There are two things going on here. First, a home-made frequency downconverter shifts the 147 MHz signal down to the 1 MHz neighborhood where the AM radio can deal with it. Then, the AM radio is tuned just slightly off the right frequency and the FM signal is slope detected.
The downconverter consists of a local tuned oscillator and a mixer. The local oscillator generates an approximate 146 MHz signal from an 18 MHz crystal, accounting for two of the three transistors. Then this 146 MHz signal and the approximately 147 MHz signal that he wants to listen to are multiplied together (mixed) using the third transistor.
If you’re not up on your radio theory, a frequency mixer takes in two signals at different frequencies and produces an output signal that has various sums and differences of the two input signals in it. It’s this 147 MHz – 146 MHz = 1 MHz FM signal, right in the middle of the AM radio band’s frequency range, that’s passed on to the AM radio.
Next, the AM radio slope detects the frequency-modulated (FM) signal as if it were amplitude modulated (AM). This works as follows: FM radio encodes audio as changes in frequency, while AM radios encode the audio signal in the amplitude, or volume, of the radio signal. Instead of tracking the changing frequency as an FM radio would, slope detectors stick on a single frequency that’s tuned just slightly off from the FM carrier frequency. As the FM signal gets closer to or farther away from this fixed frequency, the received signal gets louder or quieter, and FM is detected as AM.
At 5:23, [vk3ye] steps through the circuit diagram. As he mentions, these are old tricks from circa 50 years ago, but it’s very nice to see a junk-box hack working so well with so few parts and receiving (very) high frequency FM on an old AM car radio. A circuit like this could make a versatile front end for an SDR setup. It makes us want to warm up the soldering iron.
It isn’t much trouble programming one of TI’s MSP430 chips, but outside of the official Flash Emulation Tool, TI doesn’t make programming one of these microcontrollers cheap. The most common way of programming an MSP430 is using a Launchpad Dev board, and [Vicente] has the best looking one yet.
The MSP430 series of chips can be programmed through JTAG or Spy-By-Wire, and the official, professional engineering tool from TI for these chips costs about $100. Those of us with more sense than money have another option – use one of the TI Launchpad dev boards as an MSP430 programmer.
[Vicente]’s project uses the MSP430G2 Launchpad, with just a few wires going to the proper connector found in the official programmer from TI. There are a few limitations; the programmer only works at 3.6V, so programming 1.8V devices might not be a good idea. Also, it only works with Spy-By-Wire and no JTAG support is available. Still, it’s a great looking project, and does exactly what it’s designed to.
The setup can take IR photographs of sleeping chickens and notify [Hugo] when it’s time to collect the eggs. Naturally, an abundance of other sensors are available. The coop can tweet based on ambient temperature, nest temperature, light level, motion sensor status, or the amount of remaining chicken feed. You can easily follow whether the two fowl are in the coop or out in the yard. It’s like Big Brother, only for birds.
The application is, frankly, ridiculous. But if you’re into home (or coop) automation, there’s a lot to be learned and the project is very well documented. [Hugo] used OpenCV for visual egg detection, and custom Python code to slightly randomize the tweets’ text. All of these details are up on his Github account.
A group of multicopter enthusiasts from Argonay, France cordoned off a path through the forest and spent the day racing. The resulting video makes it look like a heck of a good time.
Twenty “drone” pilots all used first-person view (FPV) camera setups for complete immersion, racing at up to 50 kilometers per hour through a 150m course in the woods that was chosen for maximum thrills and spills. The track basically followed a footpath, but the pilots still had to be extremely alert to avoid natural obstacles (we call them “trees”). The narrator adds that the nearly random lighting and camera artifacts added an extra level of difficulty to the event.
After practicing a few times just to get around the track in one piece, they started racing each other in heats. On the final heat, at 3:40 in the video, five copters start off head-to-head and tear out into the woods. Of them, only two cross the finish line.
FPV drone crash scenes still make us wince a little bit. We wonder how many of the participants spent the next few nights in the repair bay.
CNC Foam Cutters are capable of cutting out some pretty cool shapes that would otherwise be extremely difficult to do. They do this by pulling a heated metal wire though a block of foam. Electrical current passing through the wire heats it up causing the foam to melt away, there is no dust and no mess to clean up. [batchelc] decided to make his own large-scale CNC Foam Cutter and took a lot of photos along the way.
Since machine is 4 axis, meaning both sides can move forward/back and up/down independently of each other, tapered shapes are possible. One example where this would be helpful is cutting wings that are swept or have different profiles at each end.
The main frame of the machine is made from Unistrut and measures a whopping 60 by 60 inches. Subtract the size of the mechanical components and the cutting area ends up being 48 by 42 and 22 inches high. The foam sits on an MDF bed, gravity is the only method of holding the foam down during cutting. The wire doesn’t actually touch the foam so there is no force applied to cause it to move. The hot wire moves slowly and melts the foam just a few thousands of an inch in front of the wire resulting in no contact between the two.
Both axes on each side are driven by 1/2-10″ lead screws supported by bearing blocks on both sides. The longitudinal axes smoothly traverse the length of the machine by means of skate bearings that ride on the Unistrut channel itself. The vertical axes have a plastic bushing that slides along a round shaft.
The control portion of the machine is a HobbyCC FoamPro kit that came with the 4 axis stepper motor control board and 4 NEMA 23 stepper motors. GMFC software is used to both generate the g-code and send the commands to the stepper motor control board.
[Ryan] a.k.a. [1o57] comes from an age before anyone could ask a question, pull out their smartphone, and instantly receive an answer from the great Google mind. He thinks there’s something we have lost with our new portable cybernetic brains – the opportunity to ask a question, think about it, review what we already know, and reason out a solution. There’s a lot to be said about solving a problem all by yourself, and there’s nothing to compare to the ‘ah-ha’ moment that comes with it.
[1o57] started his Mystery Challenges at DEFCON purely by accident; he had won the TCP/IP embedded device competition one year, and the next year was looking to claim his title again. The head of the TCP/IP embedded competition had resigned from his role, and through a few emails, [1o57] took on the role himself. There was a miscommunication, though, and [1o57] was scheduled to run the TCP/IP drinking competition. This eventually morphed into a not-totally-official ‘Mystery Challenge’ that caught fire in email threads and IRC channels. Everyone wanted to beat the mystery challenge, and it was up to [1o57] to pull something out of his bag of tricks.
The first Mystery Challenge was a mechanical device with three locks ready to be picked (one was already unlocked), magnets to grab ferrous picks, and only slightly bomb-like in appearance. The next few years featured similar devices with more locks, better puzzles, and were heavy enough to make a few security officials believe [1o57] was going to blow up the Hoover dam.
With a few years of practice, [1o57] is turning crypto puzzles into an art. His DEFCON 22 badge had different lanyards that needed to be arranged to spell out a code. To solve the puzzle, you’ll need to talk to other people, a great way to meet one of [1o57]’s goals of getting all the natural introverts working together.
Oh. This talk has its own crypto challenge, something [1o57] just can’t get out of his blood:
So far nobody has solved the @hackaday 10 year anniversary in-talk-mini-crypto-puzzle-of-doom…("it's only a model")
[Gadget Addict] found out about a contest being held by a shoe seller. Their mobile app has a game very much like Bejeweled. The high scorer each month gets £500. His choices were to be better at the game than everyone else, or to be smarter. He chose the latter by writing a computer vision program to play the game.
There are two distinct parts of a hack like this one. The first is just figuring out a way to programmatically detect the game board and correctly identify each icon on it. This is an iPad game. [Gadget Addict] is mirroring the screen on his laptop, which gives him easy access to the game board and also allows for simulated swipes for automatic play. Above you can see two examples where black pixels may be counted in order to identify the icon. A set of secondary checks differentiates similar entries after the first filtering. The other part of the hack involves writing the algorithms to solve for the best move.
If you liked this one, check out a super-fast Bejeweled solver from several years back. We should also mention that this was just a proof of concept and [GA] never actually entered the contest.