Pinning Tails On Satellites To Help Prevent Space Junk

September 13, 2021 by Tom Nardi 32 Comments

Low Earth orbit was already relatively crowded when only the big players were launching satellites, but as access to space has gotten cheaper, more and more pieces of hardware have started whizzing around overhead. SpaceX alone has launched nearly 1,800 individual satellites as part of its Starlink network since 2019, and could loft as many as 40,000 more in the coming decades. They aren’t alone, either. While their ambitions might not be nearly as grand, companies such as Amazon and Samsung have announced plans to create satellite “mega-constellations” of their own in the near future.

At least on paper, there’s plenty of room for everyone. But what about when things go wrong? Should a satellite fail and become unresponsive, it’s no longer able to maneuver its way out of close calls with other objects in orbit. This is an especially troubling scenario as not everything in orbit around the Earth has the ability to move itself in the first place. Should two of these uncontrollable objects find themselves on a collision course, there’s nothing we can do on the ground but watch and hope for the best. The resulting hypervelocity impact can send shrapnel and debris flying for hundreds or even thousands of kilometers in all three dimensions, creating an extremely hazardous situation for other vehicles.

One way to mitigate the problem is to design satellites in such a way that they will quickly reenter the Earth’s atmosphere and burn up at the end of their mission. Ideally, the deorbit procedure could even activate automatically if the vehicle became unresponsive or suffered some serious malfunction. Naturally, to foster as wide adoption as possible, such a system would have to be cheap, lightweight, simple to integrate into arbitrary spacecraft designs, and as reliable as possible. A tall order, to be sure.

But perhaps not an impossible one. Boeing subsidiary Millennium Space Systems recently announced it had successfully deployed a promising deorbiting device developed by Tethers Unlimited. Known as the Terminator Tape, the compact unit is designed to rapidly slow down an orbiting satellite by increasing the amount of drag it experiences in the wispy upper atmosphere.

Continue reading “Pinning Tails On Satellites To Help Prevent Space Junk” →

Scanning electron micrograph of a microfabricated lens array

Getting A Fly’s-Eye View With Microfabricated Lens Arrays

September 7, 2021 by Dan Maloney 11 Comments

Atomic force microscopy, laser ablation, and etching with a witches brew of toxic chemicals: sounds like [Zachary Tong] has been playing in the lab again, and this time he found a way to fabricate arrays of microscopic lenses as a result.

Like many of the best projects, [Zach]’s journey into micro-fabrication started with a happy accident. It happened while he was working on metal-activated chemical etching (MACE), which uses a noble metal catalyst to selectively carve high-aspect-ratio features in silicon. After blasting at a silver-coated silicon wafer with a laser, he noticed the ablation pits were very smooth and uniform after etching. This led him to several hypotheses about what was going on, all of which he was able to test.

The experiments themselves are pretty interesting, but what’s really cool is that [Zach] realized the smooth hemispherical pits in the silicon could act as a mold for an array of microscopic convex lenses. He was able to deposit a small amount of clear silicone resin into the mold by spin-coating, and (eventually) transfer the microlens array to a glass slide. The lenses are impressively small — hundreds of them over only a couple hundred square microns — and pretty well-formed. There’s always room for improvement, of course, but for an initial attempt based on a serendipitous finding, we’d call it a win. As for what good these lenses are, your guess is as good as ours. But novel processes like these tend to find a way to be useful, and the fact that this is coming out of a home lab doesn’t change that fact.

We find this kind of micro-fabrication fascinating. Whether it’s making OLED displays, micro-machining glass with plasma, or even rolling your own semiconductors, we can’t get enough of this stuff.

Continue reading “Getting A Fly’s-Eye View With Microfabricated Lens Arrays” →

Injection Molding Silicone Parts For Under $50

September 5, 2021 by Lewin Day 14 Comments

You’ve likely seen many tutorials on making silicone parts using 3D printed molds online. The vast majority of these methods use a simple pour method to fill the mold. This relies on careful degassing and gentle pouring to reduce the presence of bubbles in the final result. [Jan Mrázek] has been working on an alternative method however, that allows for injection molding at low cost in the home shop.

The process relies on the use of printed resin molds. [Jan] notes that this generally necessitates the use of condensation-cure silicones, as additive types don’t cure well in resin molds. The condensation silicone is mixed up, degassed, and poured into a standard cartridge. From there, it’s installed in a silicone delivery air gun, which uses compressed air to force the silicone out of the nozzle and into the waiting mold.

It’s basically using a bunch of home DIY gear to create a cheap injection molding solution for silicone parts. [Jan] notes that there are a few mods needed to mold design to suit the process, and that 400-800 kPa is a good pressure to inject the silicone at.

Having the silicone injected under pressure is great for complex mold designs, as it forces the material into all the little difficult nooks and crannies. Of course, we’ve seen other methods for making silicone parts before, too. Be sure to sound off in the comments with your own favored techniques for producing quality silicone parts. Video after the break.

Continue reading “Injection Molding Silicone Parts For Under $50” →

Impromptu Metal Detector Built From The Junk Bin

August 30, 2021 by Tom Nardi 12 Comments

Have you ever found yourself suddenly in need of finding a small metal object hidden in the woods? No? Well, neither have we. But we can’t say the same thing for [zaphod], who’s family was hoping to settle a dispute by finding the surveyor stakes that marked the corners of their property. It was a perfect job for a metal detector, but since they didn’t own one, a serviceable unit had to be assembled from literal garbage.

To start with, [zaphod] had to research how a metal detector actually works. After reviewing the pros and cons of various approaches, the decision was made to go with a beat frequency oscillator (BFO) circuit. It’s not the greatest design, it might even be the worst, but it could be built with the parts on hand and sometimes that’s all that matters. After packing a 2N3904 transistor, an LM386 amplifier, and every Hackaday reader’s favorite chip the 555 timer into an enclosure along with some of their closest friends, it was time to build the rest of the metal detector.

The sensor coil was made by salvaging the wire from an old fluorescent lamp ballast and winding it around the lid of a bucket 27 times. This was mounted to the end of a broom handle with some angle pieces made from PVC sheet material, being careful not to use any metal fasteners that would throw off the detector. With the handle of an old drill in the middle to hold onto, the metal detector was complete and actually looked the part.

So did [zaphod] save the day by finding the surveyor stakes and reconnoitering the family’s plot? Unfortunately, no. It wasn’t a technical failure though; the metal detector did appear to work, although it took a pretty sizable object to set it off. The real problem was that, after looking more closely into it, the surveyors only put down one stake unless they are specifically instructed otherwise. Since they already knew where that one was…

If your homemade metal detector can’t find something that was never there, did it really fail? Just a little something to meditate on. In any event, when even the cheapest smart bulb is packing a microcontroller powerful enough to emulate early home computers, we’re always happy to see somebody keep the old ways alive with a handful of ICs.

Nokia LCD Goes Transparent For Hands-Free Reminders

August 29, 2021 by Tom Nardi 1 Comment

These days everyone’s excited about transparent OLED panels, but where’s the love for the classic Nokia 5110 LCD? As the prolific [Nick Bild] demonstrates in his latest creation, all you’ve got to do is peel the backing off the the late 90s era display, and you’ve got yourself a see-through cyberpunk screen for a couple bucks.

In this case, [Nick] has attached the modified display to a pair of frames, and used an Adafruit QT Py microcontroller to connect it to the ESP32 powered ESP-EYE development board and OV2640 camera module. This lets him detect QR codes within the wearer’s field of vision and run a TensorFlow Lite neural network right on the hardware. Power is provided by a 2000 mAh LiPo battery running through an Adafruit PowerBoost 500.

The project, intended to provide augmented reality reminders for medical professionals, uses the QR codes to look up patient and medication information. Right now the neural network is being used to detect when the wearer has washed their hands, but obviously the training model could be switched out for something different as needed. By combining these information sources, the wearable can do things like warn the physician if a patient is allergic to the medication they’re currently looking at.

Relevant information and warnings are displayed on the Nokia LCD, which has been placed far enough away from the eye that the user can actually read the text; an important design consideration that [Zach Freedman] demonstrated with his (intentionally) illegible wearable display a few weeks back. That does make the design a bit…ungainly, but at least you don’t have to worry about hand-cutting your optics

Reverse Engineering A Topfield VFD Front Panel

August 14, 2021 by Tom Nardi 2 Comments

Hackers love the warm glow of a vacuum fluorescent display (VFD), and there’s no shortage of dead consumer electronics from which they can be pulled to keep our collective parts bins nicely stocked. Unfortunately, figuring out how to actually drive these salvaged modules can be tricky. But thanks to the efforts of [Lauri Pirttiaho], we now have a wealth of information about a VFD-equipped front panel used in several models of Topfield personal video recorders.

The board in question is powered by a Hynix HMS99C52S microcontroller and includes five buttons, a small four character 14-segment display, a larger eight character field, and an array of media-playback related icons. There’s also a real-time clock module onboard, as well as an IR receiver. [Lauri] tells us this same board is used in at least a half-dozen Topfield models, which should make it relatively easy to track one down.

After determining what goes where in the 6-pin connector that links the module with the recorder, a bit of poking with a logic analyzer revealed that they communicate over UART. With the commands decoded, [Lauri] was able to write a simple Python tool that lets you drive the front panel with nothing more exotic than a USB-to-serial adapter. Though keep in mind, you’ll need to provide 17 VDC on the appropriate pin of the connector to fire up the VFD.

What’s that? You don’t need the whole front panel, and just want to pull the VFD itself off the board? Not a problem. Our man [Lauri] was kind enough to document how data is passed from the Hynix microcontroller to the display itself; critical information should you want to liberate the screen from its PVR trappings.

If you manage to get your hands on one of these modules, it would be an ideal addition to a custom media streamer. Though we suppose simply turning it into a network-controlled clock would be a suitable alternative if you’re looking for something a bit easier.

Continue reading “Reverse Engineering A Topfield VFD Front Panel” →

So You Can Solder Small SMD Devices. The Question Is, Just How Small?

August 9, 2021 by Jenny List 15 Comments

A highlight of last year’s Hackaday Remoticon was a soldering competition that had teams from around the world came together online and did the well-known MakersBox SMD Challenge kit in which a series of LED circuits of decreasing size must be soldered. The Hackaday crew acquitted themselves well, and though an 01005 resistor and LED certainly pushes a writer’s soldering skills to the limit it’s very satisfying to see it working. Lest that kit become too easy, [Arthur Benemann] has come up with something even more fiendish; his uSMD is a 555 LED flasher that uses a BGA 555 and a selection of 008004 small components.

The trick with an 01005 is to heat not the tinned and fluxed solder joint, but the trace leading up to it. If components of that size can be mastered then perhaps an 008004 isn’t that much smaller so maybe the same technique might work for them too. In his tip email to us he wrote “Soldering 008004 isn’t much worse than a 0201, you just need magnification“, and while we think he might be trolling us slightly we can see there’s no reason why it shouldn’t be do-able. Sadly he doesn’t seem to have made it available for us to buy and try so if you want to prove yourself with a soldering iron you’ll have to source the PCBs and parts yourself. Still, we suspect that if you are the type of person who can solder an 008004 then that will hardly be an onerous task for you.

Meanwhile this isn’t the first soldering challenge kit we’ve brought you, and of course if you’d like to hone your skills you can find the MakersBox one on Tindie.