Chicken Coop Door Performs In Harsh Environment

April 3, 2026 by Bryan Cockfield 22 Comments

One of the pitfalls of modern engineering is that it’s entirely possible to end up in a situation where a product or solution has been designed by someone who has never left a desk. Which wouldn’t be a problem if things didn’t have a tendency to work differently in real life than they do in theory.

One of those things is automatic chicken coop doors, which have to operate reliably in not only a wide range of climates but with a number of possible physical limitations as well. [Vinnie] has taken on the challenge of building one which actually accomplishes all of these tasks, after realizing that the off-the-shelf solutions were victims of design over practicality.

[Vinnie] designed this door to be operated by the one thing that’s always 100% reliable: gravity. A linear actuator lifts the door at the beginning of the day, and then at night it’s allowed to fall back down in its track. A latch secures it against smarter intruders like raccoons. [Vinnie] has found that this lifting mechanism holds up much better in mud, snow, ice, and other difficult conditions than any other method he’s tried so far.

The system is built around a ATmega1284P, and calculates the sunrise and sunset times each day to know when to open or close the door. He’s built the system as a state machine which makes it more robust during power outages, which is a necessity since his chicken coop is mobile and is frequently out of range of WiFi and is battery powered.

The approach [Vinnie] takes to automation is something that has application outside of his own farmstead. Using state machines instead of schedules, ensuring the design is as simple as possible and works within its environment, and minimizing reliance on electric and data infrastructure can go a long way to solving problems that might not appear when designing something on paper.

He’s been automating many other things on his farm as well, and it’s worth checking it out if you haven’t seen it already.

Battery Tester Outperforms Cheaper Options

April 1, 2026 by Bryan Cockfield 4 Comments

Batteries are notoriously difficult pieces of technology to deal with reliably. They often need specific temperatures, charge rates, can’t tolerate physical shocks or damage, and can fail catastrophically if all of their finicky needs aren’t met. And, adding insult to injury, for many chemistries, the voltage does not correlate to state of charge in meaningful ways. Battery testers take many efforts to mitigate these challenges, but often miss the mark for those who need high fidelity in their measurements. For that reason, [LiamTronix] built their own.

The main problem with the cheaper battery testers, at least for [LiamTronix]’s use cases, is that he has plenty of batteries that are too large to practically test on the low-current devices, or which have internal battery management systems (BMS) which can’t connect to these testers. The first circuit he built to help solve these issues is based on a shunt resistor, which lets a smaller IC chip monitor a much larger current by looking at voltage drop across a resistor with a small resistance value. The Pi uses a Python script which monitors the current draw over the course of the test and outputs the result on a handy graph.

This circuit worked well enough for smaller batteries, but for his larger batteries like the 72V one he built for his electric tractor, these methods could draw far too much power to be safe. So from there he built a much more robust circuit which uses four MOSFETs as part of four constant current sources to sink and measure the current from the battery. A Pi Zero monitors the voltage and current from the battery, and also turns on some fans pointed at the MOSFETs’ heat sink to keep them from overheating. The system can be configured to work for different batteries and different current draw rates, making it much more capable than anything off the shelf.

Continue reading “Battery Tester Outperforms Cheaper Options” →

A Better Jogging Stroller

April 1, 2026 by Bryan Cockfield 12 Comments

Although the jogging stroller is a fixture of suburban life, allowing parents the opportunity to get some exercise while letting their young children a chance for some fresh air, it would seem like the designers of these strollers have never actually gone for a jog. Requiring a runner to hold their hands at fixed positions can be incredibly uncomfortable and disrupts most people’s strides and cadence — so [John] attempted to solve the problem after finding one of these strollers on the secondhand market.

While there are some purpose-built strollers that attempt to address these issues, they can be pricey. Rather than shell out for a top-dollar model, [John] got to work with his 3D printer and created a prototype device that allows him to attach the stroller at his waist while leaving his hands free. There were a few problems to overcome here, the first of which would cause the device to buckle under certain loading situations. This was solved with some small pieces of rope which act as flexible bump stops, keeping the hinge mechanism from binding up. Another needed to be solved with practice, which was that it took some time to be able to steer the stroller without using one’s hands.

As an added bonus, [John] also included a system that tracks the distance the stroller has traveled. Using a hall effect sensor and a magnet attached to the wheel, a small microcontroller is able to quickly calculate distance and display it on a tiny screen mounted near the handlebars. Although smartphones are handy, their GPS systems can be surprisingly inaccurate, so a system like this can be a better indicator since it’s being directly measured. All in all, not a bad few upgrades to a secondhand stroller.

Continue reading “A Better Jogging Stroller” →

Recreating One Of The First Hackintoshes

March 30, 2026 by Bryan Cockfield 22 Comments

Apple’s Intel era was a boon for many, especially for software developers who were able to bring their software to the platform much more easily than in the PowerPC era. Macs at the time were even able to run Windows fairly easily, which was unheard of. A niche benefit to few was that it made it much easier to build Hackintosh-style computers, which were built from hardware not explicitly sanctioned by Apple but could be tricked into running OSX nonetheless. Although the Hackintosh scene exploded during this era, it actually goes back much farther and [This Does Not Compute] has put together one of the earliest examples going all the way back to the 1980s.

The build began with a Macintosh SE which had the original motherboard swapped out for one with a CPU accelerator card installed. This left the original motherboard free, and rather than accumulate spare parts [This Does Not Compute] decided to use it to investigate the Hackintosh scene of the late 80s. There were a few publications put out at the time that documented how to get this done, so following those as guides he got to work. The only original Apple part needed for this era was a motherboard, which at the time could be found used for a bargain price. The rest of the parts could be made from PC components, which can also be found for lower prices than most Mac hardware. The cases at the time would be literally hacked together as well, but in the end a working Mac would come out of the process at a very reasonable cost.

[This Does Not Compute]’s case isn’t scrounged from 80s parts bins, though. He’s using a special beige filament to print a case with the appropriate color aesthetic for a computer of this era. There are also some modern parts that make this style computer a little easier to use in today’s world like a card that lets the Mac output a VGA signal, an SD card reader, and a much less clunky power supply than the original would have had. He’s using an original floppy disk drive though, so not everything needs to be modernized. But, with these classic Macintosh computers, modernization can go to whatever extreme suits your needs.

Thanks to [Stephen] for the tip!

Continue reading “Recreating One Of The First Hackintoshes” →

Training A Transformer With 1970s-era Technology

March 29, 2026 by Bryan Cockfield 15 Comments

Although generative language models have found little widespread, profitable adoption outside of putting artists out of work and giving tech companies an easy scapegoat for cutting staff, their their underlying technology remains a fascinating area of study. Stepping back to the more innocent time of the late 2010s, before the cultural backlash, we could examine these models in their early stages. Or, we could see how even older technology processes these types of machine learning algorithms in order to understand more about their fundamentals. [Damien Boureille] has put a 60s-era IBM as well as a PDP-11 to work training a transformer algorithm in order to take a closer look at it.

For such old hardware, the task [Damien Boureille] is training his transformer to do is to reverse a list of digits. This is a trivial problem for something like a Python program but much more difficult for a transformer. The model relies solely on self-attention and a residual connection. To fit within the 32KB memory limit of the PDP-11, it employs fixed-point arithmetic and lookup tables to replace computationally expensive functions. Training is optimized with hand-tuned learning rates and stochastic gradient descent, achieving 100% accuracy in 350 steps. In the real world, this means that he was able to get the training time down from hours or days to around five minutes.

Not only does a project like this help understand these tools, but it also goes a long way towards demonstrating that not every task needs a gigawatt datacenter to be useful. In fact, we’ve seen plenty of large language models and other generative AI running on computers no more powerful than an ESP32 or, if you need slightly more computing power, on consumer-grade PCs with or without GPUs.

Luthier Crafts Guitar From Cardboard

March 27, 2026 by Bryan Cockfield 18 Comments

The people at Signal Snowboards are well known not only for producing quality snowboards, but doing one-off builds out of unusual and perhaps questionable materials just to see what’s possible. From pennies to glass, if it can go on their press (and sometimes even if it can’t) they’ll build a snowboard out of it. At some point, they were challenged to build different types of boards from paper products which resulted in a few interesting final products, but this pushed them to see what else they could build from paper and are now here with an acoustic guitar fashioned almost entirely from cardboard.

For this build, the luthiers are modeling the cardboard guitar on a 50s-era archtop jazz guitar called a Benedetto. The parts can’t all just be CNC machined out of stacks of glued-up cardboard, though. Not only because of the forces involved in their construction, but because the parts are crucial to a guitar’s sound. The top and back are pressed using custom molds to get exactly the right shape needed for a working soundboard, and the sides have another set of molds. The neck, which has the added duty of supporting the tension of the strings, gets special attention here as well. Each piece is filled with resin before being pressed in a manner surprisingly similar to producing snowboards. From there, the parts go to the luthier in Detroit.

At this point all of the parts are treated similarly to how a wood guitar might be built. The parts are trimmed down on a table saw, glued together, and then finished with a router before getting some other finishing treatments. From there the bridge, tuning pegs, pickups, and strings are added before finally getting finished up. The result is impressive, and without looking closely or being told it’s made from cardboard, it’s not obvious that it was the featured material here.

Some of the snowboards that Signal produced during their Every Third Thursday series had similar results as well, and we actually featured a few of their more tech-oriented builds around a decade ago like their LED snowboard and another one which changes music based on how the snowboard is being ridden.

Continue reading “Luthier Crafts Guitar From Cardboard” →

Modern Smartphone Vs. 80s Supercomputer

March 26, 2026 by Bryan Cockfield 48 Comments

One of the most common ways of comparing the processing power of some microcontroller or older smartphone in a fantastical way was to say that they had more processing power than the Apollo Guidance Computer. While this sounds impressive on the surface, the AGC was the first integrated circuit computer ever built and is predictably under-powered by almost all modern standards. A more apt comparison would be to compare a smartphone to a supercomputer from some bygone era, and someone has recently done just that.

The linked article looks at a modern iPhone 17 compared to the Cray 2 supercomputer. When the Cray 2 was first built in the mid 80s, it was the fastest computer in the world at 1.9 GFLOPS using four vector processors. A modern iPhone is estimated to have slightly more than that, so in some ways the iPhone comes out on top.

However, the Cray 2 was built with vector processors, a specialized type of processor meant to perform rapid calculations on specific types of data sets. So the Cray 2 may have been faster at these types of tasks than the more general-purpose A19 processor, and the A19 may have the edge in other tasks.

The other major difference the article doesn’t discuss is what software runs on these computers. The Cray 2 supercomputer ran a modified version of UNIX System V, which at the time was owned by AT&T (and which ran on plenty of other computers as well). Although proprietary in some sense, it was much more open than Apple’s iOS operating system, allowing users to run whatever software they wanted to run on the supercomputers that they bought and paid for, and to modify many parts of the operating system itself. In that sense, the Cray will always maintain the edge over Apple and their walled garden.