Few things get a Hackaday staffer excited like bunches of tiny LEDs. The smaller and denser the better, any form will do as long as we can get a macro shot or a video of a buttery smooth animation. This time we turn to [Sawaiz Syed] and [Open Kolibri] to deliver the brightly lit goods with the minuscule HALO 90 reactive LED earrings.
The HALO 90’s are designed to work as earrings, though we suspect they’d make equally great brooches, hair accessories, or desk objects. To fit this purpose each one is a minuscule 24 mm in diameter and weighs a featherweight 5.2 grams with the CR2032 battery (2.1 g for the PCBA alone). Functionally their current software includes three animation modes, each selectable via a button on device; audio reactive, halo (fully lit), and sparkle. Check out the documentation for details on expected battery life in each mode, but suffice to say that no matter what these earrings will make it through a few nights out.
In terms of hardware, the HALO 90’s are as straightforward as you’d expect. Each device is driven by an STM8 at its maximum 16MHz which is more than fast enough to keep the 90 charliplexed 0402 LEDs humming along at a 1kHz update rate, even with realtime audio processing. In fact the BOM here is refreshingly simple with just 8 components; the LEDs, microcontroller and microphone, battery holder and passives, and the button. [Sawaiz] even designed an exceptionally slick case to go with each pair of earrings, which holds two HALO 90’s with two CR2032’s and includes a magnetic closure for the most satisfying lid action possible.
As with some of his other work, [Sawaiz] has produced a wealth of exceptional documentation to go with the HALO 90’s. They’re available straight from him fully assembled, but with documentation this good the path to a home build should be well lit and accessible. He’s even chosen parts with an eye towards long availability, low cost, and ease of sourcing so no matter when you decide to get started it should be a snap.
It was difficult to choose just a few images from [Sawaiz]’s mesmerizing collection, so if you need more feast your eyes on the expanded set after the break.
“You could’ve done that with a 555 timer.” But what if all you have on hand is an ESP8266? [TechColab] needed to control a solenoid valve with a short pulse via a solid-state relay (SSR) but found that the trusty 555 timer was tricky to set properly. Additionally, they wanted to add features, such as fixed pulse length, that were difficult to implement—even with multiple timers. Still wanting to keep things cheap and accessible, [TechColab] has created the WiFive55, a 555 replacement based on the ESP-01 ESP8266 board.
[TechColab] began by investigating existing ESP-01 solid-state relay boards but found that many of them momentarily enable the output on startup—a risk [TechColab] deemed unacceptable. This was resolved in the WiFive55 by adding an RC filter to the SSR output, eliminating the output glitches at the cost of slowing switching time to around 20 ms—an acceptable trade for many SSR applications.
Since they were going to design a new PCB to support this improved ESP-01 SSR controller, [TechColab] decided to go all-out. To support loads of widely varying sizes, the PCB supports an optoisolator that switches up to 1 A, a MOSFET that switches up to 2 A, and an on-board relay or SSR that can switch up to 3 A. For heavy loads, it includes connections for an off-board SSR, which allow it to switch whatever current the SSR can handle (easily over 50 A). Because the ESP-01 is slightly more capable than the 555, the WiFive55 supports control via WiFi, GPIO, serial, and push-button. Keeping with the WiFive55’s original role as a 555 replacement, it even includes a header exposing a 555-like trigger and output interface!
We always like seeing inexpensive boards like the ESP-01 being used to their full potential, and we can’t wait to see what software [TechColab] cooks up for this! If you’re interested in getting started with the ESP-01, you might consider starting with this guide to blinking an LED over WiFi.
The 1980s and early 1990s were a bit of an odd time for semiconductor technology, with the various transistor technologies that had been used over the decades slowly making way for CMOS technology. The 1991-vintage IBM ES/9000 mainframe was one of the last systems to be built around bipolar transistor technology, with [Ken Shirriff] tearing into one of the processor modules (TCM) that made up one of these mainframes.
Five of these Thermal Conduction Modules (127.5 mm a side) made up the processor in these old mainframes. Most of note are the use of the aforementioned bipolar transistors and the use of DCS-based (differential current switch) logic. With the already power-hungry bipolar transistors driven to their limit in the ES/9000, and the use of rather massive DCS gates, each TCM was not only fed many amperes of electricity, but also capable of dissipating up to 600 Watts of power.
Each TCM didn’t contain a single large die of bipolar transistors either, but instead many smaller dies were bonded on a specially prepared ceramic layer in which the wiring was added through a very precise process. While an absolute marvel of engineering, the ES/9000 was essentially a flop, and by 1997 IBM too would move fully to CMOS transistor technology.
[TheStaticTurtle] built a custom controller for automating his garage doors. He wanted to retain the original physical button and RF remote control interfaces while adding a more modern wireless control accessible from his internet connected devices. Upgrading an old system is often a convoluted process of trial and error, and he had to discard a couple of prototype versions which didn’t pan out as planned. But luckily, the third time was the charm.
The original door-closer logic was pretty straightforward. Press a button and the door moves. If it’s not going in the desired direction, press the button once again to stop the motor, and then press it a third time to reverse direction. With help from the user manual diagrams and a bit of reverse-engineering, he was able to get a handle on how to plan out his add-on controller to interface with the old system.
There are many micro-controller options available these days when you want to add IoT to a project, but [TheStaticTurtle] decided to use the old faithful ESP8266 as the brains of his new controller. For his add-on board to work, he needed to detect the direction in which the motor was turning, and detect the limit switches when the door reached end of travel in either direction. Finally, he needed a relay contact in parallel with the activation button to send commands remotely.
To sense if the motor was moving in the “open” or “close” direction, he used a pair of back-to-back opto-couplers in parallel with the motor terminals. He connected another pair of opto-couplers across the two end-limit switches which indicated when the door was fully open or closed, and shut off the motor supply. Finally, a GPIO from the ESP8266 actuates a relay to send the door open and close commands. The boards were designed in EasyEDA and with a quick turnaround from China, he was able to assemble, test and debug his boards pretty quickly.
The code was written using the Arduino IDE and connects the ESP8266 to the MQTT server running on his home automation computer. The end result is a nice dashboard with three icons for open, close and stop, accessible from all the devices connected to his home network. A 3D printed enclosure attaches outside the original control box to keep things tidy. Using hot melt glue as light pipes for the status LED’s is a pretty nifty hack. If you are interested in taking a deeper look at the project, [TheStaticTurtle] has posted all resources on his Github repository.
Building your own CPU is arguably the best way to truly wrap your head around how all those ones and zeros get flung around inside of a computer, but as you can probably imagine even a relatively simple processor takes an incredible amount of time and patience to put together. Plus, more often than not you’re then left with a maze of wires and perfboards that takes up half your desk and doesn’t do a whole lot more than blink some LEDs.
But the Pineapple ONE, built by [Filip Szkandera] isn’t your average homebrew computer. Oh sure, it still took two years for him to design, debug, and assemble, his 32-bit RISC-V CPU and all its associated hardware; but the end result is a gorgeous looking machine that runs C programs and offers a basic interactive shell over VGA. In fact with its slick 3D printed enclosure, vertically stacked construction, and modular peripheral connections, it looks more like some kind of high-tech scientific instrument than a computer; homebrew or otherwise.
[Filip] says he was inspired to build this 500 kHz (yes, kilohertz) beauty using only discrete logic components by [Ben Eater]’s well known 8-bit breadboard computer and [Robert Baruch]’s LMARV-1 (Learn Me A RISC-V, version 1). He spent six months simulating the machine before he even started creating the schematics, let alone design the individual boards. He tried to keep all of his PCB’s under 100 x 100 mm to take advantage of discounts from the fabricator, which ultimately led to the decision to align the nine boards vertically and connect them together with pin headers.
In the video below you can see [Filip] start up the computer, call up a bit of system information, and even play a rudimentary game of snake before peeking and poking some of the machine’s 512 kB of RAM. It sounds like there’s still some work to be done and bugs to squash, but we’ve already seen enough to say this machine has more than earned entry into the pantheon of master-crafted homebrew computers.
We’re accustomed to seeing giant LED-powered screens in sports venues and outdoor displays. What would it take to bring this same technology into your living room? Very, very tiny LEDs. MicroLEDs.
MicroLED screens have been rumored to be around the corner for almost a decade now, which means that the time is almost right for them to actually become a reality. And certainly display technology has come a long way from the early cathode-ray tube (CRT) technology that powered the television and the home computer revolution. In the late 1990s, liquid-crystal display (LCD) technology became a feasible replacement for CRTs, offering a thin, distortion-free image with pixel-perfect image reproduction. LCDs also allowed for displays to be put in many new places, in addition to finally having that wall-mounted television.
Since that time, LCD’s flaws have become a sticking point compared to CRTs. The nice features of CRTs such as very fast response time, deep blacks and zero color shift, no matter the angle, have led to a wide variety of LCD technologies to recapture some of those features. Plasma displays seemed promising for big screens for a while, but organic light-emitting diodes (OLEDs) have taken over and still-in-development technologies like SED and FED off the table.
While OLED is very good in terms of image quality, its flaws including burn-in and uneven wear of the different organic dyes responsible for the colors. MicroLEDs hope to capitalize on OLED’s weaknesses by bringing brighter screens with no burn-in using inorganic LED technology, just very, very small.
A decent current measurement sensor ought to be an essential part of every hacker’s workbench. One that is capable of measuring DC, as well as low and high frequencies with reasonable accuracy. And bonus credits if it can also withstand high bus voltages – such as those found in mains utility or electric vehicle work. [Undersilicon] couldn’t find one that ticked all the boxes, so he built an ACS730 based AC/DC current probe capable of measuring up to 25 A at frequencies up to 1 MHz.
Allegro Microsystems has a wide offering of current sensor IC’s. The ACS730 features a -3 dB bandwidth of 1 MHz, and -1 dB bandwidth of 500 kHz. Since it is galvanically isolated, it can be used in AC mains applications up to 297 Vrms and for DC up to 420 V. And as he intended to use it as an oscilloscope accessory, the analog output suited the application nicely. A pair of precision op-amps provide the voltage output scaled to 100 mV/A. The board is powered off a 1000 mAh LiPo battery that can run the sensor for about 15 ~ 20 hours. The power supply section consists of a charge circuit for the LiPo, and a split rail dual output power supply converter for the op-amps.
The ACS730 has a 2.5 V output when measured current is zero, and is scaled for 40 mV/A. This gives an output voltage swing from -0.5 V for -50 A to +4.5 V for +50 A. This is where the AD823ARZ dual 16 MHz, Rail-to-Rail FET Input Amplifiers step in. One pair is used to obtain a 2.5 V reference from the 5 V supply, and also to buffer the analog output from the ACS730. The second pair subtracts the 2.5 V offset, and applies a gain of 2.5 to get the 100 mV/A output. Dual power supply for the op-amps comes from a TPS65133 Split-Rail Converter, ±5V, 250mA Dual Output Power Supply. Lastly, LiPo charging is handled by the MCP73831 Single Cell, Li-Ion/Li-Polymer Charge Management Controller.