Well here we are, we’ve reached that time of year again at which our yearly ritual of resuscitating small internal combustion engines from their winter-induced morbidity is well under way. It’s lawn mowing season again, and a lot of us are spending our Saturday afternoons going up and down our little patches of grass courtesy of messers Briggs and Stratton. Where this is being written, the trusty Honda mower’s deck has unexpectedly failed, so an agricultural field topper is performing stand-in duty for a while, and leaving us with more of the rough shag pile of a steeplechaser’s course than the smooth velvet of a cricket ground. Tea on the lawn will be a mite springier this year.
When you think about it, there’s something ever so slightly odd about going to such effort over a patch of grass. Why do we do it? Because we like it? Because everyone else has one? Or simply because it’s less effort to fill the space with grass than it is to put something else there? It’s as if our little pockets of grassland have become one of those facets of our consumer culture that we never really think about, we just do. Continue reading “Something To Think About While You’re Mowing The Lawn”→
Ok, so you want a radio — but not just any radio. It has to be wireless, access a variety of music services, and must have a vintage aesthetic that belies its modern innards. Oh, and a tiny screen that displays album art, because that’s always awesome. This 1938 Emerson AX212-inspired radio delivers.
Building on the backbone of a Raspberry Pi Zero W and an Adafruit MAX 98357 mono amp chip, the crux of this single-speaker radio is the program Mopidy. Mopidy is a music player that enables streaming from multiple services, with the stipulation that you have a premium Spotify account. Once signed up, [Tinkernut] helpfully outlines how to set up Mopidy to run automatically once the Pi boots up. The addition of a screen to display album art adds flair to the design, and Adafruit’s 1.8″ TFT LCD screen is small enough to fit the bill.
Many of us develop things for one of two purposes: to hack something cool, or to sell something cool. When hacking something cool, your target market is yourself, and you already know you’ve made the sale. If your goal is to sell the thing you are making, then a lot more thought and effort is required. You could develop the coolest product in the world, but if your target market is too small, your price is too high, your lead time is too long, or any of a dozen other factors is not quite right, you’ll be spending a lot of time and effort on what will amount to a huge disappointment. The Hackaday Prize Best Product has many great examples which let us study some of these success factors, so let’s take a look. Continue reading “Will It Sell?”→
We don’t know why [stoppi71] needs to do gamma spectroscopy. We only know that he has made one, including a high-voltage power supply, a photomultiplier tube, and–what else–an Arduino. You also need a scintillation crystal to convert the gamma rays to visible light for the tube to pick up.
He started out using an open source multichannel analyzer (MCA) called Theremino. This connects through a sound card and runs on a PC. However, he wanted to roll his own and did so with some simple circuitry and an Arduino.
Last year we wrote about Hackerbot Labs’ autonomous boat, which project members hope to someday circumnavigate the globe. Now called Project Ladon, progress continues apace with a recent ocean test of their modified 18’ kayak, the TSV Disputed Right of Way. The kayak’s internal spaces contain a pair of lead-acid truck batteries controlled by a home-brewed control system that uses relays to control the craft’s trolling motor, with a Beaglebone and Arduino Mega under the hood.
The test was not exactly a success, with the boat actually avoiding the waypoints rather than sticking to them. Fortunately the team was aboard a chase boat so they were able to keep tabs on the craft. Unlike a quadcopter, which just falls down, a watercraft that borks may never be seen again.
Form Labs recently announced the launch of the Fuse 1, a desktop SLS printer that will print all your parts using nylon powder and a laser. This a fundamentally different method of 3D printing as compared to filament-based machines, and the best way to use a Fuse 1 is to fill the entire volume of the machine with 3D printed parts. [Michael Fogelman] decided to investigate the 3D packing problem, and managed to fill this printer with the maximum number of 3D printed tugboats. If you’re wondering, it’s 113, as compared with 82 tiny Benchies using naive bin packing.
The formal definition of this sort of problem is the bin packing problem, or simply calculating the maximum number of items can be packed into a finite volume. There is no general solution to this problem, and it’s probably impossible to create an algorithm that will solve this problem for any collection of 3D models. Nevertheless, it’s possible to create a solution that shows marked improvement over a naive solution.
[Michael]’s solution involves simulated annealing. This algorithm begins by randomly placing tugboats, then mutating the position or rotation of one of the boats for each iteration. The code is less than 1000 lines of Go and is available on GitHub if you already have an SLS printer at your disposal.
It should be noted this type of problem isn’t particularly new to the world of 3D printers. There have been a few tools to solve the bin-packing problem for filament-based printers, but the solutions to these problems are two-dimensional; since filling a bed is a problem that only uses the ‘shadow’ of the Z-axis of each part, it’s a slightly easier problem to solve.
Now that Form Labs’ Fuse 1 SLS printer has been announced, there is a new application for this type of problem in the space of 3D printers. It’s not a perfect solution — and it’s doubtful there will ever be a perfect solution — but if you’re looking for a way to fill the volume of your powder printer with parts, this is the best you’re going to do.
[Alonso Martinez] is an artist working on virtual characters at Pixar so it’s no wonder that his real life robots, Mira and Gertie, have personalities that make them seem like they jumped straight out of a Pixar movie. But what we really like are the tricks he’s used inside to bring them to life that are sure to get reused for the same or other things.
Mira’s head rotation mechanism
For example, Mira’s head can rotate in yaw, pitch and roll. To figure out how to make it do that he recalled having a joystick called the Microsoft Sidewinder Pro that had force feedback. That meant it might have had motors in line with the motions, much like what he wanted. To see how it worked, he bought one on eBay, took it apart, and improved on it to come up with his own design. But besides making use of the design in joysticks and heads, we can imagine it used to make robot eyeballs rotate in their sockets too. And as a side note, he’s running the robot off a Raspberry Pi, but notice the clever, space-saving way he’s mounted the whole mechanism to the Pi’s four mounting holes.
What also piqued our interest are the two tiny servos used in the head mechanism, two HD-DSM44 digital servos. These are even smaller than Tower Pro SG90s and with the added advantage of being metal geared.
Gertie’s delta jumping legs
To make the eyes blink he had to overcome the fact the head was a thin-walled sphere sliding over the body, and the eyes had to fit in the thin wall without contacting the body. His solution was to make them out of OLED screens with acrylic hemispheres for the protruding eyeballs. The circuit boards talk to the screens through ribbon cables that are around 32 connections per inch, which made for some careful soldering. And to further create a thin profile he even sanded the solder points flat.
His other robot, the yellow and green Gertie, jumps to move around and its internal mechanism is also a joy to examine. To swivel and hop, it uses much the same design as a delta 3D printer, with three legs that can move the upper body in any direction, and compress like a spring before leaping. We like how his method for determining the appropriate thickness of 3D printed PLA parts such that they wouldn’t break was simply trial an error, taking advantage of the rapid prototyping possible with 3D printers. He did cheat on one main part of each leg though, and that was to go with RC car tie rods for the lower half of each leg — but we won’t tell on him if you won’t.
And that’s only a small sample of the neat tips and tricks you’ll find in the video below (they start looking inside the robots at 7:35).
