Open Electronics

The Mactini turns a satirical BBC meme into a real, working laptop. Tucker Osman built an ultra-small device with a single button that handles every function. The project shows how you can build a minimalist computer with modern components and salvaged materials. The technical documentation is complete, and the result is surprising in both size and engineering.

At the heart of the system is a small Waveshare board based on the RP2350. This board handles the display, audio, and input. The display is a 1.69-inch SPI LCD with an ST7789V2 controller. The resolution is enough for a basic text and graphics interface. Everything is housed in a 3D-printed case designed in FreeCAD.

Firmware and video handling with DMA

The firmware uses DMA to copy screen data, freeing up CPU cycles for other tasks. Video playback is handled through Motion JPEG decoding. This approach lets the device show animations and clips on the small display without overloading the processor. DMA also reduces latency and improves image smoothness.

Audio is handled by an iPad speaker and an amplifier salvaged from a scrap board. This recycling choice lowers costs and shows that parts from broken devices can be reused. The sound is clear, considering the size of the device. The mix of salvaged and modern components makes the project accessible.

3D-printed case and materials

The case is 3D-printed and designed in FreeCAD. This open-source software lets you modify the design to suit your needs. The shell is compact and protects the internal components. 3D printing also lets you customize colors and finishes. The original project includes the files needed for printing.

To rebuild the project you need only a few key components: the Waveshare board with RP2350, the ST7789V2 display, and a salvaged speaker and amplifier. The full list is available in the documentation. In particular, anyone who wants to replicate the Mactini also needs a 3D printer and filament. The rest is soldering and patience.

The project is documented in detail. The project page on Hackaday.io collects photos, schematics, and a description of how it works. Tucker Osman also shared design choices and the difficulties he ran into. Anyone who wants to dig deeper can check that page to better understand each step. It is a valuable resource for anyone who wants to build their own Mactini.

The Mactini is an example of how a satirical idea can become a concrete object. The combination of a single button, a tiny display, and salvaged audio makes it unique. It also shows that you don’t need expensive parts to build a working computer. It is a perfect project for anyone who loves technical challenges and creative reuse.

Source: https://hackaday.io/contests

The post Mactini: the minimalist laptop built on an RP2350 appeared first on Open Electronics.

A wearable video camera can be built with a Raspberry Pi 5 and a camera module, all enclosed in a case that attaches to a hat. The project was created to record video while moving without holding anything in your hands. The board runs the Broadcom BCM2712 quad-core processor at 2.4 GHz with Arm Cortex-A76 architecture, and the 2 GB RAM version is sufficient to handle the video stream.

The assembly is shown in the reference video. The source shows how to connect the camera module to the board and how to arrange everything inside the case. Anyone following the video will find the main steps without having to look elsewhere.

The board and the camera module

The Raspberry Pi 5 is the fifth generation from the Raspberry Pi Foundation. The 2 GB RAM version uses the Broadcom BCM2712 processor, a quad-core 2.4 GHz chip with Arm Cortex-A76 architecture. This configuration handles the camera module without any issues, and the module connects directly to the board.

The camera module goes into a black plastic case designed specifically for it. The case can sit on a flat surface or be attached with Velcro. The second option is the one needed for the hat: the Velcro keeps everything steady even when walking or running.

The case and the attachment

The Raspberry Pi Camera Module case is a piece of black plastic that protects the lens and the circuit. Its shape is designed to be compact and comfortable on a hat. The Velcro included in the package lets you attach and detach it in seconds.

Only a few tools are needed for the assembly. The Velcro is cut to the right size and applied both to the case and to the hat fabric. Then the camera module is inserted into the case and the flat ribbon cable is connected to the board’s port.

The result is a lightweight camera that is not intrusive. You can use it for hikes, to document work, or to record a driving session. Video quality depends on the camera module and the light, but the board is never the bottleneck.

• Raspberry Pi 5 with 2 GB RAM
• Raspberry Pi Camera Module
• Black plastic case for the camera module
• Velcro for attachment
• Hat or flat surface

The list above summarizes the essential components. The Raspberry Pi 5 and the camera module are the two main pieces. The case and the Velcro complete the system and make it wearable.

The project also adapts to other uses. The same case with Velcro can be attached to a backpack, a bicycle, or a helmet. The board stays the same; only the mounting surface changes.

For those who want to go deeper, the reference video shows the assembly step by step. The source is especially useful for anyone who has never worked with a Raspberry Pi and needs to see the connections live.

Source: https://youtu.be/Hn-ClS2fmMU?si=BrOk84blJevP5YHU

The post A Raspberry Pi Camera Mounted on a Hat appeared first on Open Electronics.

A soldering fume extractor that turns itself on when the air gets dirty. That is what Curious Scientist built using an ESP32-C3 Super Mini, a dust sensor and a VOC sensor. The device adjusts the fan speed according to the pollution level it detects and shuts off after 3 minutes with no readings. So you no longer have to remember to switch the extractor on before picking up the soldering iron.

The heart of the circuit is a compact board, the same family many makers use in IoT projects. Next to it works an Arduino Nano, which handles the reading and control side. On the panel there is room for a 1.3″ OLED display with an SSH1106 driver, managed by a compatible Arduino library. The display shows the system status and the countdown before shutdown.

Two sensors and an adaptive baseline

Pollution detection happens in two stages. The first is based on readings from the dust sensor, the second on those from the VOC sensor. The VOC sensor does not use the raw TVOC value directly: instead it adopts a baseline that adapts slowly and looks for significant increases relative to it. This way the system is not fooled by the small fluctuations typical of these sensors.

The GP2Y1010AU0F dust sensor works with an infrared LED briefly switched on by the microcontroller. Light scattered by the particles is picked up by a photodiode, and the output signal is proportional to the amount of dust. The minimum activation threshold for the dust parameter is 40 ug/m3. For the ENS160 sensor, the AQI value it provides ranges from 1 to 5.

The controller does not trigger on a single event or on crossing a threshold. It takes several over-threshold readings within a fixed number of readings. In practice:

• LOW level: 3 over-threshold readings out of 5
• MED level: 12 over-threshold readings out of 20
• HIGH level: 15 over-threshold readings out of 25

This logic prevents false starts. The controller starts quickly at LOW speed and increases the fan speed only if the pollution persists long enough to show that the lower speed is not enough. So the fan steps up only when it is really needed.

5 V fan and shutdown after 180 seconds

The fan is a 12 V unit, but it is powered at 5 V at a very reduced speed. This is to draw air through the device so the sensors are properly exposed and false readings are avoided. If no pollution is detected for more than 3 minutes, the extractor is switched off. The device remembers the last pollution event and counts down from there.

During the last 30 seconds of the 3-minute countdown a warning message appears. Pressing the button within those 30 seconds resets the timer. So if you are about to do another soldering job you do not have to wait for the fan to stop and start again. The countdown for the automatic fan shutdown is 180 s.

Anyone who wants to rebuild the project will find the full account of the work on the Curious Scientist website. The firmware runs on a board from the ESP32-C3 family, so it is programmed with the Arduino IDE and the usual libraries. For the OLED display you need an Arduino library compatible with the SSH1106 driver.

Replicating the project takes only a few parts: an ESP32-C3 SuperMini development board, an Arduino Nano, a dust sensor, a VOC sensor, a 1.3″ OLED display and a 12 V fan. The dust sensor is the same model used in the project, the GP2Y1010AU0F from Sharp, designed to detect dust particles. For those who want to dig deeper, maker Curious Scientist’s site collects the work on this extractor.

Source: https://curiousscientist.tech/blog/automated-fume-extractor-soldering-made-easier

Related products

• Mini ESP32-C3 module
• Arduino Nano V3.3
• 1.3-inch OLED Display

The post Automatic Soldering Fume Extractor with Dust and VOC Sensors appeared first on Open Electronics.

Drawing a shape pixel by pixel in the browser and watching it light up on an 8×8 WS2812B LED matrix: that is what Rui Santos’ project does. The ESP32 connects to the Wi-Fi network and serves a web page with an 8×8 grid, where you pick the colour of each of the 64 RGB LEDs. Saved shapes stay in memory even after a reboot, so they can be shown on the physical matrix, reloaded into the editor or played back in sequence as a slideshow. All of it combines an asynchronous web server, a filesystem and the FastLED library on a single board.

How the web editor served by the ESP32 works

The web page shows an 8×8 grid and for each pixel you choose the colour. A slider sets the overall brightness of the shape, which is scaled before it reaches the physical LEDs. In the code the default brightness is 128 on a scale from 0 to 255. The matrix is also driven by a single digital pin, GPIO 2, through the FastLED library, with configuration parameters for orientation, horizontal and vertical flip and colour order.

The web server is asynchronous, so the ESP32 keeps serving the page without blocking everything else. That means you can draw while the matrix stays lit on the last shape sent. Anyone who wants to rebuild the project will find the full account of the work, from the sketches to the assembly, on maker Rui Santos’ site.

Shapes saved in LittleFS and the slideshow

Up to 10 shapes are saved in LittleFS as JSON, so they remain stored even after a reboot. The gallery shows a preview of every saved shape. Clicking it displays the shape on the physical matrix and reloads it into the editor for modification. You can also delete a shape, an operation that also turns off the physical LEDs, and enable a slideshow that cycles through all saved shapes with an adjustable interval. The default slideshow interval is 2000 ms between one shape and the next.

• ESP32 board
• 8×8 matrix with 64 WS2812B LEDs
• 5V 2A AC-DC power supply
• Jumper wires
• Arduino IDE with FastLED, ESPAsyncWebServer, AsyncTCP and ArduinoJSON 7.X

The software side relies on Daniel Garcia’s FastLED library, ESP32Async’s ESPAsyncWebServer and AsyncTCP, Benoit Blanchon’s ArduinoJSON 7.X and LittleFS. For anyone who wants to start from a larger matrix, the catalogue has the 16×16 flexible WS2812B Neopixel LED panel, with 256 LEDs instead of the 64 of this project.

Power and consumption of the 8×8 matrix

The WS2812B matrix is made up of 64 RGB LEDs arranged in an 8×8 grid and must be powered at 5V. At 5V each LED draws about 50mA at maximum brightness. Consequently, at full brightness in white, every 30 LEDs the matrix can draw up to 1.5A. The recommended power supply for the matrix is a 5V 2A AC-DC adapter, so with a useful margin above the current peak.

Those who prefer a shape other than the square one can look at the 8×32 addressable RGB LED matrix with 256 WS2812B LEDs, which uses the same type of LED but in an elongated geometry. The wiring stays simple: a single data wire to GPIO 2 and 5V power.

Source: https://randomnerdtutorials.com/esp32-8×8-matrix-shape-editor-web-server/

Related products

• 16×16 WS2812B Neopixel LED Panel – Flexible

The post Web shape editor for an 8×8 WS2812B LED matrix with ESP32 appeared first on Open Electronics.

In episode 384, the Hackaday podcast offers a curated overview of the week’s most interesting projects and news in the world of electronics and hacking. Editors Elliot Williams and Al Williams comment on their favorite articles, read listener messages in the Mailbag segment, and announce new challenges. They also showcase projects like a fully 3D-printable thermoacoustic Stirling engine, a turntable that reads MOD files optically, and a Casio F-91W watch modified for contactless payments.

Thermoacoustic Stirling engine and other projects

Among the most fascinating projects, a thermoacoustic Stirling engine made entirely with 3D printing stands out. This type of engine uses sound waves to generate motion, and the ability to print it completely makes it accessible to many makers. The podcast also discusses a turntable that reads MOD files optically, a creative solution that bridges the digital and analog worlds. Finally, a Casio F-91W modified for contactless payments shows how modern technology can be integrated into a classic object.

The podcast is not limited to projects: it also covers the history of space stations starting from the 1800s, a fascinating journey through the milestones that led to modern orbital structures. Additionally, must-read articles are discussed, including a talk on making magical PCBs and quick tips like emulating an iPod or making an LCD interface look like classic spy gear. The project page on Hackaday.io collects the details and links to dive deeper into each topic covered.

Challenges and contests for makers

The podcast announces the Retrocomputing Challenge 2026 and Supercon 2026, two unmissable events for enthusiasts. It also reminds listeners about the One Hertz Challenge, the currently open contest that asks you to design a device in which something happens once per second.

These challenges stimulate creativity and push makers to exceed their limits. The podcast also offers practical tips for those who want to get involved, such as using a TO-220 power transistor to handle more demanding loads. For those starting with prototyping, a perfboard for Arduino is the ideal starting point. Finally, for projects that require a response to movement, a high-sensitivity vibration sensor can make a difference.

Hackaday Podcast episode 384 confirms itself as a valuable resource for those following the world of electronics. Thanks to the editors’ comments and listener submissions, it offers a complete and up-to-date view. The Mailbag segment also allows direct interaction with the community, making each episode unique and participatory.

Source: https://hackaday.io/contests

Related products

• PCB – Protoshield
• High sensitivity vibration sensor with signal output and indicator LED 5-15 VDC

The post Hackaday Podcast: Stirling Engines, Contactless Watches, and Maker Challenges appeared first on Open Electronics.

A development board costing a few euros can become a complete i386 PC. That is what the Tiny386 emulator, ported to RP2350 boards by Mikhail Matveev, achieves: the Raspberry Pi Pico 2 emulates an i386 CPU, talks to an SD card, drives a VGA or HDMI output and accepts keyboard, mouse and gamepad. The firmware boots DOS, Windows 3.x, Windows 95 and Linux, so it behaves like a real Nineties PC, only much smaller.

At its heart is Chunhui He’s Tiny386 core, which emulates a complete i386 CPU with partial i486 and i586 instructions and an optional x87 FPU. The integration on the RP2350 platform is called Pico-286 and borrows the hardware patterns of QuakeGeneric. The PC peripherals code comes from QEMU, while the BIOS is SeaBIOS with its VGA BIOS. Everything runs on Raspberry Pi Pico SDK version 2.0 or later.

8MB PSRAM, SPI SD and a 378 MHz clock

Memory is the main constraint for an x86 emulator on a microcontroller, and here it is solved with the 8MB PSRAM mounted on the board. The emulator uses up to 8MB of RAM, while the settings menu lets you configure from 1 to 8 MB. The default CPU frequency in the build is 378 MHz, with a 504 MHz option, and the PSRAM runs at 133 MHz with a 166 MHz option. These are numbers to keep an eye on, because pushing the clock helps emulation but runs hotter and draws more power.

Storage goes through the SD card in SPI mode, managed by the FatFs module for the FAT filesystem. Disk images are read at runtime through a disk manager opened with Win+F12: from there you choose floppy, hard disk and CD-ROM. The maximum supported hard disk image size is 2GB, while standard floppies are 1.44MB, that is 1474560 bytes. The settings menu, on the other hand, opens with Win+F11 and changes the emulator configuration on the fly, without rebooting.

VGA or HDMI, PS/2, USB Host and I2S audio

Video output comes on a VGA or HDMI connector and covers text and graphics modes up to 640×480. For input there are two routes. The first is classic PS/2, with dedicated keyboard and mouse. The second uses the RP2350’s native USB Host, so USB keyboards and mice can be connected. On top of that there is support for the NES gamepad, which can work in mouse emulation mode. Audio comes out of an I2S DAC and the FM synthesis for AdLib comes from the MAME FM Sound Generator, the fmopl module.

• Raspberry Pi Pico 2 with RP2350 as the base board
• 8MB PSRAM to reach 8MB of emulated RAM
• SD card in SPI mode for floppy, hard disk and CD-ROM
• PS/2 keyboard and mouse, or USB keyboard and mouse via native USB Host
• NES or SNES gamepad with mouse emulation mode
• I2S DAC for audio output
• VGA or HDMI connector for video

Anyone who wants to rebuild the project will find the code to compile in Mikhail Matveev’s repository, with the Pico-286 integration for RP2350 and the third-party modules already wired in. The INI file parser is inih, the filesystem is FatFs, and the BIOS and VGA BIOS must be supplied as bios.bin and vgabios.bin. The build is done with Raspberry Pi Pico SDK 2.0 or later, so you need an up-to-date ARM toolchain and a bit of patience with clock configuration.

Assembly does not require exotic components: the board, the PSRAM, the SD slot, the connectors and the DAC. For the storage side, a MicroSD Card Reader module with SPI interface replicates the SPI read scheme used by the firmware for floppy, hard disk and CD-ROM. For audio output, a DAC audio converter with I2S interface covers the same function as the DAC described in the project.

The result is an i386 PC with complete peripherals at the cost of an inexpensive development board. DOS, Windows 3.x, Windows 95 and Linux run on pocket-sized hardware, with 8MB of RAM, 640×480 graphics and SD storage. For anyone who wants to understand how an x86 machine is emulated on a microcontroller, or simply wants a portable bench DOS machine, this is a concrete starting point.

Source: https://github.com/rh1tech/frank-386

Related products

• MicroSD Card Reader module with SPI interface
• DAC Audio Converter with I2S Interface

The post An i386 PC on the Raspberry Pi Pico 2: Tiny386 Emulator Runs on RP2350 appeared first on Open Electronics.

Koode Bot is a hospital reception system that works without internet. An AI-powered kiosk interviews patients in Malayalam, assigns a department, and generates a clinical summary for the doctor. An ESP32 robot then physically escorts the patient to the department. The project is designed to respect privacy and reduce staff workload.

The core of the system is a Raspberry Pi 5 with 8 GB of RAM. The patient selects a language on the touchscreen—Malayalam, Hindi, or English—and speaks into a USB microphone. The audio is converted to text by faster-whisper and sent to the Gemma 4 E2B model running on Ollama. The model conducts a structured clinical interview, asking follow-up questions based on reported symptoms.

Data flow and clinical report generation

At the end of the interview, the model generates a clinical report in JSON with department, urgency, and summary. The report is saved in SQLite. The kiosk displays the token and assigned department, and sends an MQTT message on the topic koode/bot/navigate. The token has the format K0419001: the initial K, month and day, and a daily counter.

Report generation takes 60–120 seconds. The system handles up to 12 question-and-answer exchanges in the interview. Model response latency is 10–30 seconds, while speech recognition takes 2–6 seconds. Model loading time is about 35 seconds. RAM usage is about 9 GB.

The Koode Bot robot and ultrasonic navigation

The Koode Bot robot uses an ESP32-S3 and receives the MQTT message with the assigned department. MQTT latency is 1 second. The robot escorts the patient using ultrasonic sensors for navigation. An HC-SR04 sensor detects obstacles, while an L298 driver controls the 12 V DC motors.

The department misclassification rate is about 15%. The system has significant potential for epidemic detection by analyzing symptoms collected from interviews. Additionally, the offline nature protects sensitive patient data.

The project is documented in the shan repository. Those who want to replicate it will find code for the Raspberry Pi and the ESP32, plus the configuration for Ollama and the models.

What you need to rebuild the project

The main components include a Raspberry Pi 5, an official touchscreen, a USB microphone, and a Hailo-8 AI HAT accelerator. For the robot, you need an ESP32-S3, an HC-SR04 ultrasonic sensor, an L298 driver, and 12 V DC motors.

The software runs on Ollama with the Gemma 4 E2B model (Q4_K_M). The backend uses Flask, SQLite, and MQTT (mosquitto). On the robot, PubSubClient and ArduinoJson handle communication. For those starting from scratch, the Raspberry Pi 5 is the recommended base for the kiosk.

• Raspberry Pi 5 with 8 GB of RAM
• Raspberry Pi Touch Display
• USB microphone
• Hailo-8 AI HAT
• DFRobot FireBeetle ESP32 IOT Microcontroller
• HC-SR04 ultrasonic sensor
• L298 motor driver
• 12 V DC motors

The system is designed for hospitals with reduced staff or areas with limited connectivity. Privacy is guaranteed by the fact that no data leaves the device. The project demonstrates how edge AI can bring intelligence to critical contexts without depending on the cloud.

Source: https://github.com/lil-shan/Koode

The post Koode Bot: Offline Hospital Triage with Edge AI on Raspberry Pi 5 appeared first on Open Electronics.

Mouser Electronics has expanded its stock of Nordic Semiconductor products with development and evaluation kits for the nRF54L15 and nRF54LM20 Bluetooth Low Energy SoCs, the nRF9151 cellular IoT system-in-package, and the nPM2100 power management IC. Mouser has distributed Nordic products since 2011.

nRF54L15 development kit

The nRF54L15 DK is an evaluation and prototyping platform for the nRF54L Series of Bluetooth Low Energy (BLE) wireless SoCs. It supports the full BLE feature set as well as other common communication protocols, and is intended for applications such as Industrial IoT (IIoT), augmented and virtual reality (AR/VR), asset tracking, and smart home products.

nRF54LM20 development kit

The nRF54LM20 DK supports application development for the nRF54LM20A and nRF54LM20B wireless SoCs. It is a low-cost, single-board kit that gives access to all SoC features, backed by the software examples, modules, and libraries in the nRF Connect SDK.

Both SoCs support Bluetooth LE, Bluetooth Mesh, Matter, Aliro, Thread, Zigbee, and proprietary 2.4 GHz protocols (up to 4 Mbps), plus High-Speed USB. The nRF54LM20B also integrates an Axon neural processing unit (NPU) that accelerates edge AI workloads such as health, biometric and activity monitoring, voice control, audio classification, and low-fidelity vision.

nRF9151 SMA development kit

The nRF9151 SMA DK is designed for in-depth evaluation of the nRF9151 LTE-M/NB-IoT system-in-package (SiP), which also supports DECT NR+ and GNSS. Its SMA connectors allow direct connection to lab instruments for conducted measurements, or to external antennas for field testing. A beta modem firmware adds support for NB-IoT NTN (satellite communication).

The kit includes antennas from Taoglas (cellular/NTN) and Kyocera (GNSS), and IoT SIM cards from Deutsche Telekom, Onomondo, and Monogoto.

nPM2100 evaluation kit

The nPM2100 EK allows no-code configuration of the nPM2100 PMIC, which is designed for low-power devices running on non-rechargeable (primary) batteries. The PMIC includes a boost regulator with automatic pass-through mode, allowing batteries to be used down to their functional end point.

With the EK connected to a PC, all PMIC settings can be configured and exported as code for direct integration into the host microcontroller application. The kit comes with six interchangeable battery holders covering a wide range of primary cell configurations.

Availability

The kits are available from Mouser. The Nordic Semiconductor products stocked by Mouser are listed on the Nordic Semiconductor manufacturer page.

The post Mouser stocks Nordic nRF54L, nRF9151, and nPM2100 development kits appeared first on Open Electronics.

ROHM has developed a fourth generation of 650 V IGBTs for automotive electric compressors and HV heaters, and for inverters in industrial equipment. The devices reach a collector-emitter saturation voltage of VCE(sat) = 1.55 V, which ROHM reports as the lowest conduction loss among 650 V automotive devices*, while offering high short-circuit tolerance and qualification to the AEC-Q101 automotive reliability standard (Fig. 1).

Application context

As electric vehicles move to higher voltages, SiC devices are increasingly used in high-power applications such as traction inverters. 650 V IGBTs, meanwhile, remain widely used as switching devices in lower-power auxiliary systems, including automotive electric compressors and HV heaters. Silicon IGBTs are also common in industrial equipment, particularly in motors and compressors, and demand is expected to keep growing.

These applications call for energy savings and smaller equipment, and therefore for power devices with higher reliability, smaller size, and higher efficiency. Heater and inverter circuits in particular need enough short-circuit tolerance to withstand a fault for the time required to detect and interrupt the overcurrent.

Device design

ROHM revised the device structure, including the manufacturing process and the edge termination structure, to combine low loss with high short-circuit tolerance while meeting higher voltage requirements. The new structure increases current density and reduces both conduction and switching losses. Despite the trade-off between lower loss and short-circuit tolerance, the devices guarantee a short-circuit withstand time of 7 µs at Tj = 25 °C.

Lineup and availability

The lineup includes 12 products in the TO-247N package, the RGAxxTS65HR and RGAxxTS65EHR, and 10 bare-die products, the SG83xxWN series. ROHM is also developing 12 products in the TO-247-4L package, the RGAxxTR65HR/RGAxxTR65EHR series.

All packaged devices share a collector-emitter voltage of 650 V, a short-circuit withstand time of 7 µs at Tj = 25 °C, an operating junction temperature of -40 °C to +175 °C, and AEC-Q101 qualification. The -HR versions have no built-in diode; the -EHR versions integrate a fast recovery diode. The TO-247N package measures 16.0 × 21.0 × 5.0 mm, the TO-247-4L 16.0 × 23.45 × 5.0 mm.

TO-247N (available)TO-247-4L (under development)IC (TC = 100 °C)VCE(sat) typ.

RGA60TS65HR / RGA60TS65EHR	RGA60TR65HR / RGA60TR65EHR	35 A	1.55 V
RGA80TS65HR / RGA80TS65EHR	RGA80TR65HR / RGA80TR65EHR	44 A	1.55 V
RGA00TS65HR / RGA00TS65EHR	RGA00TR65HR / RGA00TR65EHR	52 A	1.55 V
RGAX2TS65HR / RGAX2TS65EHR	RGAX2TR65HR / RGAX2TR65EHR	62 A	1.55 V
RGAX5TS65HR / RGAX5TS65EHR	RGAX5TR65HR / RGAX5TR65EHR	74 A	1.60 V
RGAY0TS65HR / RGAY0TS65EHR	RGAY0TR65HR / RGAY0TR65EHR	88 A	1.65 V

Packaged 650 V IGBTs: -HR without diode, -EHR with built-in fast recovery diode.

The bare-die devices are rated at 650 V with a typical VCE(sat) of 1.55 V and a short-circuit withstand time of 7 µs; all chips are 75 µm thick.

Bare dieIC (TC = 100 °C)Chip size X × Y (mm)

SG8351WN	20 A	3.08 × 3.08
SG8352WN	25 A	3.10 × 3.48
SG8353WN	30 A	3.52 × 3.52
SG8359WN	40 A	3.74 × 4.10
SG8355WN	50 A	4.20 × 4.39
SG8356WN	60 A	3.90 × 5.48
SG8358WN	75 A	4.50 × 5.70
SG8357WN	100 A	5.70 × 5.70
SG8360WN*	150 A	6.82 × 6.82
SG8361WN*	200 A	7.75 × 7.75

* Under development.

The TO-247N products and some of the bare-die products are available now, and the TO-247N parts are sold through online distributors including DigiKey and Farnell. ROHM plans to add more products in the same packages and to develop compact surface-mount IGBTs in the TO-263L package and in top-side cooled (TSC) packages.

Applications

• Automotive electric compressors
• Automotive HV heaters (PTC heaters, coolant heaters)
• Inverters for industrial equipment

Design support

ROHM’s website provides SPICE models that reproduce the electrical characteristics of the devices in simulation, PLECS models for circuit simulation, and other circuit design material. More information is available on the field-stop trench IGBT product page.

EcoIGBT is ROHM’s IGBT brand for devices and modules aimed at high-voltage power applications. EcoIGBT is a trademark or registered trademark of ROHM Co., Ltd.

*ROHM research, August 2026.

The post ROHM fourth-generation 650 V IGBTs with 1.55 V VCE(sat) and AEC-Q101 appeared first on Open Electronics.

Cambium Networks has introduced cnMaestro MarketApps Hub, an open marketplace of web apps built to extend cnMaestro, the company’s centralized network management system. The hub turns a Cambium ONE network into an application platform: network operators, MSPs, site managers, and end users can run the workflows they need, and anyone able to call the cnMaestro APIs can develop and publish apps. According to Cambium, AI coding tools and an AI-ready development platform make it possible to build custom apps in hours or days.

An open marketplace with defined responsibilities

MarketApps Hub extends Cambium’s existing MarketApps framework into a governed app ecosystem, with a public catalog, a developer SDK, single sign-on (SSO), app hosting, and platform-level security. It is built on the public cnMaestro APIs and on a developer experience designed for AI coding agents. Apps are hosted and secured on the platform, and can be written by anyone from Cambium engineers to the customers who operate the network every day.

Bruce Miller, VP of Enterprise Marketing at Cambium Networks, said that AI has turned domain experts into developers, and that a purpose-built app that once took an engineering team a quarter to build can now be created in hours or days. In his view, the hub gives these developers a marketplace and gives customers functions tailored to their vertical market and workflow.

Five app classes

Every app in MarketApps Hub belongs to one of five classes, which tell customers where it comes from and how it is supported:

• Certified: production-grade apps developed, tested, hosted, and fully supported by Cambium.
• Labs: the fast track. Apps are released quickly and shaped by customer feedback, and those that prove themselves move to Certified.
• Partner: apps developed and supported by Cambium partners.
• Community: open contributions from the developer community.
• Private: apps exclusive to a single customer account, with that customer’s data and network.

Apps can move up to a higher class as adoption shows demand, so a community idea can become a Certified app or a native cnMaestro feature.

Development and distribution

Cambium’s position is that AI has made writing apps easier, and that the harder part is now hosting, securing, and distributing them to the right audience with adequate support. MarketApps Hub addresses this with platform hosting, SSO, a public catalog, and controlled distribution through the five app classes.

The MarketApps developer platform is designed for AI coding agents. A quick-start guide provides boilerplate code, an SDK, and a live cnMaestro development environment with a single command, and a single pull request publishes an app globally.

Because apps are delivered as plain web applications that run on mobile devices or in any browser, they can serve users outside the network team, such as a resident configuring personal Wi-Fi, a site manager overseeing a location, or a technician installing equipment, without requiring a cnMaestro account.

The post Cambium Networks launches MarketApps Hub, an app marketplace for cnMaestro appeared first on Open Electronics.

Mouser Electronics has launched a digital resource center dedicated to consumer technology. It collects material on emerging technologies, reference designs, and implementation guidelines for engineers developing connected consumer products.

Design topics

As consumer devices become more portable, intelligent, and interconnected, designers have to provide seamless connectivity and adaptive user experiences. The resource center covers several of these design priorities:

• USB Type-C® charging and power management systems, which enable faster charging and simpler connectivity
• AI-based ambient systems built on sensors and contextual processing, which make devices more adaptive and responsive
• Physical buttons, which are seeing renewed interest for tactile, distraction-free interaction, while touch and voice interfaces continue to spread

Content

The center was curated by Mouser’s technical team and its manufacturer partners, and includes articles, blogs, eBooks, and new products for consumer technology design. It also features the latest episode of Mouser’s Empowering Innovation Together (EIT) series, Engineering AI for Daily Life, which looks at emerging AI technologies and their practical uses in consumer products, including voice assistants, AI-based travel planning tools, and wearables that provide more detailed health information.

Featured products

The new consumer-oriented products stocked by Mouser include:

• STMicroelectronics LSM6DSV320X: a 6-axis inertial measurement unit (IMU) that combines an accelerometer and a gyroscope in a compact, low-power package. It integrates activity tracking, gesture recognition, and context detection, and targets wearables, AR/VR systems, and smart devices.
• Molex Premo-Flex SlimStack jumpers: flexible flat cable (FFC) and flexible printed circuit (FPC) jumpers that provide high-density, flexible interconnects for compact, lightweight consumer electronics. Their thin profile helps optimize the internal layout of smartphones, tablets, wearables, and portable devices, and simplifies wiring in tight spaces while preserving signal integrity.
• Littelfuse SC3402-02ETG: an ESD protection diode that protects sensitive consumer electronics from electrostatic discharge and transient voltage events. It is designed for high-speed interfaces and compact applications, and its small footprint suits mobile devices, wearables, and other space-constrained equipment.
• TE Connectivity 0.8 mm low-profile IDC connector system: compact, high-density board-to-board and wire-to-board connections for consumer electronics. The low profile helps reduce device thickness while carrying data and power, and the system is designed for automated assembly.

The post Mouser opens a consumer technology resource center appeared first on Open Electronics.

Taoglas, a supplier of antennas and RF solutions, has completed the acquisition of Wireless Instruments Sp. z o.o. (“QuWireless”), a company that designs antennas and antenna-enclosure systems for wireless communication devices. The deal extends Taoglas’s capabilities in outdoor and industrial connectivity, where demand for rugged solutions for harsh environments is growing.

Router inside the antenna enclosure

Based in Poland, QuWireless develops integrated “all-in-one” products in which the router is installed directly inside the antenna enclosure. Enclosures are available for all major brands of industrial routers: the installer inserts and secures the router in the enclosure, leaving a single unit to mount.

With significantly less RF cabling, this type of installation can also improve signal strength and reliability in settings such as:

• industrial sites
• telecom network infrastructure
• smart grids
• transport systems
• building rooftops
• remote locations and areas with limited network coverage

Combined portfolio

Taoglas says that adding QuWireless’s antenna and enclosure expertise to its own RF and antenna design, engineering, global testing, and certification capabilities will shorten development cycles and make field performance more predictable. The combined portfolio covers a wider range of supported devices, frequency bands, and form factors.

The acquisition comes as demand for wireless connectivity in outdoor, industrial, and remote environments grows with the expansion of 5G, private networks, and IoT deployments.

“Much of the challenge today isn’t just RF performance; it’s getting reliable, robust connectivity deployed quickly, with controlled installation costs, especially outdoors,” said Dermot O’Shea, CEO of Taoglas. “Their antenna enclosures remove much of that complexity and significantly reduce installation time.”

Deal terms

QuWireless will operate as “QuWireless, a Taoglas company”, and its product portfolio will initially continue under the QuWireless brand. The team will remain in Poland as a center of excellence for antenna-enclosure design, working with Taoglas’s global engineering teams.

According to Michal Gorzad, CEO of QuWireless, joining Taoglas will allow the company to bring its designs to a much wider market and integrate them with a broader range of RF technologies.

Financial terms of the transaction were not disclosed.

The post Taoglas acquires QuWireless, maker of outdoor antenna enclosures appeared first on Open Electronics.

Mouser Electronics now stocks the ESP32-P4X-EYE, a vision development kit from Espressif Systems designed for camera-based applications such as smart surveillance cameras, detection with vision models, and edge computing in IoT systems.

A development kit shaped like a small camera

The ESP32-P4X-EYE comes in a mini digital camera form factor: a compact camera-style enclosure houses the camera module, with a 1.54-inch SPI LCD on the back, a digital microphone, a rotary encoder, and a push button. The kit is intended for real-time image processing in AI, edge computing, human-machine interface (HMI), and camera applications.

With the camera, display, microphone, and microSD card on board, the kit can monitor its surroundings in real time and collect image and audio data.

Main features

• SoC: Espressif ESP32-P4 with a 400 MHz dual-core RISC-V processor
• Memory: support for up to 32 MB of PSRAM
• Interfaces: USB 2.0, MIPI-CSI/DSI, and several other peripherals
• Video: H.264 encoder
• Display: 1.54-inch SPI LCD on the back of the enclosure
• Audio: digital microphone
• Controls: rotary encoder and push button
• Storage: microSD card
• Wireless: ESP32-C6-MINI-1 module with Wi-Fi, IEEE 802.15.4, and Bluetooth LE

Wireless module

Wireless communication is handled by Espressif’s ESP32-C6-MINI-1 module, which supports Wi-Fi, IEEE 802.15.4, and Bluetooth LE and is designed for applications including smart home, industrial automation, healthcare, and consumer electronics. The module has an on-board PCB antenna.

Developers can also connect external peripherals to the ESP32-P4X-EYE using jumper wires.

The post Mouser now stocks Espressif ESP32-P4X-EYE vision development kit appeared first on Open Electronics.

Vertiv, a critical digital infrastructure company, plans to expand its manufacturing and testing operations in Ireland and the North West. The company cites growing global demand for infrastructure supporting artificial intelligence (AI), colocation, and hyperscale data center deployments.

Planning application in Letterkenny

Vertiv will submit a planning application to Donegal County Council for an additional 7,000 m² at its Letterkenny site, supporting the planned creation of up to 300 new jobs in the region.

The company’s existing presence in Ireland and the North West includes manufacturing facilities in Letterkenny, Burnfoot, and Campsie (Derry), a Central Procurement Office in Limerick, and a sales and service office in Dublin. Vertiv is currently hiring at these locations to support ongoing organic growth.

Hiring and technical scope

Since the beginning of 2026, Vertiv has added more than 500 employees at its local sites and continues to recruit for engineering and operations roles. According to the company, the new staff will increase production capacity and expand its technical expertise in power, cooling, and industrialized infrastructure solutions, supporting faster deployment of AI, data center, and critical industrial infrastructure worldwide.

“As demand for critical digital infrastructure continues to rise, we are supporting customers as they deploy AI and data center capacity around the world,” said Paul Ryan, president of Vertiv in Europe, Middle East and Africa (EMEA).

Government support

The proposed project is supported by the Irish Government through IDA Ireland. Taoiseach Micheál Martin described the decision as good news for the region and for Ireland, while Peter Burke TD, Minister for Enterprise, Tourism and Employment, said the project underlines the role of the Irish facilities in Vertiv’s global manufacturing network.

Michael Lohan, CEO of IDA Ireland, pointed to the expansion as an example of global companies growing high-value operations from regional locations in Ireland.

The post Vertiv plans 7,000 m² expansion at Letterkenny with up to 300 new jobs appeared first on Open Electronics.

In a case study, Yokogawa describes how Integrated Optics, a Lithuanian manufacturer of lasers and photonics products, uses the AQ6373E optical spectrum analyzer to characterize its ultra-compact visible lasers during development and in production.

Visible lasers are used where spectral quality, stability, and repeatability matter: microscopy, spectroscopy, flow cytometry, quantum technologies, and LiDAR. In these applications even low-level unwanted spectral components can affect measurement accuracy and system performance, so accurate spectral characterization is needed throughout development and manufacturing.

MatchBox laser platform

Integrated Optics builds ultra-compact laser sources for scientific and industrial customers. Its MatchBox platform is a configurable family of continuous-wave, multi-wavelength, and pulsed lasers designed for integration, combining optics, control electronics, and thermal stabilization in a matchbox-sized housing.

The challenge: detecting weak spectral modes

A key problem for the company is detecting weak spectral modes and other low-level components close to the main laser line. Although much weaker than the dominant emission peak, these components can indicate unwanted laser behavior or reduced spectral purity. For both product development and production quality control, Integrated Optics needed an optical spectrum analyzer with wide wavelength coverage, high wavelength resolution, good sensitivity, and good close-in dynamic range, together with fast and repeatable measurements.

The AQ6373E optical spectrum analyzer

The company chose the Yokogawa AQ6373E optical spectrum analyzer, which covers wavelengths from 350 nm to 1200 nm. The instrument is used in single-mode laser production, spectral filtering, component characterization, quality checks, and new product development. Its close-in dynamic range of about 60 dB lets engineers separate weak spectral modes from the dominant laser peak and assess spectral purity, and so verify that each laser meets the requirements of its scientific or industrial application.

The analyzer’s sensitivity also allows measurements at relatively low optical input levels, which simplifies test setups in production. Because Integrated Optics’ lasers include control electronics and thermal stabilization, precise spectral calibration is needed to maintain performance and stability over a wide temperature range. The AQ6373E’s connectivity options support automated calibration and validation workflows, and its wavelength range lets the company use a single instrument platform across its portfolio of visible and near-infrared lasers.

The customer’s view

Evaldas Pebreza, CEO of Integrated Optics, describes the AQ6373E as a telecom-grade instrument whose performance suits laboratory laser measurements: the company gets the reliability typical of telecom equipment together with the resolution and sensitivity needed to characterize more demanding laser sources.

Results

With the AQ6373E, Integrated Optics performs fast, accurate, and repeatable measurements, identifies weak spectral components, verifies spectral purity, and evaluates new laser designs. The main benefit for the company is greater confidence that each laser meets customer requirements, both in production quality control and in the development of compact laser sources for medical, scientific, industrial, and quantum applications.

The full Integrated Optics case study is available from Yokogawa.

The post Yokogawa AQ6373E validates Integrated Optics’ ultra-compact visible lasers appeared first on Open Electronics.

SentinelOne has expanded its collaboration with Amazon Web Services (AWS) to provide unified AI governance for customers using Amazon Bedrock. The new integration combines SentinelOne’s AI runtime security with Amazon Bedrock AgentCore, so security teams can monitor, manage, and mitigate AI-related risks from a single platform. The stated aim is to narrow the gap between how fast companies deploy AI agents and how fast security teams can protect them.

The AI governance problem

As organizations deploy more AI agents, security teams need to see what those agents do, which data and applications they access, and whether that activity complies with governance requirements. In June 2026, SentinelOne and AWS addressed the first part of this problem with an integration between SentinelOne’s Prompt Security and Amazon Bedrock AgentCore, which added runtime protection to the Amazon Bedrock AgentCore gateway.

From guardrails to governance

The new integration goes beyond runtime guardrails and adds a governance layer for enterprise AI, with four functions:

• AI usage governance: visibility into how AI is used across the organization, both sanctioned and unsanctioned.
• Policy enforcement: real-time enforcement of AI usage policies on AWS and in multi-cloud environments.
• Threat detection: detection and root cause analysis of AI-related security risks.
• Automated remediation: automatic correction of misconfigurations and non-compliant AI code.

The governance layer connects Prompt Security, Singularity Cloud Security, and Singularity AI SIEM in a single platform on Amazon Bedrock, including AgentCore, with machine-speed policy enforcement, threat detection, and autonomous response across the customer’s AI environment.

SentinelOne’s view

According to Melissa K. Smith, SVP, Global Strategic Partnerships and Initiatives at SentinelOne, security teams want real-time visibility into what their AI agents are doing. The integration provides that visibility inside the workflows customers already run on AWS, so that security teams get answers immediately instead of reconstructing them afterwards.

Availability

The integration extends SentinelOne’s Prompt Security and Singularity AI Runtime offerings for companies building on AWS, and strengthens the partnership with AWS, including distribution through AWS Marketplace and joint commercial initiatives. Prompt Security, Singularity Hyperautomation, and SentinelOne AI SIEM are already available on AWS Marketplace. Full availability of the unified AI governance system is expected at AWS re:Invent 2026, at the end of November.

The post SentinelOne expands AWS integration for AI governance on Amazon Bedrock appeared first on Open Electronics.

TDK Corporation has introduced the TDK-Lambda DDSM series, programmable isolated DC-DC converters rated at 120 W and 240 W with wide input and output voltage ranges. The converters can be controlled and monitored remotely in real time via Modbus over USB, using the PowerCMC control and monitoring center. A digital display shows device settings, system alarms, and status information, and the unit can also be operated manually with three buttons on the front panel.

A wide conversion ratio

The DDSM series was designed for a wide conversion ratio. Typical uses include improving voltage stability over long cable runs, stepping up the voltage to support a system load, providing an independent isolated voltage for Power over Ethernet (PoE) requirements, and powering loads that need special operating voltages. Because the input and output ranges are wide, one model can replace several part numbers in stock.

Input, output, and efficiency

The converters accept a nominal input of 12 V to 48 V (10.8 V to 52.8 V maximum range) and deliver a nominal output of 5 V to 55 V (4.7 V to 57 V maximum range). Efficiency under standard conditions is up to 93.0%, which reduces internal heat losses and saves energy in standby. Rated output current is 5 A for the 120 W model and 10 A for the 240 W model. A 24 V/1 A DC-OK signal is accessible from the front panel. The DDSM measures 40.0 mm wide, 116.6 mm deep, and 115.0 mm high, and weighs 500 g.

Safety and EMC approvals

Safety certification covers IEC/EN/UL/CSA 61010-1, 61010-2-201, and IEC/EN/UL/CSA 62368-1 (Ed. 3). The series is CE and UKCA marked under the EU Low Voltage, EMC, and RoHS Directives and UK regulations. The models meet EN 55032 Class B for radiated and conducted emissions and EN 61000-6 for immunity.

Applications

• Factory automation
• General industrial automation
• Semiconductor manufacturing
• Renewable energy generation
• Telecommunications
• Test and measurement

Key features

• Wide input and output ranges
• High efficiency and low standby losses
• Digital control and monitoring
• Front-panel display and programming

Specifications ParameterValue

Series	DDSM
Input voltage range	10.8-52.8 Vdc
Output voltage	4.7-57 Vdc
Maximum output power	120 W or 240 W
Efficiency	Up to 93.0%
Isolation	Input to output: 2200 Vdc
Dimensions (W × D × H)	40.0 × 116.6 × 115.0 mm
Safety certifications	IEC/EN/UL/CSA 62368-1 (Ed. 3), IEC/EN/UL/CSA 61010-1, 61010-2-201

More information is available on the TDK-Lambda DDSM series page.

The post TDK-Lambda DDSM programmable isolated DC-DC converters rated 120 W and 240 W appeared first on Open Electronics.

This development board combines a 3.97-inch touch e-paper display with LoRa and NFC connectivity, all on an ESP32-S3R8 platform. The result is a compact, battery-powered device designed for those building portable IoT objects. The board also integrates a PDM microphone, buzzer, 6-axis IMU, and RTC, so it covers many use cases without needing external hardware.

At its core is the ESP32-S3R8 module with 8 MB of PSRAM and 16 MB of flash. The monochrome e-paper display has a resolution of 800×480 pixels and supports 4 gray levels. The SSD1677 driver manages it, while the FT6336G controller handles capacitive touch. The standout feature is the refresh speed: it goes from 4.71 seconds in quality mode to just 0.07 seconds in fastest mode. This makes it usable even for interfaces that require frequent updates.

Connectivity and integrated sensors

LoRa connectivity is handled by the Stamp LoRa-1262 module with a Semtech SX1262 chip, operating between 868 MHz and 923 MHz. This enables long-distance communication with low power consumption, ideal for remote sensors or mesh networks. NFC communication is managed by the ST25R3916 chip, which supports ISO14443A, ISO14443B, FeliCa, and ISO15693. The board with integrated display can therefore read NFC tags, communicate with smartphones, or function as a badge.

The BMI270 6-axis inertial sensor detects movement and orientation, while the RX8130CE RTC keeps time even when the board is powered off. The PDM microphone and buzzer complete the set: you can add voice commands or acoustic feedback. The board also includes a 1150 mAh battery, rechargeable via USB-C port. Everything fits in a 101 x 61 x 7.95 mm format and weighs only 74.7 grams.

Software and development tools

The board is compatible with Arduino IDE, UiFlow 2, and ESP-IDF. You can develop quickly with Arduino or take advantage of ESP-IDF’s advanced features. The project board is ready to use: just connect it via USB-C and start programming. For projects that require more computing power, you can pair it with additional ESP32 modules.

For those who want to experiment with displays and sensors, the ESP32 module with TFT touch display is a cheaper but less complete alternative. If you are only interested in LoRa connectivity, the RTX 433 MHz Long Range LoRa module can be integrated into custom projects. The choice depends on your needs: this board offers everything in a single package.

The combination of touch e-paper display, LoRa, and NFC in a compact board is rare. The fast display refresh makes it suitable for dynamic interfaces, while the built-in battery makes it portable. Here are the key points:

• E-paper display 800×480 with 4 gray levels and refresh up to 0.07 s
• LoRa with SX1262 for long-distance communication
• NFC with ST25R3916 for tag reading and communication
• BMI270 IMU, RX8130CE RTC, PDM microphone, and buzzer
• 1150 mAh battery and USB-C port for charging
• Support for Arduino IDE, UiFlow 2, and ESP-IDF

Furthermore, the low power consumption of the e-paper display extends battery life, making the board suitable for devices that must run for days. Finally, the generous PSRAM and flash allow handling complex applications without memory issues. If you are looking for a complete platform for portable IoT, this board is a solid choice.

Source: https://github.com/m5stack/M5PaperMono-UserDemo

The post ESP32-S3 board with touch e-paper display, LoRa, and NFC appeared first on Open Electronics.

This development board combines a 3.97-inch touch e-paper display with LoRa and NFC connectivity, all on an ESP32-S3R8 platform. The result is a compact, battery-powered device designed for those building portable IoT objects. The board also integrates a PDM microphone, buzzer, 6-axis IMU, and RTC, so it covers many use cases without needing external hardware.

At its core is the ESP32-S3R8 module with 8 MB of PSRAM and 16 MB of flash. The monochrome e-paper display has a resolution of 800×480 pixels and supports 4 gray levels. The SSD1677 driver manages it, while the FT6336G controller handles capacitive touch. The standout feature is the refresh speed: it goes from 4.71 seconds in quality mode to just 0.07 seconds in fastest mode. This makes it usable even for interfaces that require frequent updates.

Connectivity and integrated sensors

LoRa connectivity is handled by the Stamp LoRa-1262 module with a Semtech SX1262 chip, operating between 868 MHz and 923 MHz. This enables long-distance communication with low power consumption, ideal for remote sensors or mesh networks. NFC communication is managed by the ST25R3916 chip, which supports ISO14443A, ISO14443B, FeliCa, and ISO15693. The board with integrated display can therefore read NFC tags, communicate with smartphones, or function as a badge.

The BMI270 6-axis inertial sensor detects movement and orientation, while the RX8130CE RTC keeps time even when the board is powered off. The PDM microphone and buzzer complete the set: you can add voice commands or acoustic feedback. The board also includes a 1150 mAh battery, rechargeable via USB-C port. Everything fits in a 101 x 61 x 7.95 mm format and weighs only 74.7 grams.

Software and development tools

The board is compatible with Arduino IDE, UiFlow 2, and ESP-IDF. You can develop quickly with Arduino or take advantage of ESP-IDF’s advanced features. The project board is ready to use: just connect it via USB-C and start programming. For projects that require more computing power, you can pair it with additional ESP32 modules.

For those who want to experiment with displays and sensors, the ESP32 module with TFT touch display is a cheaper but less complete alternative. If you are only interested in LoRa connectivity, the RTX 433 MHz Long Range LoRa module can be integrated into custom projects. The choice depends on your needs: this board offers everything in a single package.

The combination of touch e-paper display, LoRa, and NFC in a compact board is rare. The fast display refresh makes it suitable for dynamic interfaces, while the built-in battery makes it portable. Here are the key points:

• E-paper display 800×480 with 4 gray levels and refresh up to 0.07 s
• LoRa with SX1262 for long-distance communication
• NFC with ST25R3916 for tag reading and communication
• BMI270 IMU, RX8130CE RTC, PDM microphone, and buzzer
• 1150 mAh battery and USB-C port for charging
• Support for Arduino IDE, UiFlow 2, and ESP-IDF

Furthermore, the low power consumption of the e-paper display extends battery life, making the board suitable for devices that must run for days. Finally, the generous PSRAM and flash allow handling complex applications without memory issues. If you are looking for a complete platform for portable IoT, this board is a solid choice.

Source: https://github.com/m5stack/M5PaperMono-UserDemo

The post ESP32-S3 board with touch e-paper display, LoRa, and NFC appeared first on Open Electronics.

The ThinkNode M9 from Elecrow is a fully autonomous LoRa mesh communication terminal. You don’t need a smartphone to write messages: the physical QWERTY keyboard and color TFT display are enough. The MeshCore firmware manages the mesh network between devices, so messages travel from node to node without any infrastructure. It’s designed for emergencies, hiking, and outdoor activities where cell coverage is unavailable.

At the heart of the device is an ESP32-S3R8 with a 240 MHz processor, 512KB SRAM, 8MB PSRAM, and 16MB of flash. The Semtech LR1110 LoRa module handles radio transmission in the EU868 and US915 bands. The built-in GPS supports GPS, GLONASS, Galileo, and BeiDou. A 2,300 mAh Li-Po battery powers everything, with an estimated standby time of 2-3 days.

How the LoRa mesh terminal works

The MeshCore firmware coordinates mesh communication between terminals. Each message can be forwarded by other nodes until it reaches its destination. The 37-key QWERTY keyboard sends key press events via I2C. The firmware polls the keyboard in the main loop, so response is immediate. The 2.4-inch TFT display with 320×240 resolution shows the user interface, including messages, network status, and GPS position.

The LR1110 module also includes advanced geolocation features, but the external GPS remains the primary reference for position. Wi-Fi and Bluetooth add connectivity for initial configuration and debugging. The firmware can be updated via USB-C, which also serves for battery charging. The device measures 126 x 67 x 10.3 mm and weighs 123 grams, so it fits in a pocket without any trouble.

The build is solid and compact. Operating temperature ranges from -10°C to 60°C, while storage goes down to -20°C. The price of $74.90 makes it accessible for anyone wanting to experiment with mesh messaging without spending ham-radio money. For those who want to approach the LoRa world starting from a single board, a TTGO LoRa 868 MHz with ESP32 and OLED display does the same job as the ThinkNode: same European band, same microcontroller, onboard display.

Software and customization

The MeshCore firmware is open source and can be compiled with ESP-IDF. Those who prefer simpler environments can use Arduino IDE or MicroPython. The display is managed with LVGL, a lightweight graphics library suitable for small screens. The documentation includes instructions for modifying the interface and adding custom functions.

The QWERTY keyboard is the strong point compared to other LoRa terminals. Many devices require a smartphone to type, but here everything happens on the board. Additionally, multi-constellation GPS support makes positioning reliable even in remote areas. The firmware automatically handles mesh routing, so the user doesn’t need to configure anything complex.

• Off-grid text messaging without a smartphone
• Multi-constellation GPS for accurate positioning
• EU868 and US915 bands selectable via firmware
• Wi-Fi and Bluetooth for configuration and debug
• Rechargeable Li-Po battery with 2-3 days of autonomy

Why choose a standalone LoRa terminal

In emergency situations, cellular networks can fail. A LoRa mesh terminal creates an alternative network among participants. Range depends on the environment: in open field you exceed one kilometer, in the city it decreases but the mesh extends coverage. The ThinkNode M9 is designed for this scenario, with robustness suited to outdoor use.

The physical keyboard avoids the typing errors of touchscreens, especially with gloves. The color display at 200 cd/m² is readable even in sunlight. For those who want to build a similar device from scratch, you’ll need a 2.8-inch TFT touch display and a UBLOX NEO-6M GPS module; the QWERTY keyboard, however, remains the hard part to source. The ThinkNode M9 remains a complete, ready-to-use product.

The project board on GitHub includes schematics, firmware, and assembly instructions. The community around MeshCore is active and releases regular updates. Support for ESP-IDF and Arduino IDE makes the device suitable for both beginners and experts. Finally, the competitive price and compact size make it an ideal companion for those who live outdoors.

Source: https://github.com/meshcore-dev/MeshCore/pull/2942

Related products

• TTGO LoRa32 SX1276 with ESP32 and SSD1306 OLED
• LCD TOUCH SCREEN 2.8 "SPI

The post ThinkNode M9: LoRa mesh terminal without a smartphone appeared first on Open Electronics.