[sorki] had an ESP-12F and wanted to play with nodeMCU, but found they were lacking buttons for reset & flash. We’ve all been there – mucking about with a project on a breadboard, trying to save the time required to solder up a button by shorting pins with wire or bending component legs to touch. This either doesn’t work or ends up bricking the microcontroller when it inevitably goes wrong. [Buger] found a tidier solution to adding buttons to the ESP-12F with the minimum of effort.
It’s the spirit of deadbug applied to buttons. One side of a piece of wire is soldered to the pin needing to be pulled down. Component leg offcuts are ideal for this. The other end of the wire is bent up and left to float over the metal shield of the ESP-12, which is connected to ground. When you want the pin to go low, press the wire into the shield, grounding it. Let it go, and the pin returns high again, assuming your pullup resistors are all in order.
It’s a quick hack that’s much more robust than trying to hold two ends of a piece of hookup wire in place. It’s also still easier than trying to find a tactile switch solder leads to, and you don’t end up having it hanging off the board either.
You think you’ve seen everything that there is to see regarding blinking LEDs and then a simple little trick proves you wrong. Our friend [Zach Fredin], aka [Zakqwy], added a pander mode to his blinky board which shows the Hackaday Jolly Wrencher in a Persistence of Vision mode. We love pandering, and obviously you just need to start the mode and wave the board back and forth. But in thinking the obvious you’d be wrong.
You would think that soldering all those LEDs by hand would be the trick, but [Zach] pulled off a much more difficult feat. Look closely at the image here (or click to embiggen). The two shift register footprints on the prototype were mirrored. He deadbug soldered each of them using — get this — the individual strands from some 28 AWG stranded wire. You sir, get the hardcore hand soldering badge and then some.
Okay, we’ll stop beating around the bush. The ATtiny45 on this board isn’t connected to the USB data lines, they’re only for power. That means, at its heart this is purely a blinking LED project, albeit one that uses the huge range of colors of the PICOLED family of parts. [Zach] did well with just two user inputs, but it’s the very simple POV party trick that really sucked us in. Instead of waving the board around, [Zach] uses a metal offset spatula as a mirror. Moving it back and forth unfolds the carefully timed flashes to draw your message in the air. Such a simple concept, but so satisfying to see it applied in a slightly different way.
Back in the 1980s I was a budding electronics geek working in a TV repair shop. I spent most of my time lugging TVs to and from customers, but I did get a little bench time in. By then new TVs were entirely solid-state and built on single PC boards, but every once in a while we’d get an old-timer in with a classic hand-wired tube chassis. I recall turning them over, seeing all the caps and resistors soldered between terminal strips bolted to the aluminum chassis and wondering how it could all possibly work. It all looked so chaotic and unkempt compared to the sleek traces and neat machine-inserted components on a spanking new 19″ Zenith with the System 3 chassis. In a word, the old chassis was just – ugly.
Looking back, I probably shouldn’t have been so judgmental. Despite the decades of progress in PCB design and the democratization of board production thanks to KiCad, OSH Park, and the like, it turns out there’s a lot to be said for ugly methods of circuit construction.
The first integrated circuits weren’t tiny flecks of silicon mounted to metal carriers and embedded in epoxy or ceramic. The first integrated circuits, albeit a looser definition of such, were just a few transistors, resistors, and diodes mashed together in the same package. With this in mind, [Rupert] created his own custom IC. It’s an IR receiver transmitter constructed out of a transistor, resistor, and an LED.
The attentive reader should be asking, “wait, can’t you just buy an IR receiver transmitter?” Yes, yes you can. But that’s not a hack™, and would otherwise be very uninteresting.
[Rupert]’s IC is just three parts, a 2n2222 transistor, a 220Ω resistor and an IR LED. With a good bit of deadbug soldering, these three parts were melded into something that resembled, and had the same pinout of, a Vishay TSOP4838 IR receiver. The epoxy used to encapsulate this integrated circuit is a standard 2-part epoxy and laser printer toner. Once everything is mixed up into a gooey slurry, it’s dripped over the IC producing a blob of an integrated circuit. It’s functionally identical to the standard commercial version, and looks good enough for a really cool project [Rupert]’s been working on.
[Phil] has already built a few clocks with Nixies, VFDs, and LED matrices. When his son requested his own clock, he wanted to do something a little different. Inspired by the dead bug style of [Jim Williams]’ creations, [Phil] set out to build a clock made entirely out of discrete components. That includes the counters, driver circuits, and an array of LED.
There are a few inspiration pieces for [Phil]’s clock, starting with the Transistor Clock, a mains-powered clock that uses 194 transistors, 566 diodes, and exactly zero integrated circuits. Design patterns from a clock so beautiful it’s simply called The Clockare also seen, as is a Dekatron emulator from [VK2ZAY].
[Phil]’s creation has no PCB, and all the components are soldered onto tiny wires arranged into something resembling the clocks circuit. It’s a fantastic contraption, and while we’ll still have to give the design award to the clock, [Phil]’s creation shows off the functional circuits; great if he’ll ever need to debug anything.
If you’ve never heard of “Manhattan Style” circuit construction, you’re not alone. Popular in ham radio circles, the process looks nicer than straight dead bug style circuit building, but not as involved as etching your own PCB – consider it a nice middle of the road solution.
This type of construction is often used to build circuits inside enclosures that are made of copper clad, which is a somewhat common practice among ham radio operators. Manhattan Style circuits are built using glued-on metal pads to which components are mounted. One might think that the large pads you see in the image above would limit you to through-hole components, but that’s definitely not the case. A wide array of SMD pads are available in common pin configurations as well, allowing you to use pretty much any type of component you prefer.
While it might not be appropriate for every project you work on, Manhattan Style circuits and copper clad boxes definitely add a nice touch to certain items, like the Wheatstone Bridge Regenerative Receiver you see above.
Here’s a no-PCB Arduino that doesn’t obscure the DIP footprint of the AVR chip. It’s built on an ATmega88 chip, and includes a programming header, reset button, a couple of filtering caps, and an LED. This is modeled after the Lilypad hardware, and fits nicely on top of the plastic case of the microcontroller, allowing it to be used in a breadboard or DIP socket. You can see a walk through of the components in the clip after the break.
We don’t really need most of the components on top of the chip (especially the status LED on the SCK line), but there are several things that we like about this. First off, the programming header is extremely nice. We could see this coming in handy for prototyping where you don’t want to add a header to your final design. Just use a chip socket, and this chip while you’re developing firmware. Once everything is dialed in, program a naked chip and swap the two. The same goes for the reset button, which is nice when working on firmware but may not be necessary in your final design.
This is quite an old project, and we’ve actually seen a successor to it. This is Rev. 2 and we looked in on Rev. 7 back in March. That one is a full Arduino, but the circuit board has no substrate.