Creality sent us a sample of the Falcon T1 5-in-1 modular laser engraver for review. The company describes it as the world’s first 5-in-1 laser engraving and cutting machine that supports five types of laser modules in a single unit: 20W diode laser 20W, 40W diode laser, 20w fiber laser, 60W MOPA laser, and 5W UV Laser 5W.

The laser heads can be swapped tool-free in just 15 seconds. It works with an HD camera and AI system and can engrave and cut a wide range of materials including wood, leather, acrylic, metal, glass, crystal, and plastic. Diode laser modules are ideal for cutting and engraving wood, leather, and acrylic, the fiber laser module is better suited for marking and engraving metal, the MOPA laser module supports color marking on metals such as stainless steel and titanium, and the UV laser module uses cold processing for high-precision work on glass, crystal, and plastics.

Last February we reviewed the Creality Falcon A1 Pro laser engraver, a laser engraving and cutting machine equipped with a 20W blue diode laser and an HD camera. We were very interested in reviewing their new Falcon T1 model, especially the MOPA laser for color engraving. However, the latter was not available at the time, and instead, the company sent us a kit with a 20W fiber laser module and a 40W diode laser module, along with various materials such as balsa wood sheets, acrylic, anodized aluminum pet name tags, and stone coasters.

In this review, we will cover the specifications, go through an unboxing, show how to set up the machine, and get started with the interchangeable laser modules using various materials. We might receive the MOPA laser module later on, and we’ll publish a second part if we do.

Creality Falcon T1 specifications

The specifications of the laser modules and the Falcon T1 engraver and cutter can be found in the table below.

Unboxing

We received two packages from Creality.

Let’s start by unboxing the larger (and much heavier) box first. Inside, we’ll find the Falcon T1 fully-enclosed laser cutting and engraving machine.

The top of the machine features a “LIFT TO OPEN” cover that can be removed by the user to clean the internal exhaust fan. There is also an unlock/open button to release an installed laser module, and a safety warning sticker advising that the machine should be operated under supervision, along with warnings about the laser beam, moving parts, and the risk of falling objects. The machine body features a dark gray metal structure with rounded corners, offering both strength and a sleek, premium appearance suitable for use in workshops and offices.

One side features a screen mount, an emergency stop button, and a start/pause button, while the other side comes with USB-A and USB-C ports for connecting the machine to a computer.

The top of the rear panel features a specification label identifying it as the Creality Falcon T1 Fiber Laser Cutter and Engraver, with a power rating of 360W, a “voltage” of 24V/15A, a maximum laser radiation output of 20W, a wavelength of 1064±5 nm, dimensions of 463 × 312 × 509 mm, and a weight of 24.82 kg, along with certification marks and laser safety warnings. Underneath is a sticker illustrating the steps for changing the laser module.

On the bottom of the rear panel, we’ll find a power on/off switch, a touchscreen connection port, an exhaust hose connector, a fire suppression system port, a DC jack, a safety key slot, an air purifier connection port, a conveyor belt connection port, and a rotary attachment port. The front cover can be lifted vertically to reveal the working area.
If we look up, we’ll find the (AI) camera and a “field lens” used by the laser modules.

The laser engraver shipped with a range of accessories: a 20W fiber laser module, an interactive control screen, a power adapter, an AC power cord, a slatted panel, a USB Type-A to Type-C cable, a hose clamp, an exhaust duct, an L-shaped positioning block, a hex key set, a cross-type hex wrench, a USB Type-A to Type-C adapter, ten laser calibration cards, a camera calibration card, ten F-theta lens calibration cards, a lint-free cleaning brush, a rubber air blower, a safety key, a dust-free cleaning cloth, a safety manual, and a quick start guide.

The power supply is suitable globally with a 100-240V 50/60Hz, 5A input, and it outputs 24V up to 15A (360W).

The package also includes a material kit to help users quickly get started.

The second package comes with a 40W diode laser module, a user manual for the module, a cleaning cloth, and another material kit with even more samples.

Preparing the Creality Falcon T1

The first step is to attach the touch control panel to the magnetic holder on the machine, then connect the other end of the cable to the port on the back of the machine.

Next, insert the safety key, connect the power adapter to the DC jack, and install the hose clamp on the exhaust hose to connect it to the exhaust port before securing it.

Our kit comes with two laser modules: “Fiber – 20W” and “Blue diode – 40W”. Let’s show how to install a module in the Falcon T1 laser engraver.

Start by turning the emergency stop button clockwise to unlock it, and press the power switch to turn on the laser engraving machine. After that, the Z-axis will automatically move to its highest position.

This will reveal the slot for the laser module with the connectors and an opening for the lens (protected by a cap in the photo below). At this point, turn off the machine and remove the cap.

Do the same on the laser module by removing its red cap.

We can now install the laser module onto the main unit by pushing it in until we hear a “click” sound, which indicates that the module is securely locked into place.

If later on, you’d like to change the laser module, press the unlock button to release the module with one hand, while pulling the handle on the back of the module with your other hand.

After you’ve worked with a module for a while and want to change to another laser module, press and hold the Start button for 3 seconds to activate the “laser module Quick-Replacement” mode. Once the laser module section has moved up, turn off the machine to change the module.

An alternative way is to select Replace Module on the touch panel and follow the instructions.

A third method to replace the laser module is to tap the Focus button on the screen, then tap on the Home icon to move the Z-axis back to the Home Position. After that, press the unlock button on the top of the machine and proceed to change the module.

Once the module is installed, pull the yellow hanging tab downward to remove the lens protective cover inside the machine.

When turning on the machine for the first time, the screen will prompt you to select the language and set the time zone. Various menu icons are available on the left side. An important setting for the laser engraving and cutting machine is the Calibrate section, especially since you must recalibrate the laser every time the laser module is changed.

To do so, click the settings icon to enter the Calibrate menu, and then you can calibrate the laser module, auto-focus, and the field lens.

During calibration, the “Stop on Cover Open” safety feature will be temporarily disabled while the laser beam will be active. The Falcon T1 is safe to use during normal use since it’s a fully enclosed machine and won’t work if it’s not fully closed. However, you should take precautions during the calibration process, such as potentially wearing safety goggles, not looking directly at the laser beam, and ensuring that no unauthorized persons are near the machine.

• Laser module calibration is used for first-time laser use, uneven engraving depth, incorrect focus position, and abnormal autofocus system operation.
• Autofocus calibration is used when the measurement of the workpiece base height is inaccurate, or the engraving depth is uneven.
• Field lens calibration is used when changing the lens, when the engraved pattern becomes distorted, or when the workpiece size deviates from the specified dimensions (after completing the laser module calibration).

Once you are familiar with it, the calibration process takes 3 to 4 minutes. You’ll also need to use (scrap) material each time, at least for the laser calibration, and the panel for the lens calibration can be reused.

Adding the Creality Falcon T1 to the Falcon Design Space app

The Creality Falcon T1 laser engraving machine can be used in online mode, offline mode, and via the cloud. Creality provides the Falcon Design Space app for Android and iOS, and a desktop version for macOS and Windows; no Linux support. We’ll mostly follow the instructions found on the wiki and use an Android smartphone (Redmi Note 14 Pro 5G) and a Windows PC during our review.

To get started, let’s download and install the Falcon Design Space Mobile app v1.8.3 for Android, then tap Add Device to add the T1 machine.

Before pairing the machine, make sure the Falcon T1 is properly connected to the Wi-Fi network. On the machine’s screen, tap the Robot icon and go to Account -> Scan QR Code, then use the smartphone app to scan the QR Code to log in and pair the device. Once the connection is successful, the machine will appear in the Device List with the status “Connected”.

We are now ready to use the Falcon T1.

Testing the Creality Falcon T1 with the 20W fiber module

We’ll start with the 20W fiber module. In our first test, we’ll tap the “Experience Now” button, select an image, the Line Art Filter, and the material.

Creality provides a table for each module with recommended laser power, speed, number of passes, and interval, including for the 20W fiber laser. However, we don’t need to use it here, since the app automatically selects the optimal parameters for the job.

We used a 150×150 mm paper cardboard in our first test with the following parameters:

• Layer: Picture engraving
• Speed: 2,000 mm/min
• Power: 80%
• Pass: 1

Place the workpiece into the machine, then select Autofocus -> Frame -> Start Processing. The files will be sent to the laser machine. Press the Start button on the touch panel to proceed with the job. You can also monitor the engraving process in the app from the camera.

It took 6 minutes and 42 seconds.

Let’s now switch to a computer and install Falcon Design Space v1.9.6 for Windows. The Falcon T1 can connect to the host through the provided USB-C to USB-A cable or Wi-Fi. We used both, and no problem with either.

For the first use, we must go through the Camera Calibration step.

We placed the provided ArUco calibration card inside the machine so that the software can detect the marker positions and calibrate the camera.

We then replaced the card with some material (black paper cardboard), the machine engraved four points, and on the camera alignment page, we manually selected the center of the four circles so that the camera image accurately matches the actual coordinates of the machine’s working area.

We are now ready to test engraving from the Windows program. We used aluminum dog tags (51 x 36 x 0.7 mm).

We set the parameters as follows:

• Fill engraving, speed: 4,000 mm/min, power: 50%, pass: 10
• Line engraving: speed: 4,000 mm/min, power: 50%, pass: 10

It took 1 minute and 32 seconds for all three tags.

Fiber lasers can also be used for metal coin engraving, and we had limited success when trying with the LaserPacker LP5 last year. Let’s try again with one of the blank commemorative coins provided in the kit with a 2.7mm thickness and a 40 mm diameter.

The first time we tried with one pass, but it was rather dim/pale, so we increased that to 10 passes.

Parameters:

• Line engraving, speed: 500 mm/min, power: 60%, pass: 10
• Fill engraving 1, speed: 500 mm/min, power: 60%, pass: 10
• Fill engraving 2, speed: 500 mm/min, power: 60%, pass: 10

This job took 11 minutes and 55 seconds. The result is decent for this 2D design, although the motifs at the top are not quite as dark.

Using the Creality Falcon T1 with its 40W diode laser module

Let’s now switch to the 40W diode laser module. We’ll use the parameter table on the Creality website to select the recommended speed, laser power, and passes.

The first test will be with an aluminum business card (95 x 50 x 1 mm).

Parameters:

• Image engraving, speed: 2,000 mm/min, power: 100%, pass: 1
• Line engraving, speed: 5,000 mm/min, power: 100%, pass: 1

It just took 29 seconds.

Let’s now switch to a cutting (and engraving) job using 3mm plywood and the following parameters

• Fill engraving 1, speed: 800 mm/min, power: 60%, pass: 1
• Fill engraving 2, speed: 800 mm/min, power: 60%, pass: 1
• Line cutting, speed: 16 mm/min, power: 60%, pass: 1

The workpiece fitted in a 68 x 68 mm square, and it took 1 minute and 9 seconds to complete the job.

We also tested semi-transparent red acrylic with 3mm thickness with the following parameters.

• Fill engraving, speed: 83 mm/min, power: 20%, pass: 1
• Line cutting, speed: 2 mm/min, power: 60%, pass: 3

It’s a larger 140 x 120 mm design, and the cutting and engraving process took 5 minutes and 34 seconds.

Conclusion

The Creality Falcon T1 5-in-1 laser engraver offers a modular design supporting five different laser modules, including diode (20W and 40W), fiber, MOPA, and UV. This enables it to be an ultra-versatile machine handling wood, acrylic, metals, specialty materials, color engraving, internal 3D engraving, and more.

We didn’t receive the more exotic MOPA and UV laser modules enabling color and internal 3D engraving for this review, but still conducted real-world tests with both the 20 fiber laser and 40W diode laser modules. We found the software easy to use on mobile or desktop, and the machine delivered quality pieces with decent speed and accuracy, as well as safety thanks to a fully enclosed design. The T1 machine also features an AI camera system, automatic focusing, and convenient connectivity (USB and WiFi), making it easy to get started. It is well-suited for professional users, product manufacturers, and anyone seeking a versatile all-in-one laser machine capable of handling multiple types of work. We might receive a MOPA or UV laser module, in which case, expect a second part of the review soon.

We’d like to thank Creality for providing the Falcon T1 laser engraver along with 20W fiber laser and 40W diode laser modules for review. The Creality Falcon T1 with the 20W fiber module is sold for $2,429 on Amazon or the Creality store, and you can also go the laser module page to add the $1,259 40W diode laser module and/or the $809 20W diode laser module. The $3,059 60W MOPA module and $2,339 5W UV module are not available yet, except as part of a T1 laser engraver + UV module bundle for the latter. Once all modules are available, a complete Creality Falcon T1 5-in-1 laser engraver kit should cost close to $10,000, although I don’t see why the 20W diode laser model would be needed if you already own a 40W diode laser module.

CNXSoft: This article is a translation – with some additional insights – of the original review on CNX Software Thailand by Suthinee Kerdkaew.

The post Creality Falcon T1 review – A 5-in-1 modular laser engraver tested with 20W fiber and 40W diode laser modules appeared first on CNX Software - Embedded Systems News.

NanoPi R28S is yet another tiny dual Gigabit Ethernet single board computer or router (with a metal enclosure), and the second from FriendlyELEC to implement a USB-C console port, often requested by our readers, after the NanoPi R2S Plus launched in 2024.

At just 49.5 x 49.5 x 24 mm in its metal case, the Rockchip RK3528A-powered router also features 1GB LPDDR4, a microSD card slot and an eMMC module socket for storage, WiFi 6 and Bluetooth 5.3 wireless connectivity. The USB-C ports are only used for power, eMMC firmware upgrades, and debugging, and the system doesn’t feature any USB port for data connectivity, nor any GPIO header for expansion. In other words, it’s mostly useful for networking (firewalls, soft routers), as well as IoT gateway applications.

NanoPi R28S specifications:

• SoC – Rockchip RK3528A
  • CPU – Quad-core ARM Cortex-A53 @ 1.8 GHz
  • GPU – Arm Mali-450 GPU with support for OpenGL ES1.1, ES2.0, and OpenVG 1.1 APIs
  • VPU – 4K H265/H264 60fps decoding
• System Memory – 1GB LPDDR4
• Storage
  • MicroSD Card slot up to 128GB
  • eMMC module connector
• Video Output – N/A
• Networking
  • 2x Gigabit Ethernet RJ45 ports via RTL8211F and RTL8111H controllers
  • WiFi 6 and Bluetooth 5.3 with IPEX antenna connector; the module is blurred in the photos, but I suspect an AIC8800 module.
• USB
  • USB 2.0 Type-C port for eMMC upgrade and 5V power input
  • USB Type-C port for UART debug console
• Debugging – Via USB-C port above only
• Misc
  • User and MASK buttons for eMMC upgrade
  • SYS LED and 2x user LEDs
  • 2-pin 1.27/1.25mm battery connector for low-power RTC (HYM8563TS)
• Power supply – 5V/2A via USB-C port
• Dimensions and weight
  • PCB – 45 x 45 x 1.2 mm | 14 grams
  • Enclosure – 49.5 x 49.5 x 24 mm | 77.5 grams
• Temperature Range – 0°C to 70°C (or 80°C depending on where you read…)

FriendlyELEC lists support for a range of operating systems including Alpine Linux 3.23, Buildroot – Weston (Wayland), Debian 13 Core, FriendlyWrt 25.12 (OpenWrt fork), OpenMediaVault 8.0.6 (probably not super useful with this board…), Proxmox VE 8.2.7, and Ubuntu 24.04 Core, all based on Linux 6.1 LTS and U-boot 2017.09. Additional technical information and download links for the OS image can be found in the wiki.

It’s the third Rockchip RK3528A board from FriendlyELEC, but the other NanoPi NEO3 Plus serves other purposes with a single Gigabit Ethernet port, a USB 2.0 port, and a GPIO header, and the previous-generation NanoPi Zero2 offers similar features. Other Rockchip RK3528A-based routers with multiple Gigabit Ethernet interfaces include the Radxa ROCK E20C (dual GbE + USB-C debug), the Radxa E24C (quad GbE + USB-C debug), and the Banana Pi BPI-M1S, which is more of a complete SBC with HDMI video output, GbE + 2.5GbE networking, wireless, and so on.

FriendlyELEC sells the NanoPi R28S SBC for $20, and the router with a CNC-milled metal case for $29 on the same page. They usually add various accessories like eMMC flash modules and a power supply as options, but that’s not the case for now, so you’ll need to browse their shop if you need any of those.

The post NanoPi R28S – A tiny dual GbE Rockchip RK3528A wireless SBC and router with a USB-C console port appeared first on CNX Software - Embedded Systems News.

Seeed Studio’s reTerminal Sticky is a 3.97-inch “AI-powered” magnetic touch ePaper display powered by an ESP32-S3 microcontroller, which serves as a sticky-note alternative, an e-reader, a live dashboard, or a home automation controller.

The 800 x 480 resolution display comes with 8MB PSRAM, 32MB Flash, a microSD card slot, and next/previous page buttons. It’s “AI-powered” through a microphone and an AI Voice button leveraging the ESP32-S3 Edge AI capabilities for a voice-to-note feature. It also features a temperature and humidity sensor and a 3-axis accelerometer, as well as a built-in magnet and a magnetic ring for easy mounting.

reTerminal Sticky specifications:

• SoC – Espressif ESP32-S3R8
  • CPU – Dual-core Tensilica LX7 microcontroller up to 240 MHz with vector instructions for AI acceleration
  • Memory – 8MB PSRAM
  • Wireless – 2.4 GHz WiFi 4 and Bluetooth 5.0 LE
• Storage
  • 32 MB flash
  • MicroSD card slot (SanDisk recommended)
• Display
  • 3.97-inch monochrome (B&W) ePaper
  • 4-level grayscale
  • 800 × 480 resolution
  • Capacitive touchscreen
• Audio – Single PDM microphone
• Sensors
  • Temperature & humidity sensor
  • 3-axis accelerometer
• Misc
  • Buttons: AI Voice, Page Up/Previous, and Page Down/Next
  • Buzzer
  • Internal reset button
  • Magnetic mounting: built-in magnet for metal surfaces and accessories for non-metal surfaces
• Power
  • Power Input – 5V/1A via USB-C port
  • Battery – 750 mAh
  • Standby time – Average 7 days (actual battery life may vary)
• Dimensions – 106 × 65.5 × 7.3 mm
• Weight – 70 grams

By default, the reTerminal Stick ships with firmware compatible with the Seeedash app for Android and iOS, mostly designed to facilitate uploading notes to the display, and can also be used to display photos or monitor your AI agent. The company says that only English and Simplified Chinese are supported for now, but they are working on more languages. Additional information about the hardware and firmware can be found on the documentation website.

However, one of the key selling points of the Sticky is its versatility. Seeed Studio highlights support for open-source firmware projects, namely Crosspoint for e-readers, TRMNL for live information dashboards, and ESPHome or OpenDisplay for integration with Home Assistant. I actually mentioned the Seeed Sticky when writing about the CrossPoint-compatible Xteink X4 Pro 4.3-inch touchscreen e-reader last week. I just didn’t realize it was an upcoming device at the time.

Seeed Studio told CNX Software the first 500 units would be available at an early-bird price of US$39.90. Sadly, that email sat in my inbox for a little too long, and the reTerminal Sticky can now be purchased for $49.90.  It should eventually become available on the company’s AliExpress store. The hands-on video by Gavin F below mostly goes through the default firmware, and he also tried CrossPoint firmware at the end.

The post reTerminal Sticky 3.97-inch touch ePaper display is supported by four open-source firmware projects (so far) appeared first on CNX Software - Embedded Systems News.

Maker Go ESP32-C5-EVB is an ESP32-C5 board with four relays, four opto-isolated inputs, RS-485 communication, and support for Wi-Fi 6, Bluetooth Low Energy 5.0, and Thread/Zigbee connectivity.

It’s a variant of the Maker Go ESP32-C6-EVB released in 2024 that replaces the ESP32-C6 module with an ESP32-C5 module and a GPIO header with a terminal block for RS-485. The Maker Go ESP32-C6-EVB was mostly a clone of the Olimex ESP32-C6-EVB, and while Olimex already released its own ESP32-C5-EVB, it has a different dual-channel relay design.

Maker Go ESP32-C5-EVB specifications:

• Wireless module – DOIT ESPC5-32 module
  • SoC – Espressif Systems ESP32-C5
    • CPU
      • Single-core 32-bit RISC-V processor @ up to 240 MHz
      • Low-power RISC-V core @ 40 MHz acting as the main processor for power-sensitive applications
    • Memory – 384 KB SRAM on-chip, support for external PSRAM
    • Storage – 320 KB ROM, support for external flash
    • Connectivity
      • Dual-band 802.11ax WiFi 6 in the 2.4GHz and 5 GHz bands, with 802.11b/g/n WiFi 4 standard support for backward compatibility
      • Bluetooth 5.0 Low Energy (LE)
      • 802.15.4 radio for Zigbee 3.0, Thread 1.3, and Matter
  • Storage – 4MB SPI flash
  • PCB antenna
• USB – USB Type-C port for programming and JTAG debugging
• I/Os and expansion
  • 4x relays (10A/240VAC) with 3-pin terminal blocks
  • 4x optoisolated inputs for DC voltage up to 30VDC
  • RS485 terminal block
  • 2x UEXT connectors
• Debugging/Programming – ESP-DEBUG connector suitable for ESP-PROG
• Misc
  • Reset and user buttons
  • User LED
• Power Supply – 4.5 – 40V DC input via DC jack or 2-pin terminal block
• Dimensions – 12.2 x 8.7 mm

The company provides PDF schematics and Arduino sketches for RS485, relays, and opto-isolated inputs on a wiki on ICBBuy, where the board is also known as ESP32-C5-EVBPlus.

A DIN rail mountable shell/enclosure is also offered with the board, and it looks to be the same as the one used with the previous ESP32-C6-EVB model.

I first found the Maker Go ESP32-C5-EVB on AliExpress for $21.15. However, that’s for the board only, and the company sells the DIN rail enclosure for $7.12. Alternatively, you’ll find the board and a board+shell combo on Tindie for $14.95 and $20.95, respectively.

The post Maker Go ESP32-C5-EVB board features RS-485 interface, four relays, four opto-isolated inputs appeared first on CNX Software - Embedded Systems News.

Earlier this month, the P4 KVM project was brought to my attention. It provides IP KVM functionality to a Waveshare ESP32-P4-WIFI6-POE-ETH board when paired with a Toshiba TC358743 HDMI-to-CSI adapter.

However, it doesn’t work with the new ESP32-P4 V3.x variants featuring a proper H.264 video encoding implementation, and the developer (Jonathan Rowny) says it’s just a proof-of-concept with many bugs and no security, so I decided to skip it. The good news is that he also encouraged others to fork the project, and Evgenij Spitsyn did just that with the release of the ESP-KVM project, which supports both MJPEG and H.264 encoding and a range of other features.

ESP-KVM highlights:

• Video capture, resolution changes, and MJPEG or H.264 encoding
• Keyboard, absolute and relative pointer, media keys, copy/paste
• Touch mode for phones and tablets (trackpad, on-screen keyboard)
• Multiple viewers, one in control at a time with takeover
• HTTPS, login, physical password reset
• Firmware updates over the network, with rollback
• Thermal protection
• Virtual media — boot the target from a disk image (microSD card or on-chip flash)
• Guessing the target’s OS from how it enumerates USB
• User macros — replayable key sequences
• Wake-on-LAN (no ATX wiring)
• Static IP or DHCP
• ATX power control (power/reset buttons, power-LED sensing)

Some caveats include that HDMI audio is not implemented yet, and the developer explains that apart from HTTPS login and TLS, the security of the solution has not been evaluated thoroughly. For the latter reason, it’s recommended to use it on a LAN or behind a VPN only.

Evgenij uses slightly different hardware, relying on a Waveshare ESP32-P4-ETH board ($20 on Amazon) or an ESP32-P4-Function-EV-Board, and a Geekworm C790 MIPI CSI-to-HDMI board ($23.90), as well as an optional 2-channel PC817 optocoupler module for power and reset buttons/status ($1.30 on AliExpress). The BoM cost is around $45, meaning you won’t save anything considering the price of entry-level IP KVMs these days, but it’s still an interesting project. It should also work with other ESP32-P4 boards with MIPI CSI, USB, and Ethernet or WiFi for connectivity.

As I started wiring this article, I assumed boards based on the new ESP32-P4 V3.x microcontrollers were more common, and H.264 was properly implemented. However, the ESP-KVM project status is the same as the P4 KVM project: 20 FPS with MJPEG and 7 FPS with H.264. That’s because Evgenij still used a board with an ESP32-P4 V1.3 chip that requires an additional color space conversion step that won’t be required with V3.x. As a side note, I noticed that Wireless Tag offers an ESP32-P4 V3.2 module earlier today, so these are definitely coming very soon.

You’ll find all resources to get started with the ESP-KVM project on its website and a GitHub repository with the code and additional documentation. You don’t even need hardware to test it, since there’s a live demo to give you a feel for the user interface.

Via Hackaday

The post ESP-KVM – An open-source IP KVM solution based on ESP32-P4 RISC-V MCU appeared first on CNX Software - Embedded Systems News.

Running LLMs locally on an SBC or mini PC is a good way to keep your data private, but they typically require a conventional operating system, which uses a good portion of system memory. NightRun is an experimental open-source project that addresses this issue by booting a machine directly into a local LLM from a USB drive or microSD card, without loading a conventional OS.

By removing the operating system, NightRun makes more RAM and memory bandwidth available for AI inference. The runtime is written in Rust, and during boot it loads a quantized model (1.3 GB to 2.4 GB) directly into RAM while verifying its CRC-32 checksums. After the model is loaded, storage is “sealed,” meaning any later attempt to read from the disk will trigger a hard fault.

NightRun key features:

• Hardware support – Supports 64-bit x86_64 UEFI PCs (Secure Boot disabled) and Raspberry Pi 5 via USB or microSD card
• Supported models – Supports Llama 3.2 1B (4GB RAM), Llama 3.2 3B and Granite 4.1 3B (6GB RAM), and Qwen3 4B Instruct 2507 (8GB RAM) models converted to the .nrm format
• Optimized inference – Uses AVX2/FMA/F16C kernels on x86 and NEON kernels on Raspberry Pi
• Zero-Copy design – Runs quantized Q8_0, Q4_K, and Q6_K models directly from RAM without dequantization
• Built-in UI – Includes a framebuffer chat interface with live stats, prompt editing, and conversation history

It might be easy to think that NightRun is a stripped-down Linux distribution designed to save memory, but that is not the case. It is a UEFI application that runs without a conventional operating system, so there is no Linux kernel, scheduler, or network stack underneath. Instead, it relies on UEFI Boot Services to handle keyboard input, drive the display framebuffer, access storage, and even control the Raspberry Pi 5 cooling fan.

In terms of performance, the developers say decoding speed is mainly limited by memory bandwidth. In an 8-core x86 QEMU virtual machine, Llama 3.2 1B (Q8_0) reaches around 20 tokens per second during decoding. On a real 8GB Raspberry Pi 5, Granite 4.1 3B (Q4_K_M) generates about 3.0 tokens per second using the early pre-sdot NEON kernels.

We have previously written about Binh Pham’s LLMStick that uses C/C++ inference engines like llama.cpp deployed on top of lightweight Linux distributions (such as Alpine or DietPi) to free up memory. Compared to that, NightRun is a complete, production-ready UEFI runtime appliance with parity-tested tokenization specifically designed for consumer hardware.

The majority of the code for the NightRun project was written using Claude Code with the Fable 5 model and released under the MIT license on GitHub. Installation is handled by a Linux shell script that downloads a verified model and flashes the bootable image. The developers have implemented various safety checks in the installer to explicitly type FLASH /dev/sdX to prevent accidental overwriting of system drives. More details on the technical implementation can be found at nightrun.io.

The post NightRun UEFI application boots a local LLM on Raspberry Pi 5 and x86 PCs without an OS appeared first on CNX Software - Embedded Systems News.

Luxonis has introduced a prototype of the M8 Controller Box, an industrial I/O expansion module for OAK4 AI vision edge AI cameras powered by a Qualcomm DragonWing QCS8550. Connected through the camera’s M8 interface, it adds GPIO, relay, CAN, serial, and USB connectivity, allowing OAK4-based systems to interface directly with sensors, actuators, PLCs, motors, and other industrial equipment for machine vision and automation applications.

The prototype module features 16 GPIOs, four 16A relays, CAN 2.0, RS232, two USB 2.0 Type-A ports, an isolated strobe output, onboard audio, three buttons, and three status LEDs. It is powered directly from the OAK4 M8 connector, and supports DIN rail and VESA mounting.

Luxonis M8 Controller Box specifications:

• MCU – Raspberry Pi RP2040 dual-core Arm Cortex M0+ @ 133 MHz with 264KB SRAM
• Host interface
  • M8 connector for connecting Luxonis OAK4 cameras (standalone mode only)
  • M8 passthrough port
• Audio
  • 3.5mm headphone/AUX jack with mic input
  • Onboard buzzer/speaker and
  • Speaker auto-disconnects when 3.5mm port is used
• USB – 2x USB 2.0 Type-A downstream ports (shared 500mA current limit)
• Serial – RS232 interface
• Expansion – 16x GPIOs (3.3V logic level, reverse voltage and ESD protection; 50mA max combined current)
• Other I/O
  • Isolated strobe driver output (supports 5–24V strobe lights)
  • 4x SPDT / Latching power relays (up to 16A, 400VAC max switching voltage)
  • Supports both CAN 2.0A and 2.0B, with up to 1 Mbps data rate
• Misc
  • 3x physical tactile buttons
  • 3x status LEDs
• Power
  • 5V supplied directly via the OAK4 M8 connector (no external power supply required for basic operation)
  • 2x 3.3V power outputs for external sensors (450mA max combined current)
• Dimensions – 123 x 78 x 29 mm (W x H x D)
• Weight – 300 grams
• Mounting – VESA 75 mm mounting holes; DIN rail mountable (clip included)

The M8 Controller Box is currently in its “PR1” prototype phase, meaning it provides early access to developers and professionals before the final production units are released. Luxonis notes that CAN Bus, USB-A, UART/RS232, LEDs, buttons, relays, and GPIOs are fully supported in this hardware revision, while the strobe output is only partially supported. The controller also supports FSync (Frame Synchronization) splitter functionality through the M8 interface for synchronizing multiple cameras; additional features like M8 passthrough and ADC/I2C/SPI interfaces will be added later.

On the software side, GPIO and peripheral control are exposed via the USB-to-interfaces (u2if) protocol, and applications are run directly on the OAK4 device using the oakctl CLI tool. Additionally, the oak-examples GitHub repo includes Python examples covering relay control, CAN communication, GPIO, LED and audio control, and barcode detection.

Previously, we have written about various edge AI vision cameras for industrial applications, including the Arducam KingKong and EDATEC ED-AIC2020 industrial cameras, as well as Luxonis’s OAK Thermal PoE AI camera. These products combine AI processing and cameras in a single device, but the new M8 Controller Box takes a different approach. It acts as a dedicated hardware breakout module, giving OAK4 cameras direct physical control over external industrial environments, bypassing the need for a separate PLC or host PC entirely.

The Luxonis M8 Controller Box (PR1 prototype) is available for $249 on the company’s store. The package includes the controller box, an M8 cable, a DIN rail mount clip, and plug-in screw terminals. As a prototype evaluation kit, it is designed for professional and development use and is not certified for end-user or consumer applications. Some additional information is available on the company’s blog.

The post Luxonis M8 Controller Box adds industrial I/Os, CAN Bus, and relay control to OAK4 AI vision cameras appeared first on CNX Software - Embedded Systems News.

InHand Mo 62A and Mo 68A are two credit card-sized Edge AI single board computers (SBCs) powered by Texas Instruments AM62A and AM68A processors, delivering respectively 2 TOPS and 8 TOPS of AI performance.

Designed for AI Vision applications, the Mo 62A and Mo 68A SBCs support MIPI CSI camera input, Raspberry Pi HAT expansion boards, integrate 7th-generation ISP image processing capabilities, and handle H.264 / H.265 video encoding and decoding. However, they offer a different range of ports, as you’ll see in the specifications below. Target applications include industrial equipment, access control systems, robots, smart cameras, and visual inspection terminals.

InHand Mo 62A

Mo 62A specifications:

• SoC – Texas Instruments AM62A74 SoC
  • Application processor – Up to four Arm Cortex- A53 cores @ 1.4GHz
  • Real-time cores
    • 1x Cortex-R5F @ 800MHz (MPU Channel with FFI)
    • 1x Cortex-R5F @ 800MHz to support Device Management
  • 3D GPU – None
  • VPU – 4Kp30 H.265 and H.264 codec
  • C7xV-256 + Deep Learning Accelerator up to 2 TOPS
• System Memory – 2GB, 4GB (default), or 8GB LPDDR4
• Storage – MicroSD card slot
• Video Output – Micro HDMI port
• Audio
  • 3.5mm audio jack
  • Digital audio via HDMI
  • PCM via 40-pin GPIO header
• Camera Interface – 4-lane MIPI CSI-2 connector
• Networking
  • Gigabit Ethernet RJ45 port
  • WiFi 5 and Bluetooth 4.2 with on-board snap-on antenna
• USB – 4x USB 2.0 Type-A ports
• Expansion – 40-pin Raspberry Pi-compatible GPIO header with I²C, SPI, UART, PCM…; compatible with most Pi HAT expansion boards
• Debugging – TTL UART
• Security – Secure Boot, TrustZone, OP-TEE, Hardware AES-256
• Misc
  • Reset button
  • Power and User LEDs
  • 4-pin fan connector; active fan optional
• Power Supply
  • 5V/5A via USB-C port
  • A 4-pin PoE header can be seen in the photos of the board, but it’s not documented.
• Dimensions – 85 x 56mm
• Weight – 47 grams
• Temperature Range – 0 to 50°C

The company provides a Debian 13.2 (Embedded Linux) image based on Linux 6.12. AI & vision capabilities are supported through TI TIDL AI Runtime with support for TFLite/ONNX, V4L2 camera framework, DRM/KMS display frameworks, and TI EdgeAI SDK. InHand also provides DeviceLive IoT device management platform for remote monitoring, edge app management, and remote maintenance of a fleet of boards.  You’ll find some documentation and software development resources for the Mo 62A board on GitHub.

InHand Mo 68A

Mo 68A specifications:

• SoC – Texas Instruments AM68A SoC (TDA4VE / J721S2 family)
  • Application cores – 2x Arm Cortex-A72 cores at up to 2.0 GHz with 1MB shared L2 cache with ECC
  • Real-time cores – 4x Cortex-R5F real-time cores at up to 1.0 GHz
  • GPU – Imagination Technologies BXS-4-64 up to 800 MHz
  • VPU – 4Kp60 H.265 and H.264 codec
  • AI accelerators
    • Deep Learning Accelerators up to 8 TOPS
    • Vision Processing Accelerators (VPAC) with Image Signal Processor (ISP)
• System Memory – 4GB or 8GB (default) LPDDR4
• Storage – MicroSD card slot
• Video Output
  • MiniDP port
  • Up to 2x MIPI DSI via 22-pin connectors (multiplexed with MIPI CSI)
• Audio
  • 3.5mm audio jack
  • Digital audio via mini DisplayPort
  • PCM and I2S via 40-pin GPIO header
• Camera Interface – Up to 2x 4-lane MIPI CSI-2 connectors (multiplexed with MIPI DSI)
• Networking
  • Gigabit Ethernet RJ45 port
  • WiFi 5 and Bluetooth 4.2 with on-board snap-on antenna
• USB – 4x USB 3.0 Type-A ports
• Expansion
  • 40-pin Raspberry Pi-compatible GPIO header with I²C, SPI, UART, I2S, PCM…; compatible with most Pi HAT expansion boards
  • 16-pin PCIe Gen 3.0 FFC connector 
• Debugging – TTL UART
• Security – Secure Boot, TrustZone, OP-TEE, Hardware AES-256
• Misc
  • Reset button
  • Power and Status LEDs
  • 4-pin fan connector; active fan recommended
  • RTC with backup battery
• Power Supply
  • 5V/5A via USB-C port
  • A 4-pin PoE header can also be seen on the board
• Dimensions – 85 x 56mm
• Weight – 53 grams
• Temperature Range – 0 to 50°C

Software support is about the same as for the AM62A board, including a Debian 13.2 OS image based on Linux 6.12, TI TIDL AI runtime, V4L2 camera framework, DRM/KMS display frameworks, and TI EdgeAI SDK. The Debian image appears to be built with Armbian, and Python, C/C++, OpenCV, and GStreamer are also mentioned. The company has yet to share a software repository for the Mo 68A board on GitHub, but some (hardware) documentation is available there.

I’ve already highlighted the differences between the Mo 62A and Mo 68A specifications above with bold and strikethrough text, but the company also provides a more visual side-by-side comparison shown below.

InHand Networks is primarily a B2B company, although they sometimes sell samples of their products online. However, that’s not the case for the “Mo” boards, for which we don’t have any price and availability information. Additional details can be found on the company’s website and in the press release.

Thanks to TLS for the tip.

The post Credit card-sized Edge AI single board computers are powered by TI AM62A (2 TOPS) or AM68A (8 TOPS) SoC appeared first on CNX Software - Embedded Systems News.

NXP has introduced the RW610FML, a compact, low-power, certified Wi-Fi 6 and Bluetooth Low Energy (BLE) 5.4 wireless module based on the company’s RW610 Cortex-M33 microcontroller, the little brother of the RW612 without an 802.15.4 radio.

Designed for smart appliances, home automation, industrial gateways, and medical devices, the surface-mount module combines the 260 MHz Arm Cortex-M33 wireless SoC with 8MB of on-module NOR flash, a built-in antenna, an RF matching circuit, and a crystal oscillator.

NXP RW610FML Specifications:

• Wireless MCU – NXP RW610
  • CPU Core – 260 MHz Arm Cortex-M33 core with Arm TrustZone-M
  • Internal memory – 1.2 MB embedded on-chip SRAM
  • Connectivity
    • Wi-Fi – Dual-band 2.4 GHz and 5 GHz Wi-Fi 6 (IEEE 802.11a/b/g/n/ac/ax) SISO
      • 20 MHz channel bandwidth
      • Up to MCS9 (114 Mbps) peak data rate
    • Bluetooth – Bluetooth Low Energy (LE) supporting 2 Mbit/s high-speed mode, long-range operation, and extended advertising
• Storage – 8 MB NOR flash
• Castellated holes and pads for connection to host board
  • Storage
    • 2x FlexSPI interfaces for external NOR Flash (XIP) and PSRAM
    • SDIO 3.0
  • Display  –  LCD interface
  • Audio – I2S / PCM, Digital Microphone Interface (DMIC)
  • Other I/O
    • ADC, DAC, Analog Comparator (ACOMP)
    • USB, UART, I2C, SPI, GPIOs
    • Universal Subscriber Identity Module (USIM)
    • Wireless Coexistence Interface 2 (WCI-2)
    • Packet Traffic Arbitration (PTA)
  • Debugging – JTAG, Serial Wire Debug (SWD)
• Security – NXP EdgeLock technology with secure boot, secure debug, secure firmware updates, secure life cycle management, hardware cryptography, and a physically unclonable function (PUF) for secure key management
• Misc
  • Integrated PCB antenna
  • Timers – SCTimer / PWM, Micro-tick timer (UTICK)
• Dimensions – 25.5 × 18 × 3.16 mm (Surface-mount LGA module form factor)
• Weight – Approximately 1.51 grams
• Temperature Range – Operating: -40°C to 85°C (Industrial Grade); storage: -45°C to 95°C
• Moisture Sensitivity Level (MSL) – 3

The built-in FlexSPI interface features an on-the-fly decryption engine, allowing the MCU to securely execute code in place (XIP) from external flash memory.

To simplify development, the company also offers the FRDM-RW610FML development board, which provides access to the module’s GPIOs, serial interfaces, onboard flash memory, and an MCU-Link debug probe for evaluation and prototyping. No other details are available at this stage, but the image above, provided by NXP, looks very similar to the FRDM-RW612 development board.

The module runs FreeRTOS or Zephyr OS RTOS, and development can be done through the MCUXpresso IDE, MCUXpresso for Visual Studio Code, IAR Embedded Workbench, or Keil MDK.

The RW610FML module and the FRDM-RW610FML development board are currently listed as preproduction products, and at the time of writing, the company has not mentioned anything about pricing. More details, including the preliminary RW610FML fact sheet, can be found on the product page and in the press release.

The post NXP RW610FML Wi-Fi 6 and BLE 5.4 module features RW610 Cortex-M33 wireless MCU appeared first on CNX Software - Embedded Systems News.

VIEWE UEDX12120076-HMD-RB-A and UEDX80800076-HMD-RB-A are 7.6-inch square HDMI capacitive touch displays with 1200×1200 and 800×800 resolution, respectively.

We had previously seen 4-inch square HDMI +USB touch displays with 720×720 resolution for Raspberry Pi or other SBCs, but if you need something larger, VIEWE has now launched 7.6-inch variants with higher resolutions.

VIEWE UEDX12120076-HMD-RB-A/UEDX80800076-HMD specifications:

• Display
  • 7.6-inch MIPI DSI square IPS display
  • Resolution
    • UEDX80800076-HMD-RB-A – 800 x 800
    • UEDX12120076-HMD-RB-A – 1200 x 1200
  • Display driver
    • ILI9881C for UEDX80800076-HMD-RB-A
    • Potentially HX8279 for UEDX12120076-HMD-RB-A
  • Brightness – 1,000 cd/㎡
  • Active area – 135.36 x 135.36 mm
• Touchscreen – Optional 5-point capacitive touchscreen via USB-C port (likely FT6336U touch IC)
• Video Input
  • Mini HDMI 1.4
  • ITE HDMI input to MIPI DSI bridge based on the blurry photo of the controller.
• USB – USB-C port for power and touchscreen
• Misc – LED lifetime: 30,000 hours
• Power Supply – 5V via USB-C port
• Dimensions – 148.24 x 142.56 mm
• Weight – 250 grams
• Temperature Range – Operating: -20 to 70°C; storage: -30 to 80°C

The displays can be connected to any host with a spare HDMI output and USB port, and the company highlights Raspberry Pi single board computers, NVIDIA Jetson developer kits, or PCs with a secondary stats screen running AIDA64.

You’ll need to modify config.txt on Raspberry Pi OS as follows for the 800×800 display:

hdmi_force_edid_audio=1
max_usb_current=1
hdmi_force_hotplug=1
config_hdmi_boost=7
hdmi_group=2
hdmi_mode=1
hdmi_mode=87
hdmi_drive=2
hdmi_cvt 800 800 60 6 0 0 0

The company didn’t provide a config.txt file for the 1200×1200 display, but it should be the same, except for the last line:

hdmi_cvt 1200 1200 60 6 0 0 0

Some documentation is available on the product pages, but note that there are errors, including wrong dimensions given (those are for the 4-inch models), incorrect PDF user manual for the UEDX12120076-HMD-RB-A (points to the 800×800 model PDF), etc.

The UEDX80800076-HMD-RB-A and UEDX12120076-HMD-RB-A 7.6-inch square HDMI USB touch displays can be purchased on AliExpress for $85.29 and $131.78 before taxes and shipping, and you’ll also find them for sale on the aforelinked product pages.

The post 7.6-inch square HDMI displays with USB capacitive touchscreen offered at 1200×1200 or 800×800 resolution appeared first on CNX Software - Embedded Systems News.

Sovol has launched the M1D, an open-source IDEX (Independent Dual Extruder) 3D printer with a fixed dual-extruder architecture and interchangeable toolheads that support up to seven independently heated toolheads (one fixed and six swappable). It comes in two variants: the open-frame M1D Essential and the M1D Advanced, which adds a full enclosure and a chamber heating module capable of reaching up to 60°C.

The M1D uses a patented automatic tool changer for swapping print heads in as little as five seconds and supports Single, Multi-Head, Copy, and Mirror printing modes. It enables multi-material or/and multi-color prints, including adding water-soluble support to a print to remove the need to manually remove support (see video at the end of this post for demo). The printer targets makers, engineers, prototype development, and small-scale production by reducing purge waste while improving printing efficiency and material flexibility.

Sovol M1D specifications:

• Printing technology – FDM (Fused Deposition Modeling)
• Storage – 32GB eMMC internal storage
• Display – 5-inch touchscreen display console
• Connectivity – Wi-Fi, USB
• Build volume
  • Single nozzle (Left/Right) – 300 x 300 x 350 mm
  • 6-Toolhead switching printing (6-Head Mode) – 300 x 300 x 350 mm
  • Dual overlap (7-Head Printing) – 240 x 300 x 350 mm
  • Copy printing mode – 178 x 300 x 350 mm
  • Mirror printing mode – 138 x 300 x 350 mm
• Extruder system – Dual IDEX Tool Changing System
• Nozzle
  • Number of nozzles – 1x fixed + 6x swappable (7x toolheads total) ~5-second automatic tool changes
  • Nozzle diameter – 0.4mm hardened steel (Optional: 0.6 / 0.8 mm)
  • Maximum temperature – 300°C
• Hotend – Full-metal hotend with ceramic heater and 0.4 mm hardened steel nozzle (0.6/0.8 mm optional)
• Heated bed – Textured PEI steel build plate with 800W AC heating (100°C max)
• Chamber
  • Essential – Open-frame; no chamber heating
  • Advanced – Enclosed with active chamber heating (up to 60°C)
• Printing speed
  • Maximum travel speed – 600 mm/s
  • Maximum acceleration – 10,000 mm/s²
  • Maximum volumetric flow Rate – 30 mm³/s (evaluated with Sovol PLA at 260°C)
• Cooling – 1× 5020 part cooling fan, 7× 3010 hotend fans, and 1× 4010 mainboard fan
• Supported Materials
  • Essential – PLA, PETG, TPU, PVA
  • Advanced – PLA, PETG, TPU, PVA, ABS, ASA, PA, PC
  • Filament diameter – 1.75mm
• Filament system – 6-channel Auto Filament System with clog, runout, and tangle detection for six swappable toolheads
• Other features
  • Bed leveling – Non-contact eddy current auto bed leveling with real-time Z-offset compensation
  • Motion control – Accelerometer-based vibration compensation (input shaping)
  • Vision System – Dual-camera auto X/Y toolhead calibration with nozzle cameras
  • Monitoring Camera – 1280 × 720 chamber camera for time-lapse and AI print monitoring (OTA update required for AI detection)
  • Recovery – Power loss recovery option
• Power – 115–240 VAC / 50–60 Hz input; 150W + 300W power system
• Dimensions – 714.57 x 626.2 x 1100 mm

The printer features a dual-camera Auto Vision calibration system that aligns the toolheads up to 2.5x faster than traditional probe-based methods. It also uses real-time mechanical Z-lift and Z-offset compensation, alongside an onboard accelerometer for vibration compensation to reduce ringing and ghosting during high-speed printing. It also features a six-channel automatic filament system and an AI chamber camera to detect foreign objects and failed prints. However, Sovol notes these AI detection features will be activated later via a software update.

On the software side, the M1D utilizes OrcaSlicer slicing software by default, compatible with both macOS and Windows.

The Sovol M1D is currently available on Kickstarter, which has already raised more than $2.7 million from over 1,700 backers with 29 days remaining. Early Bird rewards start at HK$10,184 (~$1,299) for the M1D Essential and HK$12,536 (~$1,599) for the M1D Advanced, excluding shipping. Deliveries are expected to begin in November 2026.

The company is also offering a two-year warranty for the printer itself, along with a three-month warranty for consumables like the nozzle and the PEI build plate. Additionally, there are optional add-ons, including the Sovol SH03 Filament Dryer and heavily discounted PLA and PETG filament bundles.

As with any crowdfunding campaign, backing the project carries some risk, but Sovol has been manufacturing 3D printers since 2019 and has previously delivered products such as the Sovol SV01, which we also reviewed back in 2019; after that, the SV02 came out in 2020 and the SV08 shortly afterward.

The post Sovol M1D hybrid IDEX 3D printer features 6 interchangeable toolheads for multi-color and multi-material prints (Crowdfunding) appeared first on CNX Software - Embedded Systems News.

SensWear is an open-source hardware development platform for wearables based on Nordic nRF54L15 Bluetooth LE SoC and a range of sensors designed for researchers, developers, and companies building smart rings, wristbands, patches, and other custom devices.

The solution is mainly comprised of flexible boards, including the SensWear core board with an nRF54L15 MCU and a 6-axis IMU sensor, and daughter boards with either PPG and optical sensing,  temperature sensing, capacitive touch sensing, or haptic feedback.

SensWear key features and specifications:

• Core mainboard
  • SoC – Nordic Semi nRF54L15
    • CPU
      • Arm Cortex-M33 @ up to 128 MHz
      • RISC-V coprocessor for “software-defined peripheral”
    • Memory – 256 KB SRAM
    • Storage – 1.5MB Non-volatile Memory (NVM)
    • Wireless
      • Bluetooth LE 6.0
      • Thread, Zigbee (IEEE 802.15.4-2020)
      • Matter
      • Amazon Sidewalk
      • Proprietary 2.4 GHz protocols with a new 4 Mbps mode
      • NFC support
  • Sensors
    • Bosch BHI360 six-axis smart IMU and sensor hub
    • Standardized interface for interchangeable daughter boards
      • I2C and interrupt-capable GPIOs
      • Support for daisy-chaining multiple daughter boards
  • Power Supply – Battery charging, power management, and fuel monitoring
  • Dimensions –  TBD (Flexible PCB)
• Daughter boards
  • MAX30101 PPG (Photoplethysmography) and optical sensing for real-time heart rate and SpO₂ monitoring
  • MAX30208 temperature sensing (skin or ambient temperature)
  • MTCH6102 15-channel capacitive touch sensing
  • DRV2605 haptic feedback
  • Supply Voltage – Programmable between 1.2 V and 5.5 V
• Debug board with standard JTAG/SWD interfaces, support for J-Link tools.

SensWear provides access to the complete stack, including hardware design files, Zephyr-based firmware, raw sensor data, Android and iOS mobile applications, and developer SDKs in Python and TypeScript for streaming raw sensor data and developing custom signal-processing or machine-learning applications. Since the solution is modular, developers can also design their own compatible sensor boards without redesigning the main wearable system. Everything has been released under an MIT license on GitHub.

What’s missing right now is a demo of the latest version of the SensWear platform with flexible PCBs. However, we can still take a quick look at software support with a video demo using previous-generation rigid PCBs.

SensWear sells as a complete developer bundle for 849,99 Euros (about $970) with the core board, four daughter boards, a debug board, and, I assume, any required cables, but that’s not explicitly stated. The core board can be purchased separately for 699.99 Euros, and daughter boards for 99.99 Euros.  The price is quite high, probably because it’s a niche developer platform.

The post SensWear – An open-source modular Nordic nRF54L15 development platform for wearables appeared first on CNX Software - Embedded Systems News.

UBPorts has announced the release of Ubuntu Touch 24.04-2.0 based on Ubuntu 24.04. It features an upgraded Morph Browser based on Chromium 134, supports smartphones with notches and rounded corners, a Widevine DRM installer, and more.

Ubuntu Touch 24.04-2.0 highlights:

• Updated browser engine for Morph Browser from Chromium 87 to Chromium 134 for improved compatibility with websites and web applications
• Installer for Widevine DRM with the following note:
  

  Widevine installer is not endorsed by Google. UBports, Ubuntu Touch, or Morph Browser is not a licensed Widevine implementation. Extracting Widevine from another source is considered by some to be in legally grey area. You are responsible for your usage of Widevine installer.

• Support for notches and rounded corners; currently configured on Fairphone 4 and 5, all Volla devices, and several community devices
• Screenshot editor allowing users to crop, rotate, and adjust color; this can be disabled by the user
• Ubuntu Touch now supports printing documents directly from the device; it’s not enabled by default, and you’ll need to go to System Settings > Printing.
• Support for legacy (X11-based) applications from Open-Store is improved
• Network indicator is improved to handle wired connection (e.g., through a dock)
• Improved mobile data connection reliability
• User data encryption for newer devices supporting fscrypt v2
• Added support for Georgian language (keyboard)
• ~2,000 new emojis are added to the emoji keyboard
• Various bug fixes and security updates

One downside of the Ubuntu Touch 24.04-2.0 is its increased system storage space requirements, so the number of supported devices has been reduced. The project released a list of phones supported by the new version of Ubuntu Touch:

• Asus Zenfone Max Pro M1
• F(x)tec Pro1 X
• Fairphone 4 and 5
• Lenovo Tab M10 HD 2nd Gen WiFi / LTE
• OnePlus Nord N10 5G and N100
• Rabbit R1
• Sony Xperia X
• Volla Phone, Phone X, Phone 22, Phone X23, Phone Quintus, and Phone Plinius
• Xiaomi Poco X3 / X3 NFC, Redmi 9 and 9 Prime, Redmi Note 9
• Zinwa Q25 (new)

A few more details, including known issues like the lack of hardware video decoding in the Morph browser, may be found in the announcement. Existing Ubuntu Touch users can upgrade directly from their phone, with a 2-step upgrade: first to 24.04-1.4, then to 24.04-2.0. New Ubuntu Touch users will find instructions for installing Ubuntu Touch on their device at https://devices.ubuntu-touch.io.

Via Liliputing

The post Ubuntu Touch 24.04-2.0 updates Morph browser, adds notch and printing support, and more appeared first on CNX Software - Embedded Systems News.

The DFI X6X-ORN is a fanless, IP67-rated Edge AI computer powered by the NVIDIA Jetson Orin NX/Nano. It is designed for extreme outdoor and industrial environments in maritime, agriculture, construction, and intelligent transportation applications.

The system fully relies on rugged M12 and M20 circular waterproof connectors for all its I/O, including PoE and GMSL2 camera interfaces. It delivers up to 157 TOPS of AI performance when equipped with the Jetson Orin NX 16GB module, complies with MIL-STD-810H shock and vibration standards, and supports an operating temperature range of -20°C to 60°C.

DFI X6X-ORN specifications:

• SoM – NVIDIA Jetson Orin NX/Nano module
  • Jetson Orin NX 16GB – Up to 157 TOPS, 16GB LPDDR5, 10W/15W/25W/ 40W power modes
  • Jetson Orin NX 8GB – Up to 117 TOPS, 8GB LPDDR5, 10W/15W/25W/40W power modes
  • Jetson Orin Nano 8GB – Up to 67 TOPS, 8GB LPDDR5, 7W/10W/25W power modes
  • Jetson Orin Nano 4GB – Up to 34 TOPS, 4GB LPDDR5, 7W/10W / 25W power modes
• Storage – M.2 2242/2280 M-key slot for NVMe SSD
• Display –  HDMI via M20 connector
• Camera – 2x GMSL2 (FAKRA SMB Plug Z code)
• Networking
  • 4x GbE M12 X-code 8-pin ports, including 2x with PoE 802.3af (supporting up to 15W per port)
  • Optional WiFi/Bluetooth via M.2 E-Key socket (USB 2.0 + PCIe x1 signals)
  • Optional 5G/LTE via M.2 Key B slot with Nano SIM card slot (M20)
  • Holes for up to 6x antennas for WiFi, cellular, and OOB
• USB
  • 2x USB 3.0 via M20 Type-A connectors (side)
  • 2x USB 2.0 via M20 Type-A connectors (front)
  • 1x USB 2.0 via M20 Type-A connector (side, for recovery)
• Serial / IO – M23 A-Code connector with
  • RS232 (Tx/Rx)
  • CAN BUS 2.0 Interface
  • 4x GPI + 4x GPO
• Expansion
  • M.2 2242/2280 M-Key socket (PCIe Gen4 x4) for storage
  • M.2 E-Key 2230 slot (USB 2.0 + PCIe x1) for Wi-Fi 6E
  • M.2 B-Key slot for 5G/LTE with Nano SIM
• Management – Optional Out-of-Band (OOB) support
• Misc – Power LED indicator; RTC with battery
• Power Supply – 9-36V DC via M12 B-Code 5-pin power input
• Dimensions – 237 (W) x 194 (L) x 105.3 (H) mm
• Weight – 5.2 kg
• Temperature Range – Operating: -20°C to 60°C; Storage: -40°C to 85°C
• Humidity – 5% to 95% RH (non-condensing)
• Enclosure – IP67 rated, fully sealed
• Shock – IEC-60068-2-27, MIL-STD-810H 514.8C-5
• Vibration – IEC 60068-2-64, MIL-STD-810H 514.8C-5
• Certifications – CE, FCC Class A, RoHS

The system combines GMSL2 and PoE interfaces for connecting long-range cameras, LiDAR, and other vision sensors. It runs Linux with NVIDIA JetPack 6.2 and supports AI vision applications. An optional Out-of-Band (OOB) management module enables remote power control and system recovery, and allows administrators to reset the system even if the operating system becomes unresponsive.

We previously wrote about the Vecow RAC-1000, a rugged Edge AI computer based on the NVIDIA Jetson AGX Orin module. It also comes with an IP67-rated enclosure, waterproof M12 connectors, and GMSL2 camera interfaces. The DFI X6X-ORN offers many of the same rugged features but is designed around the lower-power Jetson Orin NX and Jetson Orin Nano modules.

DFI has not provided pricing information for the X6X-ORN IP67-rated Edge AI computer. More details are available on the product page and in the press release.

The post DFI X6X-ORN – An IP67-rated, rugged Jetson Orin NX/Nano edge AI computer with GMSL2 camera support appeared first on CNX Software - Embedded Systems News.

Waveshare RP2350-POE-ETH is a Raspberry Pi RP2350 board with Ethernet implemented through a W6300 chip, support for optional PoE (Power-over-Ethernet), and two GPIO headers.

We had covered a few Raspberry Pi RP2350 boards with Ethernet from WIZNet, including the W5100S-EVB-Pico2, W5500-EVB-Pico2, and W6300-EVB-Pico2, but none were advertised with PoE at the time. Since then, the 4-pin WizPOE-P1 module changes that, and the Waveshare RP2350-POE-ETH board offers a more affordable option.

Waveshare RP2350-POE-ETH specifications:

• Microcontroller – Raspberry Pi RP2350A
  • CPU
    • Dual-core Arm Cortex-M33 processor @ 150 MHz
    • Dual-core 32-bit RISC-V Hazard3 @ 150 MHz
    • Up to two cores can be used at a time, either 2x Arm, 2x RISC-V, or 1x Arm + 1x RISC-V
  • Memory – 520KB internal RAM
  • Storage – 8KB OTP Storage
• Storage – 16Mbit SPI NOR flash
• Ethernet
  • 10/100Mbps Ethernet RJ45 port via WIZnet W6300 “iEthernet” chip
  • 6-pin PoE connector for optional module
• USB – USB 1.1 C-Type port for power, data, and programming
• Expansion – 2x 16-pin with through and castellated holes
  • • Up to 20x GPIO
    • 2x UART, 2x SPI, 2x I2C
    • 16x PWM
    • 4x 12-bit 500ksps Analogue to Digital Converter (ADC)
    • 3x Programmable IO (PIO) blocks, 12 state machines total
    • HSTX peripheral
    • 1x Timer with 4 alarms
    • 1x Real Time Counter
• Debugging – 3x pads for SWD (8)
• Misc
  • Boot (3) and Reset (4) buttons
  • WS2812 RGB LED (2)
• Power Supply
  • 5V via USB-C port
  • Optional PoE (802.3af)
  • Low-dropout linear regulator (LDO); 3.3V @ up to 800mA
• Dimensions – PCB: 63 x 21 mm; with USB and RJ45 connectors:  65.88 x 21 mm

Waveshare provides support for MicroPython, C/C++, and Arduino programming languages. You’ll find further technical details (e.g., PDF schematics, pinout diagram…) and instructions to get started on the wiki and a GitHub repository with the firmware and an MQTT over Ethernet code sample reporting battery information in all three languages.

Besides the RP2350 PoE boards from Waveshare and WIZNet, you’ll also find a wider range of ESP32-based PoE boards or products such as the M5Stack Unit PoE-P4, MakerFabs MAUWB for Home Assistant, Waveshare ESP32-P4-ETH, and others.

The RP2350-PoE-ETH board is available in four variants (no PoE, no PoE + soldered headers, PoE, PoE + soldered headers) for $16.19 to $20.33 on AliExpress, $21.11 to $27.99 on Amazon, and $13.99 to $18.99 on the Waveshare store.

The post Waveshare RP2350-POE-ETH – A Raspberry Pi RP2350 MCU board with PoE appeared first on CNX Software - Embedded Systems News.

LILYGO T-Display C5 is an ESP32-C5 board with a 1.9-inch IPS color LCD, a USB-C port for power and programming, Lithium battery support, and GPIO expansion via headers and two Qwiic expansion connectors.

The main benefit of using an ESP32-C5 chip is support for dual-band WiFi 6 connectivity (2.4/5.0 GHz), and it also keeps Bluetooth LE 5.x and the 802.15.4 radio for Zigbee/Thread/Matter found in the ESP32-C6.

LILYGO T-Display C5 specifications:

• Wireless module – Espressif Systems ESP32-WROOM-C5-1U-N16R8
  • SoC – ESP32-C5HR8
    • CPU
      • Single-core 32-bit RISC-V processor @ up to 240 MHz
      • Low-power RISC-V core @ 40 MHz acting as the main processor for power-sensitive applications
    • Memory – 384 KB SRAM on-chip, 8MB PSRAM
    • Storage – 320 KB ROM
    • Wireless Connectivity
      • Dual-band (2.4GHz/5 GHz) 802.11ax WiFi 6, with 802.11b/g/n WiFi 4 standard fallback
      • Bluetooth 5.0 Low Energy (LE)
      • 802.15.4 radio for Zigbee 3.0 and Thread 1.3
  • Storage – 16MB flash
  • Antenna – IPEX antenna connector
• Display
  • 1.9-inch RGB565 color IPS LCD with 172×320 resolution; ST7735 4-wire SPI display controller
  • Optional CST816S capacitive touch (I2C); note: apparently not for sale yet
• USB – USB Type-C port for power, programming, and data
• Expansion
  • 2x 12-pin headers
    • 14x GPIO, UART0; GPIO pins are configurable as I2C, SPI, PWM, I2S, CAN Bus, and UART
    • 6x ADC inputs
    • 5V, 3.3V OUT, GND
  • 2x Qwiic connectors: 1x I2C, 1x UART
• Misc
  • Reset, BOOT (IO28), and User (IO0) buttons
• Power Supply
  • 5V/500mA via USB Type-C port
  • 2-pin JST-1.25 connector for Lithium battery
  • AXP6202 low-power BMU (battery management unit)
• Dimensions – 62 x 26 x 10 mm

LILYGO has released firmware and code samples for PlatformIO and the Arduino IDE, including a factory sketch, and LCD, LVGL, SPI, Touch (CST816S), battery monitoring, and WiFi access point (AP) samples. You’ll find all that on GitHub along with basic instructions to get started.

The T-Display C5  is not the first ESP32-C5 dual-band WiFi IoT board with a built-in display, and alternatives include the Waveshare ESP32-C5-LCD-1.47  (1.47-inch TFT, non-touch, microSD card slot) and the Waveshare ESP32-C5-Touch-LCD-2.8 (2.8-inch IPS, capacitive touch, microSD card, battery support, 2x Qwicc). The LILYGO board offers a middle-ground option in terms of display size and features, and is slightly cheaper.

The T-Display C5 is sold for $14.99 on Amazon, $11.76 on Tindie, and the board is also listed on the LILYGO store for $11.83, but it’s out of stock there. It ships with two 12-pin headers, a power cable for the battery, and a dual-band WiFi/Bluetooth antenna. That’s for the model without touchscreen support, since the variant with a CST816S touch controller doesn’t appear to be for sale at this time.

The post LILYGO T-Display C5 board features 1.9-inch IPS color LCD, battery support, Qwiic expansion connectors appeared first on CNX Software - Embedded Systems News.

Jetway has introduced the SMC-ARK1, a new SMARC system-on-module built around the Rockchip RK3588 octa-core processor. Designed for embedded computing and edge AI applications, the board includes up to 16GB of LPDDR4x memory and up to 128GB of eMMC flash, dual GbE, and an Ampak AP6398SV Wi-Fi 5 and Bluetooth 5.0 module, alongside a dedicated HDMI input via an I-PEX FPC connector.

We’ve previously written about the Geniatech SOM-3568-SMARC module based on the Rockchip RK3568, but Jetway’s SMC-ARK1 appears to be the first SMARC module we’ve seen based on the Rockchip RK3588. The module follows the SMARC 2.1 short form-factor specification (82 × 50 mm) and can be used with compatible SMARC carrier boards.

Jetway SMC-ARK1 specifications:

• SoC – Rockchip RK3588 octa-core processor with
  • CPU – 4x Cortex‑A76  cores @ up to 2.4 GHz, 4x Cortex‑A55 cores @ 1.8 GHz
  • GPU – Arm Mali-G610 MP4 GPU
  • Video decoder – 8Kp60 H.265, VP9, AVS2, 8Kp30 H.264 AVC/MVC, 4Kp60 AV1, 1080p60 MPEG-2/-1, VC-1, VP8
  • Video encoder – 8Kp30 H.265/H.264 video encoder
  • AI accelerator – 6 TOPS NPU
• System Memory – 8 GB or 16 GB LPDDR4x at 4266MT/s
  
• Storage – 64GB or 128GB eMMC flash
• Video Input – HDMI input via I-PEX FPC connector
• Networking
  • Realtek RTL8111HS and RTL8111F Gigabit Ethernet controllers
  • Wi-Fi 5 and Bluetooth 5.0 via an Ampak AP6398SV module and two MHF antenna connectors (note: photo above shows unannounced AP6885SY module)
• Interfaces via 314-pin MXM edge connector (some info recovered from blurry block diagram below)
  • Storage
    • SATA (including SATA0 muxed with PCIe in places)
    • SDIO0/SDMMC
  • Networking – 2x Gigabit Ethernet
  • Display
    • HDMI 2.1 / eDP (MULTI0 and MULTI1)
    • USB 3.0 / DP 1.4 Alt mode
    • MIPI DSI
  • Camera – Multiple MIPI CSI camera inputs (CSI0/1/2/3 style labeling)
  • USB
    • Multiple USB 2.0 Host ports (HOST0, HOST1)
    • Type-C ports (TYPEC0, TYPEC1) with OTG support (OTG0, OTG1)
    • USB 3.0 / DP 1.4 (multiplexed)
  • PCIe – Multiple PCIe 2.0 ports (PORT0, PORT1, with mux options)
  • Low-speed I/O  – GPIO, I2C, SPI, UART
• Misc – Watchdog timer (255 levels); Wake-on-LAN (WoL) support
• Power – +4V input (Buck converter → SMARC Goldfinger → PMIC RK806 + discrete power)
• Dimensions – 82×50 mm as per SMARC 2.1 specification
• Temperature Range – Operating – 0°C to 60°C; Storage – -20°C to 70°C.

Note: Standard SMARC 2.1 specifications typically require a power supply of 3.0V to 5.25V (often standardized at 5V). A +4V rail requirement is within range, but unusual for a standard SMARC carrier board and might indicate a typo or a very strict internal power design limitation on Jetway’s end.

The SMC-ARK1 supports Android 12, Debian 11 with Linux 5.10, Ubuntu 22.04, and HarmonyOS. At the time of writing, the RK3588-based SMARC core board is available in three configurations: 8GB LPDDR4x with 64GB eMMC (SMC-ARK1-358886A1), 8GB LPDDR4x with 128GB eMMC (SMC-ARK1-358888A1), and 16GB LPDDR4x with 128GB eMMC (SMC-ARK1-358898A1).

While the Jetway SMC-ARK1 is the first short-size (82 x 55mm) SMARC 2.1 module based on Rockchip RK3588 we’ve seen, full-size ( 82 x 80mm) SMARC 2.1 modules include the Giada IM1-707, Advantech ROM-6881, and others.

Like for all other Jetway products, pricing and availability information have not been officially disclosed. Additional details, including mechanical dimensions and detailed block diagrams, can be found on the product page.

The post Jetway SMC-ARK1 – A Rockchip RK3588 SMARC 2.1 compliant system-on-module appeared first on CNX Software - Embedded Systems News.

Joshua Riek’s ubuntu-rockchip project helped many run working OS images for their Rockchip SBCs, but due to burnout and lack of support, he decided to step away in late 2024, and the project is now archived.

However, that doesn’t mean his efforts have gone to waste. GitHub user defcom5-rockchip forked the project and released two Ubuntu 24.04-based distributions working on the Orange Pi 5B, notably integrating two fixes for the Ampak AP6275P WiFi and Bluetooth module, including one that Cypress apparently already fixed in 2018, but took a while to be integrated into Rockchip OS images.

Short description of the two OS images

• Pi Desktop v1.0-beta – A curated daily-driver desktop: 4K@120 via custom VOP2/dclk kernel work, hardware video decode, zero snaps/telemetry, and “self-healing Bluetooth”: a background service that revives the AP6275P radio wedge in ~10 seconds automatically.
• Pi Studio v1.3 — the audio-production sibling: low-latency kernel (PREEMPT/1000Hz), Sonic Pi with an on-NPU AI copilot running locally on the RK3588’s NPU, JACK-transparent PipeWire.

The two repositories above only link to binary images and some documentation. The actual code can be found in a fork of ubuntu-rockchip, and the parts specific to the two WiFi and Bluetooth issues are detailed in the issue tracker.

• Fix 1 — Bluetooth dead on boot whenever WiFi is active, aka “the boot race”: the vendor bring-up script runs rfkill unblock all, waking WiFi during the BT firmware (patchram) load on the shared chip. Result: hci0 DOWN, null MAC, HCI Reset opcode 0x1003 failed: -49, Bluetooth greyed out until a lucky reboot. Fix: rfkill unblock bluetooth + verify/retry. A/B on the same board: stock 0/1, fixed 4/4, then first-boot-validated on a fresh flash with WiFi live.
• Fix 2 — Bluetooth audio stutter whenever WiFi scans, aka “runtime coexistence”: the 2018 Cypress values (btc_params8/1/50), ported to the AP6275P’s nvram. WiFi-scanning stutter: gone. The developers told us it’s the first time these parameters are implemented in a Rockchip-focused distribution.

The downside is that it only works on the Orange Pi 5, and other Rockchip RK3588 boards are not supported at this time, including variants like the Orange Pi 5 Pro. As a side note, it appears that the Raspberry Pi SBCs also share an issue similar to the one mentioned in “Fix 2”.

The post Orange Pi 5B SBC gets Ubuntu Desktop and Audio Production OS images with Bluetooth/Wi-Fi fix missed for 8 years appeared first on CNX Software - Embedded Systems News.

Redragon K688 Pro is a 96% layout tri-mode mechanical keyboard that aims to deliver enthusiast-level typing comfort and full productivity features in a compact, affordable package suitable for both work and gaming.

The keyboard features a 98-key ANSI layout that retains a full numpad, function row, and arrow keys while claiming roughly 10% less desk footprint than a traditional full-size keyboard. It uses a gasket-mounted structure with five layers of silencer foam and PCB single-key slotting. Redragon states this design cuts hollow/cavity sound by about 90% and improves noise reduction by 20%, producing the soft, bouncy “creamy” acoustics and elastic feel highlighted in product materials.

A dedicated multi-function rotary knob provides tactile control: turn for volume, press for mute, and combine with FN for backlight brightness or mode switching. This offers faster access than FN-layer shortcuts found on many competitors.

Key specifications include:

• Layout / Keys – 96% (98 keys + knob), N-key rollover / full anti-ghosting
• Switches – Pre-lubricated linear “cream” / Longyin (or Growl) switches — 40 ± 10 gf actuation force, 2.0 ± 0.6 mm actuation point, 3.6 mm total travel; hot-swappable PCB supporting 3-pin and 5-pin MX-style switches
• Keycaps – Double-shot MDA-profile (some listings note PBT shine-through)
• Mounting – Gasket with multi-layer foam dampening
• Tri-mode connectivity
  • Bluetooth 5.0 (multi-device)
  • 2.4 GHz wireless
  • USB-C wired
  • Up to five devices with easy switching; 1,000 Hz polling on wired/2.4 GHz
• Lighting – North-facing RGB (16.8 million colors), 16 preset effects plus software-driven music rhythm modes; adjustable brightness and speed
• Misc – Extras: Keycap/switch puller and spare switches typically included
• Battery – 8000 mAh battery good for about ~200 hours with backlight off / ~30 hours with backlight on; battery level check via FN+Del
• Dimensions / Weight – 390 × 144 × 40 mm, 906 grams

Redragon positions the K688 Pro as a high-value “productivity powerhouse” for hybrid gamers, content creators building aesthetic desks, office professionals who type for hours, and light enthusiasts seeking a gasket-style feel without DIY cost or complexity. It emphasizes ready-to-use convenience, versatile dual-use scenarios, and an out-of-the-box premium experience at a budget-friendly price point.

The keyboard is compatible with Windows, macOS, Android, and select consoles, notably Xbox/PS4/PS5,  via appropriate modes. The Redragon Pro program for Windows enables macros, key remapping, and advanced RGB.

The Redragon K688 Pro mechanical keyboard is available on Amazon for $48.99 with the coupon code K688RGBPRO valid until July 31 (US Time). For users wanting a compact board that keeps the numpad, offers creamy linear typing, wireless flexibility, and a handy media knob without custom-keyboard pricing, the Redragon K688 Pro is a practical contender in the current mid-range mechanical keyboard market.

The post Redragon K688 Pro – An affordable tri-mode gasket mechanical keyboard with full numpad and media knob (Sponsored) appeared first on CNX Software - Embedded Systems News.

Jake Wachlin (Energy Labs) has designed the Open RC Spotter, an open-source ESP32-based telemetry and data logger for RC cars, trucks, and other RC toys. The platform logs synchronized GPS position, acceleration, gyroscope, steering, throttle, battery voltage, temperature, and lap timing data to a microSD card, while also supporting live telemetry over Wi-Fi and ESP-NOW.

We have previously written about telemetry and data-logging hardware designed for full-sized vehicles, such as the Macchina A0 ESP32 CAN board, which supports OBD-II diagnostics and RaceChrono integration. Compared to that, this ESP32-based logger is designed for RC cars and trucks, combining GPS, IMU, PWM monitoring, and onboard data logging in a compact ESP32-based platform.

Open RC Spotter specifications:

• Wireless module – ESP32-WROOM-32E-N4
  • SoC – ESP32 dual-core Tensilica LX6 microcontroller @ up to 240 MHz with 520KB SRAM
  • Storage – 4MB flash
  • Wireless – 2.4 GHz WiFi 4 and Bluetooth 4.2 BR/LE connectivity, built-in PCB antenna
• Storage – MicroSD card slot for high-rate, synchronized CSV logging (FAT32, Class 10 recommended)
• Inputs – 3-channel PWM input (800–2200 µs) for RC receiver signal monitoring (throttle, steering, and one auxiliary)
• Sensors
  • LSM6DS 6-axis IMU (accelerometer + gyroscope)
  • ADS1115 ADC for dual 10K NTC thermistor inputs (ESC/motor temp monitoring) and battery voltage monitoring
  • IRM-56384 IR receiver for lap counting and split times
• GPS – u-blox MAX-M10S GPS with MAX2659ELT+T low-noise GPS amplifier (5 Hz, multi-constellation, UBX binary protocol)
• USB – USB Type-C port via CH340C for power, serial monitoring, and configuration
• Misc
  • Power, Reset, and User(INP) buttons
  • Dedicated board connectors for THERM1, THERM2 (Thermistors)
  • Onboard LED1 (Status) and LED2
• Power Supply
  • 5V via USB-C port
  • Battery connector
• Dimensions – 63.627 mm x 38.1 mm

The firmware is completely written in C++ and can be compiled using the Arduino IDE. The standout feature is its 10-state Extended Kalman Filter (EKF), which fuses the IMU and GPS data to calculate accurate NED (North-East-Down) position, velocity, and attitude. During a run, data is written locally to the microSD card at a default sample rate of 100 Hz.

When the run is over, double-pressing the onboard button stops the logging and spins up a Wi-Fi SoftAP, allowing users to download their CSV logs via a web browser at http://192.168.4.1. The data can then be uploaded to the Open RC Crew Chief web app for detailed plotting, visualization, and lap comparisons.

Additionally, the board supports live telemetry. By flashing a pit-side ESP32 with the provided receiver sketch, the Open RC Spotter can broadcast telemetry at 20 Hz using ESP-NOW. This data can then be fed into the Crew Chief interface via USB serial for real-time race engineer feedback.

The hardware design files, including schematic diagrams, PCB layouts, and BOM, are available on OSHWLab, while the Arduino library, firmware, and setup documentation are found in the project’s GitHub repository.

Energy Labs is currently selling the fully assembled Open RC Spotter on Tindie for $169.00.

The post Open RC Spotter is an open-source ESP32 telemetry and data logger for RC cars and toys appeared first on CNX Software - Embedded Systems News.

Creality has sent us a sample of the Pika portable 3D scanner for review. In some ways, it shares some of the features of the high-end Sermoon S1 3D scanner, like blue laser and NIR structured light scanning modes, but it’s more affordable and compact, works out of the box through USB or WiFi, and is offered in a design and form factor similar to an action or point-and-shoot camera.

In this review, I’ll start by listing the specifications, then go through an unboxing, and test the Pika with an (underpowered) Android phone, before spending more time testing the portable 3D scanner on a powerful mini PC connected to an NVIDIA GeForce RTX 5060 Ti 16GB GPU dock.

Pika 3D scanner specifications and host requirements

Pika (CRS15CPK) specifications:

• SoC – Unnamed octa-core processor clocked at 2.0 GHz
• Display – 3-inch touchscreen display
• Scanning Modes
  • 7 Parallel Blue Laser Lines
    • Working distance – 160 to 400mm
    • Outdoor Scanning – Below 110,000 lux
    • Accuracy – Up to 0.03mm; note: evaluated in laboratory conditions, and actual results may be affected by operating environments such as vibration, temperature, and other factors.
    • 3D Resolution – 0.1 – 2mm
    • Scanning Speed – Phone: up to 40fps; PC: up to 110 FPS
    • Min. scan volume – 10 x 10 x 10mm
    • Single Capture Range
      • 101 x 107mm at 160mm
      • 344 × 268mm at 400mm
    • Alignment mode – Marker / Global Marker
  • NIR structured light
    • Working distance – 200 to 1200mm
    • Outdoor Scanning – Below 80,000 lux with NIR filter
    • Accuracy – Up to 0.1mm
    • 3D Resolution – 0.15 – 2mm
    • Scanning Speed – Phone: up to 20fps; PC: up to 30 FPS
    • Min. scan volume – 150 x 150 x 150mm
    • Max. scan volume – 2000 x 2000 x 2000mm (multiple passes required)
    • Single Capture Range
      • 144 × 134mm at 200mm
      • 764 × 670mm at 1000mm
    • Alignment mode – Marker / Global / Geometry / Texture
• Color Mapping supported
• Laser Safety – Class I (eye safe)
• Color Supplemental Light – 4x Blue/White LEDs + 4x Infrared LEDs
• Calibration board – High-precision glass calibration board
• Host interface (selected in the user interface)
  • USB 3.0 Type-C port for charging and data
  • WiFi 6
• Power Supply
  • Battery compartment – 2x 3350 mAh rechargeable batteries provided
  • Charging via USB-C cable
• Dimensions – 100 x 60 x 35mm
• Weight – 260 grams with battery
• Temperature Range – -10°C to 40°C
• Humidity – 10-90% RH

It’s much more compact and cheaper than the Sermoon S1 3D scanner, and it works with phones out of the box without an additional accessory. However, the downsides are a lower number of blue laser and NIR structured light scanning modes, lower accuracy (0.03 vs 0.02mm in laser mode, and 0.1 vs 0.075 mm in NIR mode), a smaller maximum volume (2000 x 2000 x 2000 mm vs 4000 x 4000 x 4000 mm), fewer supplemental lights, and so on.

3D scanning requires a host that’s powerful enough. Here are the recommended software and hardware specifications for acceptable performance using the Pika 3D scanner:

• Android OS – Android 10.0+; SoC: Qualcomm Snapdragon 8 Gen4 or above; RAM: 12GB+
• iPhone iOS – iOS 15+; SoC: A17 or above; RAM: 8GB+
• PC – Windows 10/11 64-bit, CPU: Intel i7 Gen13 or above; GPU: NVIDIA RTX 3060 or above; RAM: 32GB
• mac – macOS 14 or above; CPU: Apple M3/M4/M5; RAM: 32GB

We own two Android phones: an OPPO A98 5G (Qualcomm Snapdragon 695 @ 2.21 GHz, 8 GB RAM) and a Redmi Note 14 Pro 5G (Mediatek Dimensity 7300 Ultra @ 2.5 GHz, 12 GB RAM). The OPPO phone is not ideal due to a weak processor and not enough RAM, and the Redmi model could be used in theory, but Creality doesn’t appear to have optimized the app for MediaTek SoC. So I’ll probably have a quick try with the OPPO A98 5G smartphone to go through the onboarding process and show why you shouldn’t use an underpowered smartphone with the Pika 3D scanner, and spend most of the review using a Khadas Mind 2 mini PC connected to the Mind Graphics 2 dock with an NVIDIA GeForce RTX 5060 Ti 16GB GPU, which I previously used with the Sermoon S1.

Creality Pika unboxing

I received the Creality Pika 3D scanner in a retail box reading “portable AI Blue Laser & NIR 3D scanner”.

The package contains the camera-looking 3D scanner (center in the photo below), a quick start guide with a QR code and link to the wiki, a turntable, an after-sales service card,  a calibration board, a lanyard, a 2-meter USB-C cable for data and charging, a cleaning cloth, a foldable tripod, six marker towers, two 3,350 mAh rechargeable Li-Ion batteries, a plastic box with two NIR filters, and two packs of reflective markers of different sizes (6mm and 3mm).

The front of the camera features two 3D scanning cameras on the left and right, each with two blue and white fill lights and two infrared lights, and the center camera features a DOE (Diffractive Optical Element) projector, a parallel laser, and an RGB camera. The top comes with a start and stop button and a power button, as well as a ventilation grid, since the scanner is cooled by a fan.

One of the sides features a lockable USB-C port for data and power, and an opening to install the lanyard.

The other side comes with a USB/WLAN mode switch and the cover for the battery compartment.

We’ll also find a mounting thread on the back along with another ventilation grid (that’s easy to block by mistake when holding the camera).

I took a photo showing the Creality Pika and Sermoon S1 3D scanner together to show the differences in size and features.

First test with an underpowered Android smartphone

The first step is to insert one of the 3350 mAh batteries into the camera 3D scanner.

I then turned on the camera, but I was quickly informed about the battery charge level( 1%), and I could barely go past language selection. So instead I connected it to my smartphone charger to charge the battery first… After that, I could go through the setup wizard. There are plenty of languages to select, not only English and simplified Chinese, as even Thai is in there!

After selecting English, a 3-step connection guide was started. First, I had to select between USB or WLAN mode. Wi-Fi connection mode was already selected, so I didn’t change anything here.

The second step is where the user selects between phone or PC host.

After selecting “Connect phone”, the scanner will start a “Hotspot” (enter AP mode), and the display will show the ESSID and a password to connect.

I did that on an OPPO A98 5G phone, and the message on the display changed to let me install the Creality Scan app (for Android).

Once I started the Creality Scan, it automatically connected to the Pika scanner and started a performance test, which, as expected, my phone failed with the message “Performance result: Not recommended”. I still tapped on Continue to carry on and have a quick check.

The display on the 3D scanner now shows “+ Start Scanning”.

It defaults to Blue laser mode. But I got the same results as when I tried using an underpowered laptop with the Sermoon S1, and I didn’t get any data in this mode. So I switched to NIR mode, and it “worked”, but only captured data at about 1 FPS, so basically unusable.

The 3D scanner display shows a range bar to help the user scan at the right distance (green color), the live capture, and buttons to pause, cancel, and stop/complete the scan. The Android app shows pretty much the same, plus the output from the RGB and depth cameras.

If I were more patient, I could probably have completed the scan with the phone in 30 minutes or so, but knowing that it will probably take 2 to 3 minutes with a proper machine, I didn’t even try. Interestingly, I was never asked to calibrate the scanner with the Android app, and it was just a plug-and-play experience that didn’t turn out well because my phone is just too slow. The takeaway here is that you have to make sure your phone is powerful enough before purchasing the Pika 3D scanner.

Creality Pika NIR and Blue laser 3D scanning using a mini PC with an NVIDIA GPU

I’ll now reproduce the setup I used in the last part of the Creality Sermoon S1 review, namely an Intel Core Ultra 7 258H-powered Khadas Mind 2 running Creality Scan 4, and connected to the Khadas Mind Graphics 2 dock with an NVIDIA GeForce RTX 5060 Ti 16GB that’s shown excellent 3D scanning performance. I’ll perform NIR scanning of a Mickey Mouse “toy” and blue line (parallel) scanning of a comb using some of the markers.

You’ll need to download Creality Scan 4 from the company’s website for Windows or macOS (still no Linux program, and it doesn’t look like there’s any plan for it). I had Creality Scan 4.2.2 installed from the Sermoon S1 review, but I had to update to 4.3.0 since it adds support for the Pika scanner.

I connected the 3D scanner over Wi-Fi. It’s basically the same process as on Android.

After the update and connecting the Pika scanner, I noticed a calibration was recommended. However, when I tried to do the calibration, I was told it only works over USB.

Oh well… after I switched the mode from WLAN to USB using the switch on the device, it rebooted into USB mode and was detected in the program. But instead of doing the calibration, I was invited to upgrade the firmware…

Once completed, I could finally start the calibration process with the provided calibration board. At first, nothing seemed to work, and I was repeatedly told to adjust the position of the 3D scanner. I eventually figured out that the USB-C port on the 3D scanner must be oriented in the direction of the bottom side of the calibration board. So I rotated the scanner by 180 degrees, and it went smoothly. The calibration process is the same as on the Sermoon S1.

At this point, I could finally start a scan. I used the same Mickey Mouse figurine as with the Android test, and selected Infrared mode, normal scan, small size, and texture tracking mode in the Creality Scan 4 program.

Scanning the black parts is tricky since black absorbs light instead of reflecting it. The raw scan had quite a few holes in the black parts like the ears despite my best efforts and some exposure adjustments.

I thought that Fusion and Mesh processing might automatically fix that, but the final result was underwhelming…

I further manually increased the exposure for the RGB and infrared cameras, and black sections were better detected, but this also picked up a lot of artifacts along the way. I eventually decided to cover the figurine with baby powder and gave it another try.

The scan was much easier and faster to complete, and the result was vastly superior. The image below is the raw scan of the figurine with baby powder, before any cleanup. The artifacts around the scan are normal and can be easily removed automatically or manually within the Creality Scan interface, as demonstrated in the Sermoon S1 review.

Here’s the final result after meshing.

However, it’s not ideal for color mapping, done with the “One-click Process” option, since the baby powder also shows up in the final version.

To have a complete figurine scan, I should have made a second scan under the figurine. I’ll skip it for that part, since I’ll perform two scans and alignment with the next test: a comb scan with parallel laser blue light.

I started using the turntable, but I eventually moved the comb on top of the white desk. Blue laser scanning requires markers, but instead of placing sticky markers on the desk as I did during the Sermoon S1 review, I used four of the marker towers provided with the Creality Pika 3D scanner.

The screenshot shows what the raw scan looks like. It looks like a headache to clean, but it’s not that difficult after orienting the scan, since the desk is not at the same level as the comb.  I didn’t clean the raw scan, and only after it was fused (Fusion step). Now that we have one side of the comb, we need to start a new scan in the same project for the other side of the comb, then fuse the scan and clean it up.

At this point, we can click Alignment in the top bar of the interface, select Feature Alignment->Manual, add three strategically placed points on the left (top side) scan, and repeat the same on the right (bottom side) scan, before clicking Preview and applying our alignment options.

After that, I could click on Meshing to complete the design.

Here’s the “final” result.

Since there are some small issues with some of the fins, I further cleaned it up in Mesh mode. However, this will create some holes (shown in yellow). I could fill the holes automatically by going to the Mesh Processing tab and selecting the “Hole Filling” option.

Update July 27. I’ve decided to 3D print the comb to demonstrate the scan is usable.

Conclusion

I have to say I quite like the Creality Pika 3D scanner. It’s small and portable, supports both WiFi 6 and USB connection to the host without hardware add-ons, and infrared and blue laser scan modes. It’s a better option for beginners than the Sermoon S1, and for the tests I did, the scans were quite similar between the two devices. Having said that, as the scan requirements and skills of the users improve, there are benefits from getting a Sermoon S1.

One downside is that the 3D scanner doesn’t work with Linux, so you’d need a Windows PC or a Mac computer, an Android smartphone, or an iPhone/iPad. It’s really important to check the hardware requirements to have a decent user experience, or the 3D scanner will be hard or impossible to use. As noted in the Sermoon S1 review, 3D scanning has a learning curve, but after a few days, you should be able to get fairly good scans.

I’d like to thank Creality for sending the Pika portable 3D scanner for review. As noted above, it’s much cheaper than the $2,000+ Sermoon S1 scanner, and is now up for pre-orders for $629 on the Creality store until August 12, 2026, after which the regular price should be $699. It’s also listed on the company’s Amazon store for $899.

The post Creality Pika review – An affordable, portable 3D scanner with infrared and blue laser modes appeared first on CNX Software - Embedded Systems News.

Beelink “ME Pro 2-Bay AI NAS Mini PC Intel Wildcat Lake 304” offers two SATA bays and one M.2 PCIe Gen4 socket, and is equipped with 16GB DDR5-5600 SO-DIMM memory and 512GB UFS 3.1 storage.

It’s an upgrade to the Beelink ME Pro Intel N95/N150 NAS and mini PC introduced last February. Besides the faster processor, the NAS trades its 5GbE port for a faster 10GbE port, and gains two Thunderbolt 4 ports. One downside, due to the lower number of PCIe lanes in Wildcat Lake processors, is that it now features one M.2 PCIe Gen4 socket, instead of three M.2 PCIe Gen3 sockets, which is somewhat mitigated by the built-in UFS storage.

Beelink ME Pro specifications:

• SoC – Intel Core 3 304 “Wildcat Lake”
  • 5-core CPU – 1x P-cores @ 1.5/4.3 GHz (Turbo) + 4x LPE-cores @ 1.4/3.3 GHz (Turbo)
  • Cache – 6MB L3 cache
  • GPU – 1-core Intel Xe3 Graphics @ 2.3 GHz (9 TOPS)
  • NPU – 15 TOPS
  • Manufacturing processor – Intel 18A
• Memory – 16GB DDR5 5600MHz SO-DIMM memory; upgradeable up to 64GB
• Storage
  • 512GB UFS 3.1 flash for OS
  • 2× SATA bays for 2.5-inch / 3.5-inch HDDs (up to 30 TB per drive, 60 TB total)
  • M.2 2280 PCIe 4.0 x1 NVMe slot (up to 4 TB)
  • Up to 64 TB total storage
• Video Output
  • HDMI 2.1 up to 4K @ 60Hz
  • DisplayPort Alt Mode via USB4 ports
  • Up to 3x independent displays
• Audio – 3.5mm headphone/mic jack
• Networking
  • 10GbE RJ45 jack
  • 2.5GbE RJ45 jack
  • Wi-Fi 6 and Bluetooth 5.4 via MediaTek MT7922 chipset
• USB
  • 2x Thunderbolt 4/USB4 (40 Gbps) Type-C ports with DisplayPort Alt mode.
  • 1x USB 3.2 Gen 2 (10Gbps) Type-A port
  • 2x USB 2.0 Type-A ports
• Misc
  • Power button
  • Reset and CLR CMOS pinhole buttons
  • Internal blower fan with high static pressure
  • Dual-sided silicone anti-vibration dampeners for HDDs
  • Tool-less slide-out motherboard for dust cleaning and maintenance
• Power Supply – 19V / 6.32A (120W) DC jack
• Dimensions – 166 × 121 × 112 mm
• Weight – ~2.1 kg
• Temperature – Operating: −10 °C to 45 °C; Storage: −20 °C to 60 °C
• Humidity – Operating: 30% to 90%; Storage: 10% to 90%
• Enclosure – Aluminum unibody chassis with thermal pads and conduction brackets

Besides the SATA bays, the hardware configuration is quite similar to the Beelink EQi Wildcat Lake mini PC we’ve just reviewed. The computer ships with Windows 11 Pro pre-installed, but it should also work with NAS operating systems and virtualization solutions such as TrueNAS, Unraid, Proxmox, and Ubuntu Server. The dual-port configuration may also be useful for software routing or link aggregation.

An additional benefit of getting an Intel Core 3 304 processor in a NAS is the 15 TOPS Intel AI Boost NPU (24 TOPS combined for the SoC), which can be used to accelerate AI-powered media search and other AI workloads. Beelink shared some LLM performance results using Gemma4 and Qwen3.5 models with 2B or 4B parameters.

Beelink has started taking pre-orders for the “ME Pro 2-Bay AI NAS Mini PC Intel Wildcat Lake 304” for $619, and shipping should start sometime in August. The “ME Pro Wildcat Lake” ships with a 120W power adapter, a 1-meter HDMI cable, a 1-meter Ethernet cable, a user manual, and two mounting screw packs. I wish Beelink would just come up with shorter names for their mini PCs…

Via AndroidPC.es

The post Beelink ME Pro 2-bay hybrid NAS gets Wildcat Lake treatment, gains 10GbE and Thunderbolt 4, loses a few M.2 slots appeared first on CNX Software - Embedded Systems News.

BOSGAME E6 ECO is another Wildcat Lake mini PC powered by the penta-core Intel Core 3 304 processor, paired with 12GB LPDDR5X and a 512GB NVMe SSD, and an attractive price tag of $379.99 and up.

It also features three video outputs via HDMI 2.0, DisplayPort 1.4, and USB-C ports, dual GbE networking, Wi-Fi 7 and Bluetooth 5.4 wireless connectivity, a 3.5mm audio jack, and five USB ports.

BOSGAME E6 ECO mini PC:

• SoC – Intel Core 3 304 “Wildcat Lake”
  • 5-core CPU – 1x P-cores @ 1.5/4.3 GHz (Turbo) + 4x LPE-cores @ 1.4/3.3 GHz (Turbo)
  • Cache – 6MB L3 cache
  • GPU – 1-core Intel Xe3 Graphics @ 2.3 GHz (9 TOPS)
  • NPU – 15 TOPS
  • Manufacturing processor – Intel 18A
• System Memory – 12GB LPDDR5x @ 6400MT/s
• Storage
  • 512GB M.2 NVMe SSD; upgradable to up to 4TB
  • Additional M.2 2242 PCIe Gen4 slot
• Video Output
  • HDMI 2.0 video output up to 4Kp60
  • DisplayPort 1.4 up to 4Kp60
  • DisplayPort via USB-C port up to 4Kp60
• Audio
  • 3.5mm audio jack
  • Digital audio output via HDMI, DP, and USB-C
• Networking
  • 2x Gigabit Ethernet RJ45 ports
  • Wi-Fi 7  and Bluetooth 5.4
• USB
  • 1x USB4 (40 Gbps) port with DisplayPort Alt mode
  • 2x USB 3.2 Gen2 (10Gbps) Type-A ports
  • 2x USB 2.0 (480Mbps) Type-A ports
• Misc
  • Power button
  • Power status LED (white when powered on)
  • Kensington lock slot
  • Dual copper heat pipes and a silent hydraulic-bearing fan (30dB)
• Power Supply – 19V/3.42A via DC jack
• Dimensions – 129 x 128 x 44mm
• Weight – xxx grams

The mini PC ships with Windows 11, but the company says that Ubuntu Linux is also supported. Earlier this month, we tested a prototype of the Beelink EQi Wildcat Lake Core 3 304 mini PC with both Windows 11 Pro and Ubuntu 26.04, and found the processor delivers excellent single-core performance, but multi-core and 3D graphics are on the low side. Thunderbolt 4/USB4 interoperability still needed some work, especially on Windows 11, and Linux support still needed some tweaks.

Both the BOSGAME E6 ECO and Beelink EQi are scheduled to ship in August, with the BOSGAME mini PC specifically scheduled for global launch on August 15, 2026. If the current $379.99 price on the company’s store is real, the E6 ECO looks like a decent deal considering the Beelink model is up for pre-order for $509, albeit with higher memory capacity (16GB LPDDR5x), 512GB UFS 3.1 storage, faster wired networking (2.5GB + 10GbE), and slightly different interfaces. It’s also the first time I’ve heard about BOSGAME, which appears to be based in China and Hong Kong.

Via TechPowerUp.

The post $379.99 BOSGAME E6 ECO mini PC features Intel Core 3 304 Wildcat Lake CPU, 12GB LPDDR5X, 512GB NVMe SSD appeared first on CNX Software - Embedded Systems News.

CirkitScape Top HAT (also written as TopHAT) is a multi-function expansion board for the Raspberry Pi that adds 16 extra GPIOs through an I2C expander, an RS-485 header, a 3-channel, 12-bit ADC, and four extra USB 2.0 ports through a USB hub.

The company told us it’s designed to “make prototyping, sensing, automation, and real-world Raspberry Pi builds cleaner and easier without stacking as many separate add-ons”. A Top HAT Sensor Kit is also offered with a few sensors for environmental monitoring builds.

CirkitScape Top HAT specifications:

• Supported SBCs – Raspberry Pi boards and compatible SBCs with 40-pin GPIO header, including the Pi 5, 4, 3B+, and Zero
• USB
  • 4x USB 2.0 (480 Mbps) ports
  • 4-port USB 2.0 hub
  • Per-port current-limited switches (500mA per port)
• I/O Expansion
  • 16x GPIOs via MCP23017 I²C expander
  • Analog Input – 3-channel 12-bit ADC via ADS1015 chip
  • RS-485 interface via MAX485 half-duplex transceiver
• Communication Interfaces – I²C, SPI, UART via 40-pin Raspberry Pi GPIO header
• Power Supply
  • Operating Voltage: 5V DC input
  • Logic Voltage: 3.3V
• Dimensions –  76.2 x 50.8 mm (Raspberry Pi HAT form factor)

The Top HAT is compatible with Raspberry Pi OS and doesn’t require any custom firmware. It supports pyserial, smbus, and gpiozero Python libraries and is fully controllable via I2C, UART, and GPIO pins through the 40-pin GPIO header on the Pi. The company also provides a script to configure/enable I2C and UART, install required packages, and set up a Python virtual environment.  Finally, some Python code samples and a Python-based web dashboard to control the hardware without programming by simply accessing http://<raspberrypi-ip>:5000.

Target applications include industrial control systems, IoT gateways and edge devices, sensor data acquisition, USB device multiplexing, serial network bridging via RS-485, and generic embedded system prototyping.

CirkitScape sells the Top HAT for $48.6 on Amazon or $45 on the company’s website, where you’ll also find all resources to get started (datasheet, Python samples, dashboard program, installation script, etc.). The Top HAT Sensor Kit mentioned in the introduction goes for $73 with the Top HAT, a USB-C to USB-A cable, a temperature and humidity sensor,  a capacitive soil moisture sensor, a water/rain detection sensor, and a few jumper wires. I’m just not sure the sensor kit is available, as the company used ChatGPT to create the product image below… There’s no address on the website, but CirkitScape appears to be a small business from Alabama.

The post CirkitScape Top HAT adds extra GPIOs, RS-485, 3-channel ADC, four USB 2.0 ports to Raspberry Pi SBCs appeared first on CNX Software - Embedded Systems News.

AMD has announced a range of new AI solutions at Advanced AI 2026, including 6th Gen AMD EPYC CPUs, AMD Instinct MI400 Series GPUs, and AMD Helios AI rackscale solutions designed for data centers. In this article, we’ll focus instead on embedded and physical AI solutions like the new Ryzen AI Embedded X100 processor family and a new Kria AI SoM integrated into the Kria AI Robotics Developer Platform pairing an X100 embedded CPU with a Kria FPGA.

AMD Ryzen AI Embedded X100

AMD first introduced the Ryzen AI Embedded X100 family when it launched the Ryzen AI Embedded P100 family in January 2026, delivering up to 50 TOPS of AI performance. However, at the time there were very few details about the X100 family, just that it would offer higher CPU core counts and AI TOPS performance, and this latest announcement fixes that.

AMD claims the new X100 series processors are expected to deliver up to 2.1x higher multithread CPU performance (CoreMark), an 1.7x graphics performance uplift (OpenGL, GFXBench 5.0.0), and 3.5x higher token generation throughput with 1.4x faster time-to-first-token (TTFT on llama-bench) for accelerating physical AI workloads compared to equivalent Intel Core Ultra Series 3 processors.

Compared to the NVIDIA Jetson Thor T5000 platform, the Ryzen AI Embedded X100 Series processors deliver up to 3x higher peak FP32 performance and an average of 1.7x higher performance compared to discrete GPUs like NVIDIA RTX 4000 Ada on workloads such as beamforming for advanced medical ultrasound.

Six models are currently available in the Ryzen AI Embedded X100 family: AMD Ryzen AI X199, X188, and X168, and the industrial-grade equivalent: X199i, X188i, and X168i. All feature Zen 5 cores, an up to 50 TOPS NPU, and a TDP of 55 Watts.

Other shared specifications:

• Memory – LPDDR5x-8533 up to 8 channels with Link-ECC
• HW Video Encode / Decode – up to 4K @60Hz
• Display Interfaces – Up to 5 ports using HDMI 2.1, DP 2.0, eDP v1.4, DVI, and USB4 up to 8K (7680×4320) @ 60 FPS
• USB – 2x USB4, 2x USB 3.2, 3x USB 2.0
• PCIe – 16-lane PCIe Gen4
• Low Speed Interfaces – I2C, SMBus, SPI, UART
• Package – 45 x 37.5 mm
• Reliability – Standard: 2.5 years; extended: 10 years

The new family is supported by an open software stack, which includes Linux, the AMD ROCm software stack for iGPU workload acceleration, Xen Hypervisor for virtualization, and support for standard AI frameworks like PyTorch, ONNX, and TensorFlow. The company is also offering tools to migrate existing codebases from CUDA to ROCm.

Sampling of Ryzen AI Embedded X100 Series processors began in June 2026, but mass production is scheduled for Q4 2026, and they will be available for purchase at least until 2037. System-on-modules from Arbor, Congatec, iBase, IEI, Sapphire, and Seavo are also planned. Additional information can be found on the product page and in the announcement.

Kria AI SOM and AI Robotics Developer Platform

AMD also launched a new Kria AI SOM system-on-module based on the X100 CPU along with the Kria AI Robotics Developer Platform.

The AMD Kria AI SOM is a COM-HPC module based on the AMD Ryzen AI Embedded X100, but contrary to earlier Kria modules, there’s no FPGA on the module itself

AMD Kria AI SoM key features:

• SoC – 16-Core ”Zen 5” CPU with AMD RDNA 3.5 iGPU, AMD XDNA 2 NPU (not sure why AMD didn’t mention the SKU here, but “advanced detective work” implies it’s the AMD Ryzen AI Embedded X199…
• System Memory – 64GB or 128 GB LPDDR5x unified memory
• Camera support – Up to 8x camera inputs via FAKRA connectors (on the carrier)
• Interfaces – CAN-FD, RS485, 10GbE QSFP

AMD says the Kria AI SOM can scale up to 234 concurrent agents and make up to 8,000 decisions per second based on tests using mimik Agentix computer benchmarking and a Bosch Rexroth controller, respectively.

The Kria AI SOM is integrated into a robotics developer platform, which has an FPGA on the carrier board, with the following specifications:

• System-on-Module – AMD Kria AI SoM described above with X199 16-core CPU, 64GB or 128GB LPDDR5X memory
• FPGA – Unspecified Spartan UltraScale+ FPGA
• Storage – M.2 NVMe SSD socket
• Display – 2x Mini DisplayPort
• Camera
  • 1x FAKRA x4 camera (GMSL2/3)
  • 1x FAKRA x4 camera (GMSL1/2)
• Audio
  • 3.5mm audio jack
  • A2B (Automotive Audio Bus)
• Networking
  • 2x 5Gbps Ethernet RJ45 ports
  • 2x Gigabit Ethernet RJ45 ports (EtherCAT/TSN)
  • 10GbE (x4) QFSP cage
  • Optional WiFi and Bluetooth
• USB
  • 2x USB4 Type-C ports
  • 3x USB 3.2 Type-C ports
  • 2x USB 2.0 Type-A ports
• Serial – 2x CAN-FD,RS485
• I/O Expansion
  • 1x Oculink PCIe connector
  • M.2 (NVMe SSD) 2280 Key-M socket
  • M.2 (WiFi/Bluetooth) 2230 Key-E socket
  • Pmod connector
• Misc – Inertial Measurement Unit (IMU)

The AMD Kria AI SOMs will be available from ODM partners starting in Q4 2026, and the Kria AI Robotics Developer Platform is sampling with early access customers, while generally available is also planned for Q4 2026. The developer platform may have different names depending on the vendor, and I just received an email from Sapphire about their EDGE+ Apex SOM/Carrier Robotics Platform based on the same design (no product page yet).  More details can be found on the product page and in the announcement specific to the Kria AI SoM and devkit. While no pricing has been announced, it will compete against the Jetson AGX Thor Developer Kit, so expect $5,000+ for the new AMD Developer Platform.

The post AMD launches Ryzen AI Embedded X100 processors, Kria AI SoM, and physical AI/robotics developer platform appeared first on CNX Software - Embedded Systems News.

Check your calendar, everyone! It’s definitely not April 1st. Now that we have that out of the way, we can start with the main topic of this article. Ready? Augmental MouthPad is a Bluetooth HID device you put into your mouth and control like a touchpad or trackpad, but with your tongue (and head) instead of your fingers.

MouthPad is a custom-fit mouthpiece that turns tongue and head movements into HID data to control a phone, a tablet, and a computer like you would normally do with using a mouse, touchpad, or a touchscreen. When I first read about it and saw a short video on X, it felt like a joke, but it can be a lifesaver for people with disabilities.

MouthPad specifications:

• Compatibility – macOS, Windows, Linux, iOS, Android
• Connectivity – Bluetooth
• Battery
  • 7+ hours of continuous use
  • 1.5 hours charge time
• Dimensions
  • MouthPad – About ~30 × 50 × 80 mm; membrane thickness about 1mm
  • Charging case – 38 × 75 × 95 mm
• Weight
  • MouthPad – 10 grams
  • Charging case – 112 grams
• Material- Dental resin
• Battery certifications – UN 38.3, IEC 62133

Besides the MouthPad and charging case, the package also includes a USB-C to USB-A cable for charging. The documentation website explains how to left-click, right-click, scroll, click-and-drag, and pause/resume the cursor with your tongue and head movements. The company claims over 100 people already use it, and some for up to 16 hours/day. Just don’t drink hot drinks or eat while wearing it. Cold water is fine.

Since the MouthPad needs to be customized, it requires an intraoral 3D scan from your dentist or a local scanning provider, which should cost $50 to $150 in the US. The MouthPad is only available in the US, as the company follows US regulations. They also only ship to US addresses, but don’t geo-fence any of the device’s features or capabilities, so the MouthPad will work outside the US as well. I assume that means you could probably order a MouthPad through a US forwarding agent if you live outside of the US, although it might be harder to have the 1-year warranty work that way. You might also want to wait since Augmental plans to eventually go global.

As one would expect, such a device is quite pricey, and goes for $1,400 excluding the cost of the dental scan. But it can be worth it if it’s the only way to use a computer or phone. Buying two devices is $1,900 and something that may be useful if it needs to be used more than 7 hours a day. The company is also working on the VOX wearable microphone to simulate a keyboard.

The post Augmental MouthPad is a tongue-controlled touchpad/trackpad appeared first on CNX Software - Embedded Systems News.

If you feel like a cheapskate when purchasing a Raspberry Pi 5 16GB for only $305, there’s now a solution to avoid embarrassment. Meet the $3,640 AIS-1000 AI Surveillance Station, pairing a Raspberry Pi 5 with a Hailo-8L AI accelerator and a 1TB NVMe SSD, all housed in a metal enclosure.

Jokes aside, the device from Planet Technology is real, and users mostly pay for NDAA (National Defense Authorization Act) and TAA (Trade Agreements Act) compliance for use in AI-enhanced surveillance applications for government, critical infrastructure, or other high-security applications.

Planet AIS-1000 hardware specifications:

• SBC – Raspberry Pi 5 (only advertised features listed)
  • SoC – Broadcom BCM2712
    • CPU – Quad-core Arm Cortex-A76 processor @ 2.4 GHz
    • GPU – VideoCore VII GPU @ 800 MHz with support for OpenGL ES 3.1, Vulkan 1.2, 4Kp60 HEVC decoder
  • System Memory – 8GB LPDDR4X-4267 SDRAM
  • Networking – Gigabit Ethernet RJ45 port via Broadcom BCM54213PE transceiver
  • USB – 2x USB 3.0 ports, 2x USB 2.0 ports
• Storage – 1TB NVMe SSD
• AI accelerator – Hailo-8L (13TOPS)
• Video support
  • H.264, H.265, or H.265+ codec
  • Live RTSP streaming up to 8 channels at 704 × 576 resolution at 10 FPS (Sub-stream)
  • Video Recording –  Up to 1080p at 25/30 FPS per channel
• Misc
  • Recessed power button for under 10-second shutdown
  • Power LED
• Power Supply – 5V/5.1A via USB-C port on Pi 5; up to 27W power consumption
• Dimensions – 88.7 × 68.2 × 41 mm (metal enclosure)
• Weight – 287 grams
• Temperature Range – 0 to 50°C

The solution works as an NVR handling up to eight RTSP IP cameras. AI video analytics functions include parking management (illegal vehicle intrusion / idle timeout detection), human detection (line crossing, area intrusion, crowd counting), pet detection for cats and dogs, face recognition with whitelist and blacklist support, industrial safety detection to check for helmet and vest wearing, and fire & smoke detection. The software also implements an intelligent video search system using an LLM.

Management is performed through a web interface running on the Raspberry Pi 5 to handle system configuration, real-time display, AI search, AI model upload/update, user access, historical data, and so on. While the company talks about the “centralized management platform”, that’s just for the eight IP cameras, and as I understand it, there’s no additional cloud-based web dashboard to manage a fleet of AIS-1000 devices. In other words, each has its own web interface.

You’ll find more details, including a datasheet and a user manual for the web interfaces, on the product page, where you can purchase the AIS-1000 AI surveillance station for $3,640 with free shipping within the US. I estimate an equivalent hardware-only solution would sell for $400-$500, so the main costs are NDAA and TAA compliance and custom software development. The niche market for this type of solution also requires a higher-than-usual profit margin per device. Additional information may also be found in the announcement.

Thanks to TLS for the tip.

The post AIS-1000 AI Surveillance Station – A $3,640 NDAA and TAA-compliant AI box based on Raspberry Pi 5 and Hailo-8L appeared first on CNX Software - Embedded Systems News.

The nRF54L15 Discovery is a compact wireless devboard with an onboard nRF52820 CMSIS-DAP debugger/programmer for Bluetooth LE, Thread, Zigbee, Matter, NFC, and other 2.4 GHz IoT applications. Its snap-apart design allows the target and interface MCUs to be used as two standalone boards.

In the past, we have written about various development boards based on the Nordic nRF54L series, including the Seeed Studio XIAO nRF54L15, XIAO nRF54LM20A (Sense), and the nRF54L15 Connect Kit. The nRF54L15 Discovery board further integrates a snappable nRF52820 CMSIS-DAP debugger/programmer section. It also includes an nPM1300 PMIC with 32–800 mA LiPo battery charging and USB-C PD support, along with a TPS63900 buck-boost converter configurable for 1.8V or 3.3V I/O operation.

nRF54L15 Discovery specifications:

• Wireless MCU – Nordic Semi nRF54L15 wireless SoC
  • CPU
    • Arm Cortex-M33 @ up to 128 MHz
    • RISC-V coprocessor for “software-defined peripheral”
  • Memory – 256 KB SRAM
  • Storage – 1.5MB Non-volatile Memory (NVM)
  • Wireless
    • Bluetooth LE 6.0 (includes Channel Sounding)
    • Thread, Zigbee (IEEE 802.15.4-2020)
    • Matter
    • Amazon Sidewalk
    • Proprietary 2.4 GHz protocols with a new 4 Mbps mode
    • NFC support
• Debugger SoC – Nordic Semi nRF52820
  • CPU – Arm Cortex-M4 @ 64 MHz
  • Memory – 32 KB RAM
  • Storage – 256 KB Flash
  • Wireless – Bluetooth 5.2 LE and 802.15.4 radios (Inactive in default CMSIS-DAP debugger mode, but fully functional if snapped off and used as a standalone devboard)
• USB – 2x USB Type-C ports (1x Mainboard; 1x Debugger)
• Expansion via dual-row 20+10 headers (60 pins in total) with castellated edges
  • Up to 48x GPIOs (29x for nRF54L15, 15x for nRF52820, 2x for nPM1300)
  • 5x SPI/UART/I2C, 3x PWM, 2x QDEC, I2S, PDM, 14-bit ADC
• Misc
  • User LEDs and buttons
  • U.FL connector for 2.4 GHz antenna
  • U.FL connector for NFC antenna
• Power supply
  • 5V in via USB Type-C port
  • Nordic nPM1300 USB-C compliant PMIC with 32–800 mA LiPo charging, power path, 200 mA buck-boost converter, and configurable 1.0V–3.3V output
  • The nPM1300 PMIC features a 50 mA switchable LDO for LEDs and supports USB-C PD with 100 mA, 500 mA, and 1500 mA current limits
  • TPS63900 ultra-low quiescent current buck-boost converter with selectable 1.8V/3.3V I/O supply
• Dimensions – 91 x 21 mm

The nRF54L15 Discovery board supports Zephyr RTOS, the nRF Connect SDK, and (Micro)Python. It ships with the nRF52820 pre-flashed with CMSIS-DAP firmware, enabling onboard debugging and programming without external hardware, and can be programmed out of the box using the west flash -r pyocd command. More information is available on the relevant GitHub repository.

The Icy Electronics nRF54L15 Discovery devboard with CMSIS-DAP debugger is available for $27.19 on Tindie or Lectronz, with minor volume discounts starting at five units. Shipping out of France starts at $3.20 for untracked global delivery.

The post Icy Electronics nRF54L15 Discovery devboard integrates a snappable nRF52820 CMSIS-DAP debugger appeared first on CNX Software - Embedded Systems News.

Xteink X4 Pro is a pocket-sized eReader with a 4.3-inch touchscreen E Ink display, a 16GB microSD card, 2.4 GHz WiFi and Bluetooth wireless connectivity, and a 1100 mAh battery.

Previous models like the X3 and X4 e-readers were based on the ESP32-C3 RISC-V microcontroller, but the company has not disclosed which MCU was used in the new X4 Pro model. However, a recent comment on GitHub points to the X4 Pro featuring a more powerful ESP32-S3 microcontroller.

Xteink X4 Pro specifications:

• SoC – ESP32-S3 microcontroller (TBC) with 2.4 GHz WiFi 4 and Bluetooth wireless connectivity
• Storage – 16 GB microSD card; expandable up to 256 GB
• Display
  • 4.3-inch touchscreen E Ink display; 219 PPI
  • Dual-tone front light
• Supported Formats (with default firmware)
  • Document: EPUB, TXT
  • Image: JPG, BMP
  • Fonts: BIN, XTF
• Misc
  • Home touch button
  • Power, Reset, Previous, and Next buttons
  • Built-in magnets for connection to the back of a smartphone
• Power Supply
  • Charging via pogo pins (no USB-C port)
  • 1,100 mAh battery
• Dimensions – 111 × 69 × 5.95 mm
• Weight – 72 grams

It ships with firmware supporting English, Japanese, and Simplified/Traditional Chinese. However, other languages including Korean, Turkish, Russian, Vietnamese, Malay, Indonesian, Thai, and a range of European languages will be rolled out via future firmware updates.

However, one good thing with the Xteink X4 Pro is that its firmware is unlocked, so you can also install a third-party firmware as explained on the website:

All units purchased directly through the XTEINK official website are exempt from third-party firmware restrictions. This commitment applies to all past and present models, with no planned changes for future updates.

One such firmware will be CrossPoint Reader, which is already working on the previous generation Xteink X3 and X4 models, as well as Seeed Sticky, M5Stack M5Paper, and Lilygo T5. The X4 Pro doesn’t appear supported yet, but The Verge reports that Xteink is now an official partner of the project, and the Xteink X4 Pro will be officially supported once it starts shipping to customers.

The X4 Pro E-Reader ships with a 16GB microSD card, a matte screen protector, two e-Reader adhesive metal rings, a magnetic charging adapter, and a quick start guide. It can be pre-ordered now for $99. Deliveries are scheduled to start by August 2, 2026. The older ESP32-C3-based X4 model without touchscreen and backlight, and a smaller battery, goes for $69 on Amazon.

Via Liliputing

The post $99 Xteink X4 Pro 4.3-inch touchscreen eReader to support CrossPoint Reader open-source firmware appeared first on CNX Software - Embedded Systems News.

Arduino UNO Q Arcade bundle is a kit designed to build a custom retro gaming console. The bundle combines the Arduino UNO Q 4GB SBC with a set of daisy-chainable Modulino I2C modules, allowing users to create a custom game controller without requiring complex wiring or breadboards.

The kit relies on Qwiic connectors to daisy-chain the input modules, whose data is then processed by the UNO Q to emulate standard HIDs for gaming or development. The plastic parts can be 3D printed by the user.

Arduino UNO Q Arcade bundle contents:

• Arduino UNO Q 4GB – A Qualcomm Dragonwing QRB2210 SBC with an onboard eMMC, running Debian Linux
• Modulino Joystick – An I2C node configurable for simulating D-pad inputs or mouse cursor movements
• Modulino Buttons – An I2C node for mapping single keystrokes, multi-key shortcuts, or mouse clicks
• Modulino Movement – A 6-axis IMU node for motion-based mouse control

Note: The UNO Q requires an external USB-C hub (for HDMI out, USB ports, and power passthrough) and a power supply, as neither is included in the bundle.

On the software side, the Arduino UNO Q runs a standard Debian Linux desktop environment. To use the I2C-based Modulino nodes as a game controller, Arduino also provides a Python script (injector.py) that acts as a bridge translating the raw sensor events into standard virtual keyboard and mouse data via the Linux uinput system.

Because the OS recognizes the Modulino setup as a standard input device, you can easily run native Linux applications. Arduino’s official tutorial shows how to set up RetroArch to emulate classic consoles (using cores like mGBA, bsnes, and Nestopia) and utilize the open-source Godot engine for custom game development. The documentation also includes instructions for configuring systemd services, allowing both the controller bridge and the emulator to launch automatically whenever the board boots up.

Arduino also provides open-source 3D-printable files for a base, button pad, and cover, which you need to print and assemble yourself. You will also need fourteen M3x6 flathead screws, two M3x10 screws, and Qwiic cables for assembly.

The Arduino UNO Q Arcade bundle is available on the Arduino store for $90.90 or €96.36 (including VAT), which at the time of writing is a 20% discount over the regular price until September 10, 2026. A few more details may also be found on the Arduino blog.

The post Arduino UNO Q Arcade bundle lets you build a custom retro gaming console for about $90 appeared first on CNX Software - Embedded Systems News.

Previously offered with AMD Ryzen 5/7/9 CPUs, the ACEMAGICIAN Kron K1 mini PC has now gotten a low-cost version based on the AMD Ryzen R2544 quad-core processor paired with 8GB DDR4 and a 256GB SSD and sold for just $279.99 on Amazon.

With everything going up in price, companies are trying to offer more affordable systems. ACEMAGICIAN has decided to use a processor from the Ryzen Embedded R2000 family introduced in 2022 that’s typically found in industrial-grade modules and motherboards such as the AAEON COM-R2KC6 COM Express module or the Kontron D3723-R mini-ITX motherboard.

ACEMAGICIAN Kron K1 R2544 specifications:

• SoC – AMD Ryzen Embedded R2544
  • CPU – Quad-core/eight-thread processor @ 3.35 / 3.7 GHz (Turbo)
  • GPU – 8-core AMD Radeon Vega Graphics
  • TDP – 45 Watts
• System Memory
  • 8GB DDR4 RAM pre-installed
  • Upgradeable up to 64GB via two DDR4 SO-DIMM sockets
• Storage
  • 256GB M.2 SSD pre-installed
  • 2x M.2 2280 slots: one PCIe Gen3/SATA, the other SATA-only.
• Video Output
  • HDMI 2.0 port
  • DisplayPort 1.4 connector
  • DisplayPort via  USB Type-C port
  • Up to 3x independent displays; single screen up to 4Kp60; triple screen setups up to 4Kp30 per screen
• Audio
  • 3.5mm audio jack
  • Digital audio output via HDMI, USB-C, or DP
• Networking
  • Gigabit Ethernet RJ45 port
  • WiFi 5 and Bluetooth 4.2
• USB
  • 2x USB 3.2 Gen2 Type-A ports
  • 4x USB 3.2 Gen1 Type-A ports
  • 1x USB Type-C port
• Misc
  • Power button
  • Aluminum heatsink, 80mm “quiet fan”, and air outlets for cooling
• Power Supply – Via DC jack
• Dimensions – 128.3 × 128.3 × 40.9 mm
• Weight – 500 grams

The mini PC comes with an activated version of Windows 11 Pro.  It ships with a power adapter, an HDMI cable, a VESA mount with screws, and a user manual.

The company claims the AMD Ryzen Embedded R2544 processor is faster than AMD Ryzen 3 4300U and Ryzen 5 3500U processors, at least in Cinebench R23 multi-core benchmark… All these are fairly old processors, so don’t expect amazing performance. It will still deliver higher performance than an Intel N150, but should consume more power.

The $279.99 price tag on Amazon is said to be a “limited time deal”.  It’s also listed on AliExpress for $235.55 in the same 8GB/256GB configuration,  and there’s a 16GB/512GB option for $267.67 on the same page.

Via AndroidPC.es

The post ACEMAGICIAN Kron K1 mini PC gets low-cost variant with AMD Ryzen R2544 CPU, 8GB DDR4, 256GB SSD appeared first on CNX Software - Embedded Systems News.

Raspberry Pi has just announced the “Raspberry Pi Touch Display 2 – 10″ Portrait”, and we’ve received an early review sample for testing. The new 10.1-inch Raspberry Pi Touch Display 2 is a multi-touch portrait panel for Raspberry Pi designed for interactive projects such as tablets, entertainment systems, and information dashboards.

The new 10.1-inch model follows the original 7-inch Raspberry Pi Touch Display 2 launched in 2024 and the 5-inch Raspberry Pi Touch Displays 2 introduced in 2025. I reviewed the 7-inch model with the Raspberry Pi AI camera and a Pi 4, and I just gave away the 5-inch variant. In this short review, I’ll show how to mount a Raspberry Pi 5 to the Raspberry Pi Touch Display 2 – 10″ Portrait, install it on a VESA mount, and test features such as 10-pin capacitive touch and display orientation. But let’s start with the specifications first.

Raspberry Pi Touch Display 2 – 10″ Portrait specifications

• Display diagonal – 10.1-inch
• Active area – 135.4 mm × 216.6 mm
• Viewing angle – 85 degrees
• Display format – 24-bit RGB
• Resolution – 1200 × 1920 pixels
• Display type – True multi-touch capacitive touch panel supporting ten-finger touch
• In-plane switching thin-film transistor (IPS TFT) liquid-crystal display; Normally black; Transmissive
• Touch response time – Up to 35 ms
• Surface treatment – Anti-glare
• Color configuration – RGB stripe
• Backlight brightness – 400 cd/m2
• Dimensions – 247.3 x 161.8 mm
• Operating temperature −20°C to +70°C
• Production lifetime – To remain in production until at least January 2030

One important point is that the 10-inch Raspberry Pi Touch Display 2 is only compatible with Raspberry Pi 5 and Raspberry Pi Compute Modules connected to a carrier board with an MIPI connector.

Unboxing

I received the display in a retail package reading “Raspberry Pi Touch Display 2” and  “10” Portrait” with basic specifications and installation instructions on the bottom side.

The display ships with a power cable, an FPC cable, and eight M2.5 screws.

Raspberry Pi 5 installation on the 10.1-inch Raspberry Pi Touch Display 2

I was also given a Raspberry Pi 5 4GB for the review. The installation is pretty straightforward, since only two connections are required to connect the display to the Raspberry Pi: a power cable for the GPIO header and a ribbon cable for the DSI port.

The only potential mistake is inserting the cables in the wrong way. Once done, we can secure the Raspberry Pi 5 with four of the M2.5 screws, connect the FPC cable to one of the MIPI connectors, and the power cable to the 40-pin GPIO header.

Insert a microSD card with Raspberry Pi OS, connect a 5V/5.1A USB power adapter, and Raspberry Pi OS will boot in portrait mode with everything working out of the box, including ten-finger touch and the on-screen keyboard. More on that below.

3D printed VESA mount for the Raspberry Pi Touch 2 – 10″ Portrait

I usually use a wooden stand to test displays, but the 10.1-inch is a bit too big. I’m sure there will eventually be an enclosure for this model, like we’ve seen with the $10 Waveshare PI5-CASE-TD2 for the 7-inch model.

Since at the time of the review, no enclosure was provided/announced, I designed a VESA mount adapter for the 10.1-inch Raspberry Pi Touch 2 in Blender, and printed it with PLA filament (rainbow style, hence the colorful design).

I didn’t think everything through during the design, so I had to use a power drill to enlarge four of the holes to mount the display to the adapter. But I eventually got something working, and the old VESA stand I use can also rotate, which will be great to test landscape/portrait orientation.

Landscape mode after mount.

The STL file can be found on GitHub, but I don’t recommend it unless you have a few 35mm M2.5 screws lying around…

Multi-touch, portrait/landscape mode, and on-screen keyboard

I also tested the advertised features, including 10-point capacitive touch using the WBBMTT website.

All ten points were detected, and tracking works smoothly when moving the fingers on the screen.

Raspberry Pi OS supports four orientation modes: Normal (default, in portrait mode), Left, Inverted, and Right. I tested Normal and Left (Landscape), and it worked well after I updated Raspberry Pi OS. Before the update, I had some issues when switching the orientation, as the display rotated but not the touchscreen, so I could not confirm the change.

Another important option is the touchscreen mode, for which there are two options:

• Mouse Emulation – Simulate a mouse: one tap converts to left click, long tap to right-click, and double tap to double-click. It’s better for app not made for touch screens.
• Multitouch – Multitouch support, similar to using a tablet

In previous touchscreen reviews, I had to change that manually by editing config files, but it’s now available right in the Control Centre.

While browsing the web in multitouch mode, you can scroll a web page, pinch and zoom, etc… like you would on a tablet. In Mouse Emulation mode, trying to scroll a web page will select the text instead.

If you select a text field, the soft keyboard will show up automatically. No need to install a keyboard manually like in the past.

You can watch a short demo showcasing the features covered above.

I’d like to thank Raspberry Pi for sending the “Touch 2 – 10″ Portrait” for review. The new display can be purchased from your favorite reseller for $80 plus potential shipping and taxes.

The post Raspberry Pi Touch Display 2 – 10″ Portrait review, 3D printed VESA mount appeared first on CNX Software - Embedded Systems News.