Glen Atkins wrote a blog post about a recent project for the holidays:
The lighted tree in the video above gets both the power and data for its RGB LED pixels using a single Ethernet cable. Power for the pixels is supplied from an Ethernet switch using the 802.3at PoE+ standard. Data for the pixels comes from software running on a PC that generates Art-Net packets at 40 Hz. Each Art-Net packet contains the RGB levels for all the pixels on the tree. Let’s take a closer look at the technical details and how this tree came into existence.
A few weeks ago, I wrote a post where I reversed engineered some Philips Color Kinetics iColor Flex RGB LED string lights. These lights require an AC power data supply that supplies 24 volts to power the pixels and transforms DMX or UDP packets of pixel data into the protocol used by the pixels. In addition to the power supply, the pixels require a proprietary leader cable to connect them to the power supply.
In a typical setup, you have to run AC mains power and Ethernet data to the power supply then run the leader cable to the pixels. To avoid having a large stack of power data supplies in larger setups, Color Kinetics makes a rack mount power supply that can power up to eight strings of lights. This rack mount power supply still requires a leader cable for each string of lights.
I’ve been wanting to build an 802.3af/at/bt Power over Ethernet design for a few years now and have always come up short on ideas and then it hit me, what if the Ethernet cable could connect closer to the pixels in the photo above? With a small box of electronics between the Ethernet cable and the connector on the end of the pixels, the pixels could receive both power and data from the Ethernet switch. No more AC mains wiring and no more proprietary leader cables.
We were amazed to see the incredible wearable project by Zach from NeuroTinker while at CrowdSupply Teardown:
Cordwood-style blinky ring, powered by an ATtiny85 and a pair of size 10 Zn-air hearing aide batteries.
From Jeremy S Cook on Tindie blog:
Ted Yapo had a small problem. As an amateur atronomer and astrophotographer, he needed a way to mark his expensive equipment so that he wouldn’t trip over it in the dark. Glow-in-the-dark materials were out because of they only glow for a short time, and glow sticks were also less than ideal because of their single-use nature. Tritium light sources would be perfect, barring the small details that they’re radioactive, expensive, and in the US only a few uses are allowed, most are prohibited by law.
So Yapo instead came up with an LED light that can run for not 20 hours, or even 20 days, but 20 yearson a single CR2032 coin cell battery!
Alex on Hackaday.io is working on a smart miniature (DIP6) 5×7 LED Matrix:
This project is based on my #reDOT project. Basicly it is a 5×7 SMD LED Matrix an a microcontroller on one PCB. I started wirh 0201 LEDs (see first project log), but this was not reliable. So a second version with 0402 is in development. 0402 LEDs do have some benefits over 0201:
- bigger and you can solder them better
- more colors availible
The microcontroller (a low coast STM8) drives all LEDs directly with multiplexing. For controlling a UART interface is available. The dimensions are like a DIP-6 package. For easy connection of multiple PCBs, the pads are castellated. Also the supply rails are available on both sides. So multiple of these display can be soldered together to a bigger display without the need of additional wiring.
Jens Hauke designed this charlieplexed 20 LED blinker controlled by an ATTiny45 for the Hackaday Coin Cell Challenge:
This is a small blinky with 20 LEDs powered by one CR2032 coin cell
and with an ATTiny45 brain. The firmware is written in plain C and
compiled with the avr-gcc toolchain. The PCB is a two layer design made
Space efficient daisychained LED placing with shared anode/cathode soldering pads.
Firmware and gerbers are available on GitHub:
Jens has shared the board on OSH Park:
Here is a video of the LED in action:
A holiday project by Dan Hienzsch (@rheingoldheavy) to build a little Snowbot with an adjustable speed larson scanner for an eye:
The Snowbot has three major subsystems: Power and Timer and Display.
The power subsystem uses a 3.7V LiPO battery boosted to 5V with an SC4503 boost converter to power the fully analog circuit. It requires a set of passive components in order to generate the higher voltage.
The timer subsystem is comprised of a 555 IC that generates a clock signal. The speed of the clock is adjusted by twisting the potentiometer (the nose of the snowbot). The clock signal ticks through the outputs of a CD4017 decade counter, lighting each LED in sequence, then moving back through them again.
The display subsystem involves taking the output of the CD4017, and directing it to six red LEDs in the form of a larson scanner. In addition to lighting the LED, the current also charges a 22uF capacitor through a diode. When the output moves to the next LED, the cap discharges through a 2.2K resistor (part of a resistor network), fading the LED out gracefully.
From Ken Olsen of The Maker’s Box:
You never know what people will do with your ideas, and it is always fun to see someone do something I would have never thought off with them. I got a video from someone who built my Programmable Fidget Spinner, and used a leaf blower to see how fast they could get it to go. Fortunately, they were wearing safety glasses, and no, it didn’t come flying apart. It did, however, start displaying erratically at about 3600 RPM (the fastest I can get it by hand is just under 2000 RPM).
So, what is going on? TLDR: I figured it out and fixed it, and here is proof.