Widgets, hacks, ideas, and tricks. A grab bag of ideas for prototyping, experimenting, or designing.

If you come across any technologies that you think might belong here in future issues, please let me know!

#1: Power-line communication

Instead of using a wireless connection to communicate data between devices in your home, consider using the common wired connection of the power cable!

I’ve heard conflicting reports that this only really seems to work for devices plugged into the same power point. Also, anything involving messing with mains power is very dangerous and I cannot in good conscience recommend you experiment with this unless you have some sort of appropriate certification or are 100% sure you know what you are doing. But it might be interesting to see what kinds of signals you can transmit and receive through the mains in different settings.

#2: Gimmick capacitor

If you’re prototyping and need a small (pF) [variable] capacitor that you don’t have, try twisting two pieces of wire together.

#3: Conductive glass

If you need conductive glass for a project but find market solutions too expensive and aren’t afraid of doing some chemical work, this experimenter has developed a method for producing it quite cheaply and written up his results.

#4: Chorded keyboards

One way you might get around a limited number of buttons on an input device is by setting up some key chording arrangement, whereby pressing multiple keys at once results in a new input. This is taken to the extreme with stenography/stenotype, which is used by court reporters and closed captioners to type at the speed of talking.

Obviously this has the tradeoff of requiring the user to learn the chords.

For regular keyboards and personal computer software, you might want to look into Plover, which is a FOSS stenography engine that can work with regular keyboards. (Figuring out whether your keyboard has enough rollover capability to handle this might be kind of annoying though.)

#5: Candles work as infrared (IR) beacons

An ancient hack for the Nintendo Wii (but which could be applied to similar IR-based systems): the Wii “sensor bar” is effectively just two point infrared light sources that the Wiimote can “see” and use as a position reference. You know what else can act as a point infrared light source? Anything incandescent! As this video from 2006 demonstrates, if you have a broken Wii sensor bar and are unfazed by fire hazards, you can replace it very cheaply by lighting two candles and placing them where the two ends of the sensor bar would be.

…Y’know, TV remotes usually use sequences of IR signals… I wonder if you could build a “caveman remote” that has a candle behind some opaque card or disk with the right pattern of slits or holes punched out. If you spun the disk at the right frequency, could you get the pattern of light leaving the disk to match up with a TV remote control code? Or is there some reason this wouldn’t work?

#6: VR “Lighthouse” tracking system

(see also this Gizmodo article)

The technology being described here is actually not very different to the Wiimote technology above – you have stationary beacons that emit infrared light and serve as reference points for the controllers whose position you want to determine – with the addition of “rapidly spinning lasers”.

#7: Video game interfaces

It might make sense to reuse interface elements or paradigms from video games for a couple reasons: users might be familiar with the interface if they’ve played the sorts of games it’s based on; development might be faster and easier working in a game engine rather than from scratch; most games try to make player interactions as smooth as possible, so you have yourself a tried and true design; and, of course, there’s an undeniable whimsy to the whole idea.

Consider the following:

  • Ren’Py: Ren’Py is a very capable visual novel engine that makes writing dynamic cross-platform visual novels relatively easy. It also allows you to run (arbitrary?) Python code as part of the novel scripting, meaning you could very well embed a fairly complex application inside of it.
    • MSCHF’s Tax Heaven 3000 is written in Ren’Py and implements tax filing software as a visual novel. (If you’re not familiar with MSCHF and are thinking “WTF” after reading “tax filing software visual novel”, you have to check out some of their other stuff. I promise you won’t regret it.)
    • Doki Doki Literature Club was written in Ren’Py and does some fairly complex stuff with files as it’s running, although it is just a game.
  • Tom’s Timer is written in GameMaker.
  • I feel like text adventure interfaces have a lot of potential in terms of acting as user interfaces for non-game applications, but while text adventure programming languages like Inform have been around for a very long time, I haven’t heard of anyone doing anything like setting up triggers or hooks in the game environment to run other pieces of code. But imagine if you could do something like boot up a text adventure, flip a light switch in the game and have it come on in real life through IoT magic.
    • This would require a modified Z-machine implementation at any rate.
  • Strategy games are another one. What if you had a project management app that played like Civilization?

#8: Kite antennas

A kite is one possible way to set up a portable, long, vertical antenna for amateur (or other) radio. The military has used them historically, although usually only when there aren’t any other options, like in emergencies. Lots of hams have written about their experiments with kite antennas – a quick internet search yields posts by KM1NDY and VK3YE.

#9: DIP abuse

This article covers how the author managed to contort a DIP MCU into something resembling surface-mount for a PCB by bending the chip’s pins (!). Even if you don’t plan on doing something that scuffed in the near future, the article has a good section on how to do custom chip footprints in KiCad.

#10: Diode sampling

Measuring an 18 GHz signal on a 4 MHz oscilloscope might seem impossible… well, not if you use diode sampling. The basic idea is to generate a lower frequency oscillation that you can actually measure, then use that to trigger extremely short pulses at regular intervals, then sample your signal using those pulses and Schottky diodes. Don’t worry, the video explains the whole thing properly. And yes, DSP enjoyers, this basically weaponises aliasing. (Of course the bandwidth of your signal is still going to be limited; there’s no easy fix for that.)

I am a bit of a silly goose and still didn’t quite understand how the whole thing worked, because how are you sampling 18 GHz signals without having an 18 GHz reference signal? What the heck? I was combing through the documentation the YouTuber helpfully uploaded looking for the 18 GHz generator, misunderstanding the theory of operation completely. But there is no 18 GHz generator, of course. Just reread the previous paragraph or watch the video to realise that you just need a relatively low VFO (in this case it “only” goes up to 133.3 MHz), step recovery diodes, and the diode sampling circuit.