PURE Modules Aim to Make Prototyping Easier

[Sashi]’s PURE modules system wants your next wireless microcontroller and sensor module project to be put together using card-edge connectors. But it’s a lot deeper than that — PURE is an entire wireless gadget development ecosystem. Striking a balance between completeness and modularity is very difficult; a wire can carry any imaginable electronic signal, but just handing someone a pile of wires presents them a steep learning curve. PURE is at the other end of the spectrum: everything is specified.

So far, two microcontroller options are available in the system, the nRF52 series and TI’s CC2650. Both of these run the Contiki OS, so it doesn’t matter which of these you choose. Wired data is all transmitted over I2C and connects up via the previously-mentioned card-edge connectors. On the wireless side, data transport is handled through an MQTT broker, using the MQTT-sn variant which is better suited to small radio devices. At the protocol layer everything uses Protocol Buffers, Google’s newest idea for adding some structure to the data.

Continue reading “PURE Modules Aim to Make Prototyping Easier”

Make Mulled Wine With A Processor Heatsink!

Now, over the holiday season there seems to be a predilection towards making merry and bright. As many an engineer and otherwise are sure to note, fine alcohols will facilitate this process. One such warm holiday beverage is mulled wine; there are many traditions on how to make it, but a singular approach to preparing the beverage would be to re-purpose an old PC and a CPU liquid cooling unit into a mulled wine heating station.

Four years ago, [Adam] found himself staring at a pile of mostly obsolete PCs in his IT office and pondering how they could be better used. He selected one that used a power-hungry Pentium 4 — for its high heat output — strapped a liquid cooling block to the CPU and pumped it full of the holiday drink. It takes a few hours to heat three liters of wine up to an ideal 60 Celsius, but that’s just in time for lunch! The Christmastime aroma wafting through the office is nice too.

Continue reading “Make Mulled Wine With A Processor Heatsink!”

Simon Says Smile, Human!

The bad news is that when our robot overlords come to oppress us, they’ll be able to tell how well they’re doing just by reading our facial expressions. The good news? Silly computer-vision-enhanced party games!

[Ricardo] wrote up a quickie demonstration, mostly powered by OpenCV and Microsoft’s Emotion API, that scores your ability to mimic emoticon faces. So when you get shown a devil-with-devilish-grin image, you’re supposed to make the same face convincingly enough to fool a neural network classifier. And hilarity ensues!

Continue reading “Simon Says Smile, Human!”

Cat vs. Human Escalates with Armor-Plated Feeder

Wars generally increase innovation as the opposing sides try to kill each other in ever more efficient ways. Even the soft war waged daily between felines and their human servants results in innovation, to wit we offer this armor-plated automated cat feeder.

The conflict between [Sprocket H.G. Shopcat] and her human [Quinn Dunki] began with a thoughtfully provided automatic food dispenser. Like human vending machine customers who witness a just-purchased bag of Cheesy Poofs dangling on the end of the dispense auger, [Sprocket] learned that the feeder would dispense a few fishy nuggets when nudged. [Quinn] embarked on an iterative design process to control [Sprocket]’s off-schedule snacking. Fastening the feeder firmly to the floor, and adding obstructions to prevent her from pawing up the dispense chute — nothing seemed to stop the clever feline’s raids. [Quinn] then pulled out all the stops and whipped up a [Sprocket]-safe enclosure for the feeder from 1/8″ plate steel and copper. This seems to have put the cat back on the straight and narrow, and it doesn’t look half bad either.

All kidding aside, [Quinn]’s approach to this problem is pretty instructive. Careful observations informed several cycles of reasonable modifications until it became clear that only the most extreme solution would work. There’ve been tons of cat feeders here before, from the simple to the complex, but we think all would fall prey to the clever [Sprocket] without a little up-armoring.

Continue reading “Cat vs. Human Escalates with Armor-Plated Feeder”

Halogen Lamp Abused for Desoldering

[Moony] thought that it was unconscionable that IR soldering stations sell for a few hundred Euros. After all, they’re nothing more than a glorified halogen lightbulb with a fancy IR-pass filter on them. Professional versions use 100 W 12 V DC bulbs, though, and that’s a lot of current. [Moony] tried with a plain-old 100 W halogen lightbulb. Perhaps unsurprisingly, it worked just fine. Holding the reflector-backed halogen spotlight bulb close to circuit boards allows one to pull BGAs and other ornery chips off after a few minutes. Voila.

[Moony] reasons that the IR filter is a waste anyway, since the luminous efficiency of halogen lights is so low: around 3.5%. And that means 96.5% heat! But there’s still a lot of light streaming out into a very small area, so if you’re going to look at the board as you de-solder, you’re really going to need a pair of welding goggles. Without, you’ll have a very hard time seeing your work at best, and might actually do long-term damage to your retinas.

So the next time you’re feeling jealous of those rework factory workers with their fancy IR soldering stations, head on down to the hardware store, pick up a gooseneck lamp, a 100 W halogen spotlight, and some welding goggles. And maybe a fire brick. You really don’t want your desk going up in flames.

We love make-do hacks, but we love doing it right, too. Just watch [Bil Herd] extol the virtues of a real IR desoldering station. And then giggle as you do the same thing with a few-dollar halogen bulb.

Continue reading “Halogen Lamp Abused for Desoldering”

Light Pipes and LEDs Team Up for a Modern Take on the Nixie Tube

There’s no doubting the popularity of Nixie tubes these days. They lend a retro flair to modern builds and pop up in everything from clocks to weather stations. But they’re not without their problems — the high voltage, the limited tube life, and the fact that you can have them in any color you want as long as it’s orange. Seems like it might be time for a modern spin on the Nixie that uses LEDs and light pipes. Meet Nixie Pipes.

Inspired by an incandescent light-pipe alphanumeric display from a 1970s telephone exchange, [John Whittington]’s design captures the depth and look of a Nixie by using laminated acrylic sheets. Each layer is laser etched with dots in the shape of a character or icon, and when lit from below by a WS2812B LED, the dots pick up the light and display the character in any color. [John]’s modular design allows one master and an arbitrary number of slaves, so large displays can simply be plugged together. [John] is selling a limited run of the Nixie Pipes online, but he’s also open-sourced the project so you can build your own modules.

We really like the modularity and flexibility of Nixie Pipes, and the look is pretty nice too. Chances are good that it won’t appeal to the hardcore Nixie aficionado, though, in which case building your own Nixies might be a good project to tackle.

Continue reading “Light Pipes and LEDs Team Up for a Modern Take on the Nixie Tube”

Practical Deep Learning

Deep Learning — the use of neural networks with modern techniques to tackle problems ranging from computer vision to speech recognition and synthesis — is certainly a current buzzword. However, at the core is a set of powerful methods for organizing self-learning systems. Multi-layer neural networks aren’t new, but there is a resurgence of interest primarily due to the availability of massively parallel computation platforms disguised as video cards.

The problem is getting started in something like this. There are plenty of scholarly papers that can be hard to wade through. Or you can grab some code from GitHub and try to puzzle it out.

A better idea would be to take a free class entitled: Practical Deep Learning for Coders, Part 1. The course is free unless you count your investment in time. They warn you to expect to commit about ten hours a week for seven weeks to complete the course. You can see the first installment in the video, below. Continue reading “Practical Deep Learning”