCU and external peripherals.

I also spent time resolving ERC warnings and setting power flags by tracing each power rail back to its source and confirming that the STM32WB power pins, regulators, and USB input were correctly defined. On this project in particular, the large number of separate MCU power pins meant it was easy to accidentally leave a rail unpowered or incorrectly labelled, especially for VDDA and VREF+, which are critical for RF and analogue performance. I checked that all required decoupling capacitors were present and connected to the correct ground, and that the 3.3 V rail was consistently distributed to both the MCU and external headers.

After that, I added the peripherals. Since this is a module-based v1 prototype, most of these are connected using generic headers instead of fixed parts. The headers are labelled to match common module pinouts (NFC, RFID, Sub-1 GHz RF, IR, vibration motor, display, and expansion), which keeps the design flexible.

What each CORE component will do, context for the project:

• Software (Flipper firmware): the operating logic, user interface, and support for all peripherals. Running the Flipper software ties the hardware together, allowing NFC, RF, IR, and user inputs to be managed through an interface rather than as separate test systems.
• NFC / RFID: enables interaction with contactless cards and tags, which is a core feature of a Flipper-style device and necessary for testing access control and identification systems.
• Sub-1 GHz RF: allows communication with common low-frequency wireless devices such as remotes and sensors, this forms a major part of the device’s intended wireless interaction capability.
• IR: infrared transmit and receive functionality, allowing the device to interface with IR-controlled consumer electronics.
• Vibration Motor: simple haptic feedback so the device can signal actions or events without relying solely on visual or audio output.
• Buzzer: audible feedback for user interactions, alerts, improving usability during testing.
• Switches: the primary input method, enabling navigation and control of the device.
• Display: displays menus, information, and feedback from the system, making the device usable as a standalone tool.
• Expansion (Flipper-style expansions): additional functionality to be added later, supporting experimentation and future upgrades without redesigning the core board.

Due to my lower skill level in PCB building this step took alot longer than it would for many others.

12/22/2025 - Fixed ERC errors and truly finalised the schematic

Spent this session going back through the schematic and cleaning up a bunch of issues that were causing ERC errors and general confusion. Most of the work was around power domains and RF, especially separating and correctly labeling VDD, VDDA, VREF+, VBAT, and VDDUSB so they match how the STM32WB module actually expects to be powered.

On the RF side, I reworked the Bluetooth antenna path so RF_OUT is handled properly, adding a DC-blocking capacitor and a u.FL connector instead of tying the pin off incorrectly. This clears the ERC errors and also makes the BLE side of the board actually usable.

12/26/2025 8:20 PM - Routed the PCB

I am now a new man...

This was my third time routing a PCB, and it definitely showed improvement AND pain. The Xiao MCU that I used for the Macropad, and the ESP32-S3-WROOM-1u that I used for the development board can not compare to the struggle I have just gone through. The STM32WB5MMG is cruel in the way that it is so small but has so many pins. But, I can't complain anymore, since I am DONE!

There were multiple reroutes during this step, mostly caused by pin congestion around the MCU and the number of external connectors. Some compromises were needed, such as tighter clearances and smaller vias, but overall the board reached a fully routable and manufacturable state.

While not perfect, this routing pass really makes me feel proud that I've come so far from the Macropad project. It was also a reminder that schematics are the easy part, routing is where designs either make your or break you.

Firstly I trace the board edge cuts. Initially I had chosen dimensions of 100x60mm, but I soon realised that this was far too small for what I was trying to do so I increased it to 120x80mm. After recommendation from a Redditor, I made the board a 4 layer board (my first board that is more than 2 layers) in order to manage density and RF performance, particularly for the RF/Bluetooth section, which required creating a dedicated antenna keep-out zone to avoid copper interference. Next I moved the USB-C, U.FL and MCU, and I placed them on the edge of the board in the top right corner, I was told that these three components were very important and they should be placed on the boards edge so that’s what I did. Next I moved everything to do with the board power near it, such as the regulators, capacitors and resistors for the pins nearby, this made it easy to see what routes would go where. I then ‘made’ the modules, I added silkscreen boxes around each functional module, making the board easier to understand and populate. I then scattered everything around until it looked right. Once everything was placed on the board, I began routing. USB D+ and D− were routed as a matched, parallel pair to maintain signal integrity, while power traces were widened to 0.5mm to reduce voltage drop. After doing the important routes (power, antenna) I worked my way around the MCU’s pins, starting on the top ones and going around clockwise (I’m not sure why I did it this way, but it seemed to work). I did a lot of fiddling with routes. Lastly, once everything was done, I added Gojo onto it, more information around that is below.

The most difficult thing I found on this step once wrapping my head around what was on the top, and what was on the bottom. To solve this issue I used the 3D viewer after almost every change. It really helped visualise where things are. I highly recommend it.

The GND layer and +3V3 layer definitely made the step easier to a degree, and is much appreciated now looking back at all the extra routing I would've had to do.

Due to the very tight board size (even after increasing it), I had to selectively use smaller vias and reduced trace widths in congested areas to complete the routing without breaking design rules. For this I made a change in the Design Rules options.

I found most module specifications on the website where I initially found them, but for cases like the X-Nucleo that didn’t have exact measurements I had to follow the philosophy that “if it looks right, it is right” - My Engineering Teacher, DGA. Of course, I tried to make the most educated guess that I could.

The benefits to using modules can also been seen in this step as, if I had made each module as a component on the board e.g. the IR Transmitter and Reciver I would have a higher chance of making a mistake, would take considerably longer, and if there were an issue with the board later, it would be harder to diagnose. For these reasons, I really support and encourage the use of modules for this kind of thing.

For art purposed I added two Gojo images, these two silkscreens are my staples for all the projects I do. It has kind of become a unique thing that I do. Why? Because Gojo is the goat. He died to plot because Gege made him too strong. (Cope).

Estimated Time:
~2 hours of planning & component placement
~6 to 8 hours of routing + rerouting
~2 to 3 hours of cleanup, checks, and second-guessing everything

(Plus bonus time staring at the screen being stupid)

12/26/2025 8:59 PM - NFC Card Micro Project

For V2 of the Sorta Flipper Zero I would most likely have too make a copper trace antenna for the NFC. For V1 I am using a module, but I don't want to go into V2 with no idea so I decided that I would spend a good bit of time understanding how the module im using works, what I can bridge across, and built my own little NFC Card as a little introduction to what I will have to do in V2.

Here is the datasheet for the NFC module I hope to use: https://www.st.com/resource/en/data_brief/x-nucleo-nfc08a1.pdf

One thing I quickly noticed is that the antenna matters a lot. The orientation, size, and shape of the coil strongly affect the read range, and even small changes in trace layout or nearby copper can weaken the signal. This made me realize that designing a custom antenna will require careful attention.

I also learned that matching the antenna to the module is key. The module expects the antenna to have roughly the right inductance, otherwise communication becomes unreliable. Experimenting with the card showed me how adding or removing a turn in the coil changes performance noticeably.

Using the module is convenient because it handles a lot of tuning internally. But seeing it in action made it clear that if I want a custom copper trace for V2, I’ll need to account for these details myself.

This is the PCB I made off a simple schematic, I had a LED included but for the sake of readability I removed it in the PCB.

In v2 of the Sorta Flipper Zero, an antenna similar to the one that I have designed here will be used, that is where the main importance of this step can be seen (although it isn’t necessary).

And of course, I had to add my favourite two silkscreens of Gojo on it to make it uniquely mine.

12/27/2025 - Parts and getting prices

At this final stage, all that is left is to lock in the parts and get some quotes to figure out how much it would all cost. For most parts I merely used what was recommended as I don’t have a strong knowledge on what is better than something else, but for important ones I reach out for help from others and used AI. Some parts had to be changed during this part but luckily I didn't have to change anything on the PCB.

Sourcing parts from multiple suppliers was necessary due to cost and manufacturing constraints. Using JLCPCB’s standard, low-cost assembly option limited the number of components they could install, so some modules had to be sourced separately. Cheaper, non-critical modules such as the display and IR components were ordered from AliExpress to keep costs down, while critical parts like the MCU module, X-NUCLEO NFC board, and SD card reader were sourced from reputable supplier Digi-Key to ensure reliability and authenticity.

1/22/2026 2 PM - Firmware and NFC

Today I worked on getting the Flipper Zero firmware to work with the X‑NUCLEO NFC board. The chip on the board is the ST25R3916B, which is slightly different from the ST25R3916 the firmware expects. When investigating the firmware, I found that the firmware would fail because of an IC identity check.

To fix it, I forked the official Flipper Zero firmware on GitHub and updated the IC identity defined in the NFC driver header file so that the B variant would be recognised. I also made sure to keep all the original license files and credit the Flipper Zero repo, so everything stays GPL‑3.0 compliant. I also reviewed the OTP generation process and documented how firmware will be installed later using qFlipper.

This wouldn't effect the functionality since the ST25R3916B is essentially the same as the ST25R3916.

The issue was easy to isolate once I understood how the identity check worked, and the fix itself was small and low‑risk. Doing this early should avoid problems later when I start installing and using the Flipper hardware.

I also thought about using a custom NFC circuit instead (like a reviewer suggested), but that wouldn’t work. The Flipper firmware is tightly coupled to the ST25R3916 family, and using a different NFC chip would require writing a completely new driver and modifying large parts of the NFC stack. Using the ST25R3916B is the best option because it has almost identical functionality and works with minimal firmware changes.
This also means that I won't lose any of the Flipper functionality that I am aiming for.

All relevant materials, including the original and modified firmware, OTP documentation, and qFlipper references, have been included in my GitHub repository.

1/22/2026 4 PM - Create a basic case

Today I also did some basic CAD work in Fusion to design a simple housing for the Flipper project. I’m still not very confident with CAD, so I kept the design straightforward and focused more on functionality than aesthetics. At this stage the model is more of a placeholder, as I don’t yet know the exact final thickness once all modules are assembled, so the dimensions may not be completely accurate.

To deal with any size issues and stop the PCB from moving around inside the case, I’m planning to use small-ish foam blocks glued to the walls to prevent the PCB from moving around in the case. The top and bottom halves of the case are designed to attach using sixteen (eight on the top and eight on the bottom) small neodymium magnets (2 mm × 1 mm), which should make assembly easy while still keeping the case secure.

This step went reasonably well despite my limited CAD experience, and it helped me better visualise how the final device might come together physically. I also won’t need to apply for a grant to manufacture the case or magnets, as I’ll be returning to school soon and will have access to a 3D printer there.

2/15/2026 - PLEASE READ: The NFC expansion board

The NFC module was selected because it uses an NFC controller closely aligned with the one used in the Flipper Zero, which ensures access to the same hardware-level NFC capabilities and avoids functional limitations introduced by alternative controllers. Although it may appear that a cheaper NFC module could be substituted with a small firmware change, the existing firmware is tightly coupled to the ST25R3916 at the hardware abstraction level, including its register map, interrupt signalling, calibration procedures, RF field control, and timing behaviour. Substituting a different NFC controller would therefore require reimplementation of the entire NFC driver layer, including antenna tuning and RF control logic, even though the higher-level ISO14443 protocol handling could largely remain unchanged. Consequently, at the NFC subsystem level, this change constitutes a redesign rather than a modification, making the selected component the most technically appropriate choice for maintaining firmware compatibility and full functionality within the project scope.

In fact, the same board is not cheaper from Aliexpress either.