The core of the build is a 16 bit microcontroller a dsPIC33FJ128GP802 from Microchip. It’s a humble chip to be doing so much. It uses a UBlox NEO-6M positioning module for the location and a custom built QFH antenna built after calculations done with an online calculator for the GPS half. The audio half is based around a VLSI VS1003b decoder chip.
The whole build is done with protoboard. Where the built in traces didn’t suffice enamel and wire wrap wire were carefully routed and soldered in place. There’s a 48pin LQFP package chip soldered dead bug style that’s impressive to behold. You can see some good pictures in this small gallery below.
A bit ago I wrote an article called, “Death To The 3.5mm Audio Jack, Long Live Wireless.” A few readers were with me, a few were indifferent, many were vehemently against me, and there was a, not insubstantial, subset in a pure panic about the potential retirement of a beloved connector. Now I used a lot of opinionated language dispersed with subjectively evaluated facts to make a case that the connector is out. Not today maybe, but there is certainly a tomorrow not so far off where there are more wireless headsets at the electronics store than wired ones.
The bumper design is particularly interesting. The magic happens with two rings of conductive filament. the bottom one is stationary while the top one is a multi material print with a flexible filament. When the ball runs into the bumper the top filament flexes and the lower rings contact. Awesome. Who wants to copy this over to a joystick or bump sensor for a robot first? Send us a tip!
The whole document can be read as a primer on pinball design. [Tony] starts by describing the history of pinball from the French courts to the modern day. He then works up from the play styles, rules, and common elements to the rationale for his design. It’s fascinating.
Then his guide gets to the technical details. The whole machine was designed in OpenSCAD. It took over 8.5 km of eighty different filaments fed through 1200+ hours of 3D printing time (not including failed prints) to complete. The electronics were hand laid out in a notebook, based around custom boards, parts, and two Arduinos that handle all the solenoids, scoring, and actuators. The theme is based around a favorite bowling alley and other landmarks.
It’s a labor of love for sure, and an inspiring build. You can catch a video of it in operation after the break.
We got quite a few tips in about a paper from Vanderbilt about a cool scrap metal battery they’ve been playing with. They made some pretty bold claims and when we fed the numbers in they pretty much say they’ve got a battery you can make at home, that can hold half as much as a lead acid, can be made out of scraps in a cave (even if you’re not Tony Stark), charge super fast,and can cycle 5,000 times without appreciable capacity loss. Needless to say that’s super cool.
Of course, science research is as broken as ever and the paper was hidden behind a paywall. Through mysterious powers such as the library and bothering people we were able to get past this cunning defense and read the paper. Unfortunately the paper reads more like a brag track than a useful experimental guide on how to build the dang battery. It’s also possible that our copy was missing some pages. Anyway, we want to do science!
Anyway, here’s what we know. The battery is based on an ancient battery called the Baghdad Battery. The ancient battery supposedly used iron and copper with a mystery electrolyte. The scrap battery, however, is made from scrap iron and scrap brass. The iron makes sense, but why brass? Well, brass has copper in it, and you can still get at it chemically even if it’s alloyed.
To that end, the next step was to throw some oxygen atoms in with those pesky Fe and Cu ones. The goal is to get a redox reaction going. If you do it right you can achieve pseudocapacitance. To to this the researchers used “common household chemicals and voltages” to anodize the iron and copper inside the brass. The press photo have them holding a gallon of muratic acid, if that helps. We don’t know, but if they can jam a few oxygen atoms in there then so can we!
After that it’s all about sitting the electrodes in a bath of potassium hydroxide. We guess you can scrape the inside of an AA for that. Anyway, the paper’s light on process but the battery seems really cool. They’re not pursuing this research for commercialization, instead going the OSHW route. They hope to get to the point where anyone can just grind up a bunch of scrap steel and brass, maybe throw it in a birdcage, anodize it, and get a super long life battery for grid use for less than a lead acid. If any of you manage to build one of these drop us a tip!
Raspberry Pi’s are a little weird. They mostly get crammed into the slots microcontrollers used to live in. The nice part about microcontrollers is that they just turn on and start going. There’s no OS to boot. No file system to mount. Of course the downside to microcontrollers is often that there’s no OS to boot and file system to mount. Regardless, mostly you’ve got to spend a bit configuring a Raspbian install before a Raspberry Pi really starts to encroach on the microcontroller’s territory.
Pi Bakery abstracts all this. You can drag blocks, representing scripts, in the order you’d like them run. If you want to your Pi to boot up, connect to WiFi, and then start a VNC server it’s as easy a dragging the blocks in the right order and filling in the blanks. You can see an example of it in operation in the video after the break.
One only has to ship one or two things via a container, receiving them strangely damaged on the other end, before you start to wonder about your shipper. Did they open this box and sort of stomp around a bit? Did I perhaps accidentally contract a submarine instead of a boat? Did they take a detour past the sun? How could this possibly have melted?
[Jesus Echavarria]‘s friend had similar fears and suspicions about a box he is going to have shipped from Spain to China. So [Jesus] got to work and built this nice datalogger to discover the truth. Since the logger might have to go for a couple of months, it’s an exercise in low power design.
The core of the build is a humble PIC18. Its job is to take the information from an ambient light, temperature, and humidity sensor suite and dump it all to an SD card. Aside from the RTC, this is all powered from a generic LiPo power cell. The first iteration can run for 10 days on one charge, and that’s without any of the low power features of the microcontroller enabled. It should be able to go for much longer once it can put itself to sleep for a period.
It’s all housed in a 3D printed case with some magnets to stick it to shell of the shipping container. Considering the surprisingly astronomical price of commercial dataloggers, it’s a nice build!
Most of us have had a science teacher desperately try to alleviate the drudgery of standardized test centric science education by dramatically putting a copper nail and a zinc nail into a potato or lemon. “Behold, we can measure a voltage with this voltmeter. If you get asked what a voltmeter is on a test, here is a definition none of you have enough experimental basis to understand,” the teacher would say as their dreams of being a true educator were crushed a little more.