Dodecahedron Speaker Is Biblically Accurate

Once upon a time, many radios and TVs only came with a single (mono) speaker. Then someone decided all audio hardware should have as many speakers as we have ears. That was until [Olivia] came along, and whipped up a dodecahedron speaker as an educational piece for workshops. Really, it shows us that twelve speakers should be the minimum standard going forward.

The speaker relies on a 3D-printed frame. The dodecahedron shell is assembled from 12 individual faces, each of which hosts a small individual speaker. Multichannel audio fans shouldn’t get too excited—all twelve speakers are wired to the same input in four groups of three, making this essentially an exceptionally complicated mono device. It might sound silly, but it’s actually a great way to deliver audio in many directions all at once. [Olivia] even went to the effort of running some sweep tests in anechoic and reverberation chambers to see how they performed, which is a fun bit of extra detail in the build log.

[Olivia] notes that these unique speakers are great as a beginner workshop build. They’re easy to modify in various ways to suit different ideas or levels of ability, and they can be made for less than $30 a pop. We’d love to see an advanced version that maybe packed in a lithium battery and a Bluetooth module to make them a standalone audio device. Video after the break.

Continue reading “Dodecahedron Speaker Is Biblically Accurate”

How A Failed Video Format Spawned A New Kind Of Microscope

The video cassette tape was really the first successful home video format; discs just couldn’t compete back in the early days. That’s not to say nobody tried, however, with RCA’s VideoDisc a valiant effort that ultimately fell flat on its face. However, the forgotten format did have one benefit, in that it led to the development of an entirely new kind of microscope, as explained by IEEE Spectrum.

The full story is well worth the read; the short version is that it all comes down to capacitance. RCA’s VideoDisc format was unique in that it didn’t use reflective surfaces or magnetic states to represent data. Instead, the data was effectively stored as capacitance changes. As a conductive stylus rode through an undulating groove in a carbon-impregnated PVC disc, the capacitance between the stylus and the disc changed. This capacitance was effectively placed into a resonant circuit, where it would alter the frequency over time, delivering an FM signal that could be decoded into video and audio by the VideoDisc player.

The VideoDisc had a capacitance sensor that could detect such fine changes in capacitance, that it led to the development of the Scanning Capacitance Microscope (SCM). The same techniques used to read and inspect VideoDiscs for quality control could be put to good use in the field of semiconductors. The sensors were able to be used to detect tiny changes in capacitance from dopants in a semiconductor sample, and the SCM soon became an important tool in the industry.

It’s perhaps a more inspiring discovery than when cheeky troublemakers figured out you could use BluRay diodes to pop balloons. Still fun, though. An advertisement for the RCA VideoDisc is your video after the break.

Continue reading “How A Failed Video Format Spawned A New Kind Of Microscope”

Build Your Own 6K Camera

[Curious Scientist] has been working with some image sensors. The latest project around it is a 6K camera. Of course, the sensor gives you a lot of it, but it also requires some off-the-shelf parts and, of course, some 3D printed components.

An off-the-shelf part of a case provides a reliable C mount. There’s also an IR filter in a 3D-printed bracket.

Continue reading “Build Your Own 6K Camera”

Calculator Battery Mod Lets You Go The Distance

Disposable batteries seem so 1990s. Sure, it’s nice to be able to spend a couple of bucks at the drugstore and get a flashlight or TV remote back in the game, but when the device is a daily driver, rechargeable batteries sure seem to make more financial sense. Unfortunately, what makes sense to the end user doesn’t always make sense to manufacturers, so rolling your own rechargeable calculator battery pack might be your best option.

This slick hack comes to us from [Magmabow], who uses a Casio FXCG50 calculator, a known battery hog. With regular use, it goes through a set of four alkaline AA batteries every couple of months, which adds up quickly. In search of a visually clean build, [Magmabow] based the build around the biggest LiPo pillow-pack he could find that would fit inside the empty battery compartment, and planned to tap into the calculator’s existing USB port for charging. A custom PCB provides charging control and boosts the nominal 3.7-volt output of the battery to the 5-ish volts the calculator wants to see. The PCB design is quite clever; it spans across the battery compartment, with its output feeding directly into the spring contacts normally used for the AAs. A 3D-printed insert keeps the LiPo and the PCB in place inside the battery compartment.

Almost no modifications to the calculator are needed, other than a couple of bodge wires to connect the battery pack to the calculator’s USB port. The downside is that the calculator’s battery status indicator won’t work anymore since the controller will just shut the 5-volt output down when the LiPo is discharged. It seems like there might be a simple fix for that, but implementing it on such a small PCB could be quite a challenge, in which case a calculator with a little more room to work with might be nice. Continue reading “Calculator Battery Mod Lets You Go The Distance”

Automatic Feeder Keeps Fish Sated

[Noisy Electrons] is a maker who also likes to keep fish. He sometimes needs to travel and keep his fish fed in the meantime, so he created an automated solution to handle that for him.

The build is based around an STM32 microcontroller, paired with a MCP7940N real-time clock to keep time. The microcontroller is hooked up to a few buttons and a small display to serve as an interface, allowing the feeding times and dosage amounts to be configured right on the device. Food is distributed from a 3D printed drum with a hole in it, which is rotated via a stepper motor.  Each time the drum rotates, some food falls through the hole and into the tank. Dosage amount is measured in rotations — the more times the drum rotates, the more food is delivered to the fish.

[Noisy Electron] built three of these devices for three separate tanks. Thus far, it’s been three weeks and all the fish are still alive, so we’ll take that as a vote of confidence in the build. We’ve featured some other great pet feeders over the years, too Continue reading “Automatic Feeder Keeps Fish Sated”

2025 Hackaday Superconference: Announcing Our Workshops And Tickets

Can you feel the nip of fall in the air? That can only mean one thing: Supercon is just around the corner. The next few weeks are going to bring a blitz of Supercon-related reveals, and we’re starting off with a big one: the workshops.

Supercon is the Ultimate Hardware Conference, and you need to be there to attend a workshop. Both workshop and general admission tickets are on sale now! Don’t wait — they sell out fast.

Continue reading “2025 Hackaday Superconference: Announcing Our Workshops And Tickets”

Moondream title with man's face visible in background.

Using Moondream AI To Make Your Pi “See” Like A Human

[Jaryd] from Core Electronics shows us human-like computer vision with Moondream on the Pi 5.

Using the Moondream visual language model, which runs directly on your Raspberry Pi, and not in the cloud, you can answer questions such as “are the clothes on the line?”, “is there a package on the porch?”, “did I leave the fridge open?”, or “is the dog on the bed?” [Jaryd] compares Moondream to an alternative visual AI system, You Only Look Once (YOLO).

Processing a question with Moondream on your Pi can take anywhere from just a few moments to 90 seconds, depending on the model used and the nature of the question. Moondream comes in two varieties, based on size, one is two billion parameters and the other five hundred million parameters. The larger model is more capable and more accurate, but it has a longer processing time — the fastest possible response time coming in at about 22 to 25 seconds. The smaller model is faster, about 8 to 10 seconds, but as you might expect its results are not as good. Indeed, [Jaryd] says the answers can be infuriatingly bad.

In the write-up, [Jaryd] runs you through how to use Moonbeam on your Pi 5 and the video (embedded below) shows it in action. Fair warning though, Moondream is quite RAM intensive so you will need at least 8 GB of memory in your Pi if you want to play along.

If you’re interested in machine vision you might also like to check out Machine Vision Automates Trainspotting With Unique Full-Length Portraits.

Continue reading “Using Moondream AI To Make Your Pi “See” Like A Human”