If you’ve ever wondered what lies beneath the barely visible hole in the can of a MEMS microphone, you’re in luck, because [Zach Tong] has a $10 pair of earbuds to sacrifice for the cause and an electron microscope.
For the uninitiated, MEMS stands for microelectromechanical systems, the tiny silicon machines that power some of the more miraculous functions of smartphones and other modern electronics. The most familiar MEMS device might be the accelerometer that gives your phone a sense of where it is in space; [Zach] has a deep dive into MEMS accelerometers that we covered a while back.
MEMS microphones seem a little bit easier to understand mechanically, since all they have to do is change vibrations in air into an electrical signal. The microphone that [Zach] tore down for this video is ridiculously small; the SMD device is only about 3 mm long, with the MEMS chip under the can a fraction of a millimeter on a side. After some overall views with the optical microscope, [Zach] opened the can and put the guts under his scanning electron microscope. The SEM shots are pretty amazing, revealing a dimpled silicon diaphragm over a second layer with holes etched right through it. The dimples on the diaphragm nest into the holes, forming an air-dielectric capacitor whose capacitance varies as sound waves vibrate the diaphragm.
The most visually interesting feature, though, might be the deep cavity lying behind the two upper surfaces. The cavity, which [Zach] says bears evidence of having been etched by the deep reactive ion etching method, has cool-looking corrugations in its walls. The enormity of the cavity relative to the thin layers covering it suggests it’s a resonating cavity for the sound waves.
Thanks to [Zach] for this in-depth look at a device that’s amazingly complex yet remarkably simple.
Continue reading “Up Close And Personal With A MEMS Microphone”
We’ve been eagerly following [ProjectsInFlight]’s stepwise journey toward DIY semiconductors, including all the ups and downs, false leads, and tedious optimizations needed to make it possible for the average hacker to make chips with readily available tools and materials.
Next up is metallization, and spoiler alert: it wasn’t easy. In a real fab, metal layers are added to chips using some form of deposition or sputtering method, each of which needs some expensive vacuum equipment. [ProjectsInFlight] wanted a more approachable way to lay down thin films of metal, so he turned to an old friend: the silver mirror reaction. You may have seen this demonstrated in high school chemistry; a preparation of Tollen’s reagent, a mix of sodium hydroxide, ammonia, and silver nitrate, is mixed with glucose in a glass vessel. The glucose reduces the reagent, leaving the metallic silver to precipitate on the inside of the glass, which creates a beautiful silvered effect.
Despite some issues, the silvering method worked well enough on chips to proceed on, albeit carefully, since the layer is easily scratched off. [ProjectsInFlight]’s next step was to find an etchant for silver, a tall order for a noble metal. He explored piranha solutions, which are acids spiked with peroxide, and eventually settled on plain old white vinegar with a dash of 12% peroxide. Despite that success, the silver layer was having trouble sticking to the chip, much preferring to stay with the photoresist when the protective film was removed.
The solution was to replace the photoresist’s protective film with Teflon thread-sealing tape. That allowed the whole process from plating to etching to work, resulting in conductive traces with pretty fine resolution. Sure they’re a bit delicate, but that’s something to address another day. He’s come a long way from his DIY tube furnace used to put down oxide layers, and suffering through the search for oxide etchants and exploring photolithography methods. It’s been a fun ride so far, and we’re eager to see what’s next.
Continue reading “Simple Chemistry To Metallize And Etch Silicon Chips”
Building circuits on a silicon chip is a bit like a game of Tetris — you have to lay down layer after layer of different materials while lining up holes in the existing layers with blocks of the correct shape on new layers. Of course, Tetris generally doesn’t require you to use insanely high temperatures and spectacularly toxic chemicals to play. Or maybe it does; we haven’t played the game in a while, so they might have nerfed things.
Luckily, [ProjectsInFlight] doesn’t treat his efforts to build semiconductors at home like a game — in fact, the first half of his video on etching oxide layers on silicon chips is devoted to the dangers of hydrofluoric acid. As it turns out, despite the fact that HF can dissolve your skin, sear your lungs, and stop your heart, as long as you use a dilute solution of the stuff and take proper precautions, you should be pretty safe around it. This makes sense, since HF is present in small amounts in all manner of consumer products, many of which are methodically tested in search of a practical way to remove oxides from silicon, which [ProjectsInFlight] has spent so much effort recently to learn how to deposit. But such is the ironic lot of a chip maker.
Three products were tested — rust remover, glass etching cream, and a dental porcelain etching gel — against a 300 nm silicon dioxide layer. Etch speed varied widely, from rust remover’s 10 nm/min to glass etching cream’s blazing 240 nm/min — we wonder if that could be moderated by thinning the cream out with a bit of water. Each solution had pros and cons; the liquid rust remover was cheap easy to handle and clean up, while the dental etching gel was extremely easy to deposit but pretty expensive.
The good news was that everything worked, and each performed differently enough that [ProjectsInFlight] now has a range of tools to choose from. We’re looking forward to seeing what’s next — looks like it’ll be masking techniques.
Continue reading “Testing Oxide Etchants For The Home Semiconductor Fab”
Most of us are content to get our semiconductors from the usual sources, happily abstracting away the complexity locked within those little epoxy blobs. But eventually, you might get the itch to roll your own semiconductors, in which case you’ll need to start gearing up. And one of the first tools you’ll need is likely to be something like this DIY tube furnace.
For the uninitiated, [ProjectsInFlight] helpfully explains in the video below just what a tube furnace is and why you’d need one to start working with semiconductors. Perhaps unsurprisingly, a tube furnace is just a tube that gets really, really hot — like 1,200° C. In addition to the extreme heat, commercial furnaces are often set up to seal off the ends of the tube to create specific conditions within, such as an inert gas atmosphere or even a vacuum. The combination of heat and atmospheric control allows the budding fabricator to transform silicon wafers using chemical and physical processes.
[ProjectsInFlight]’s tube furnace started with a length of heat-resistant quartz glass tubing and a small tub of sodium silicate refractory cement, from the plumbing section of any home store. The tube was given a thin coat of cement and dried in a low oven before wrapping it with nichrome wire. The wrapped tube got another, thicker layer of silicate cement and an insulating wrap of alumina ceramic wool before applying power to cure everything at 1,000° C. The cured tube then went into a custom-built sheet steel enclosure with plenty of extra insulation, along with an Arduino and a solid-state relay to control the furnace. The video below concludes with testing the furnace by growing a silicon dioxide coating on a scrap of silicon wafer. This was helped along by the injection of a few whisps of water vapor while ramping the furnace temperature up, and the results are easily visible.
[ProjectsInFlight] still needs to add seals to the tube to control the atmosphere in there, an upgrade we’ll be on the lookout for. It’s already a great start, although it might take a while to catch up to our friend [Sam Zeloof].
Continue reading “Start Your Semiconductor Fab With This DIY Tube Furnace”
The first solar panels may have rolled out of Bell Labs in the 1950s, with major press around their inconsistent and patchy adoption in the decades that followed, but despite the fanfare they were not been able to compete on a price per kilowatt compared to other methods of power generation until much more recently. Since then the amount of solar farms has increased exponentially, and while generating energy from the sun is much cleaner than most other methods of energy production and contributes no greenhouse gasses in the process there are some concerns with disposal of solar panels as they reach the end of their 30-year lifespan. Some companies are planning on making money on recycling these old modules rather than letting them be landfilled. Continue reading “There’s Cash In Them Old Solar Panels”
Few CPUs have had the long-lasting influence that the 8086 did. It is hard to believe that when your modern desktop computer boots, it probably thinks it is an 8086 from 1978 until some software gooses it into a more modern state. When [Ken] was examining an 8086 die, however, he noticed that part of the die didn’t look like the rest. Turns out, Intel had a bug in the original version of the 8086. In those days you couldn’t patch the microcode. It was more like a PC board — you had to change the layout and make a new one to fix it.
The affected area is the Group Decode ROM. The area is responsible for categorizing instructions based on the type of decoding they require. While it is marked as a ROM, it is more of a programmable logic array. The bug was pretty intense. If an interrupt followed either a MOV SS or POP SS instruction, havoc ensues.
Continue reading “Silicon Sleuthing: Finding A Ancient Bugfix On The 8086”
Since the widespread adoption of USB 1.1 in the 90s, USB has become the de facto standard for connecting most peripherals to our everyday computers. The latest revision of the technology has been USB 4, which pushes the data rate capabilities to 40 Gbit/s. This amount of throughput is mindblowing compared to the USB 1.x speeds which were three to four orders of magnitude slower in comparison. But data speeds haven’t been the only thing changing with the USB specifications. The amount of power handling they can do has increased by orders of magnitude as well, as this DIY USB charger demonstrates by delivering around 200 W to multiple devices at once.
The build comes to us from [tobychui] who not only needed USB rapid charging for his devices while on-the-go but also wanted to build the rapid charger himself and for the charger to come in a small form factor while still using silicon components instead of more modern gallium nitride solutions. The solution he came up with was to use a 24 V DC power supply coupled with two regulator modules meant for solar panel installations to deliver a staggering amount of power to several devices at once. The charger is still relatively small, and cost around $30 US dollars to make.
Part of what makes builds like this possible is the USB Power Delivery (PD) standard, which has enabled all kinds of electronics to switch to USB for their power needs rather than getting their power from dedicated, proprietary, and/or low-quality power bricks or wall warts. In fact, you can even use this technology to do things like charge lithium batteries.
Continue reading “DIY USB Charging The Right Way”