There was a time when a microphone for most people was a cheap plastic affair that probably came for free with their sound card, but in the age of pandemic video streaming no desktop is complete without a chunky model that looks for all the world as though it escaped from a studio. Few people make their own microphones, so the work of [DJJules] in building very high quality condenser microphones is a particularly fascinating read.
A condenser microphone is a capacitor in which one plate is formed by a conductive diaphragm. A bias voltage is supplied to the diaphragm via a resistor, and since the charge on the plate remains constant as its capacitance changes with the sound vibrations, the voltage on the capacitor changes accordingly. This is picked up by a high impedance buffer and from there fed to a normal microphone input. This Instructable uses a commercial condenser microphone capsule, and takes the reader through generating the bias voltage for it before describing the op-amp buffer circuit.
The most interesting part comes at the end, as we’re shown how the sensitivity pattern of a dual-microphone array can be tuned to be omnidirectional, cardoid, or figure-of-eight. This is probably the norm among audio engineers, but we rarely see this sort of insight in our community. We may never build a microphone of our own, but it’s fascinating to see this one from the ground up in the video below the break.
Most people find two problems when it comes to flip-dot displays: where to buy them and how to drive them. If you’re [Pierre Muth] you level up and add the challenge of driving them fast enough to rival non-mechanical displays like LCDs. It was a success, resulting in a novel and fast way of controlling flip-dot displays.
If you’re lucky, you can get a used flip-dot panel decommissioned from an old bus destination panel, or perhaps the arrivals/departures board at a train station. But it is possible to buy brand new 1×7 pixel strips which is what [Pierre] has done. These come without any kind of driving hardware; just the magnetized dots with coils that can be energized to change the state.
The problem comes in needing to reverse the polarity of the coil to achieve both set and unset states. Here [Pierre] has a very interesting idea: instead of working out a way to change the connections of the coils between source and sink, he’s using a capacitor on one side that can be driven high or low to flip the dot.
Using this technique, charging the capacitor will give enough kick to flip the dot on the display. The same will happen when discharged (flipping the dot back), with the added benefit of not using additional power since the capacitor is already charged from setting the pixel. A circuit board was designed with CMOS to control each capacitor. A PCB is mounted to the back of a 7-pixel strip, creating modules that are formed into a larger display using SPI to cascade data from one to the next. The result, as you can see after the break, does a fantastic job of playing Bad Apple on the 24×14 matrix. If you have visions of one of these on your own desk, the design files and source code are available. Buying the pixels for a display this size is surprisingly affordable at about 100 €.
We’re a bit jealous of all the fun displays [Pierre] has been working on. He previously built a 384 neon bulb display that he was showing off last Autumn.
[K6ARK] likes to operate portable, so he puts together very lightweight antennas. One of his latest uses tiny toroids and SMD capacitors to form trap elements. You can see the construction of it in the video below.
You usually think of toroid winding as something you do when building transmitters or receivers, especially small ones like these. We presume the antenna is best for QRP (low power) operation since the tiny core would saturate pretty quickly at higher power. Exactly how much power you should pass through an FT50-43 core depends on the exact application, but we’ve seen numbers around 5 watts.
From the sound of reports in the press, graphene is the miracle material that will cure all the world’s ills. It’ll make batteries better, supercharge solar panels, and revolutionize medicine. While a lot of applications for the carbon monolayer are actually out in the market already, there’s still a long way to go before the stuff is in everything, partly because graphene can be very difficult to make.
It doesn’t necessarily have to be so hard, though, as [Zachary Tong] shows us with his laser-induced graphene supercapacitors. His production method couldn’t be simpler, and chances are good you’ve got everything you need to replicate the method in your shop right now. All it takes is a 405-nm laser, a 3D-printer or CNC router, and a roll of Kapton tape. As [Zach] explains, the laser energy converts the polyimide film used as the base material of Kapton into a sort of graphene foam. This foam doesn’t have all the usual properties of monolayer graphene, but it has interesting properties of its own, like extremely high surface area and moderate conductivity.
To make his supercaps, [Zach] stuck some Kapton tape to glass slides and etched a pattern into with the laser. His pattern has closely spaced interdigitated electrodes, which when covered with a weak sulfuric acid electrolyte shows remarkably high capacitance. He played with different patterns and configurations, including stacking tape up into layers, and came up with some pretty big capacitors. As a side project, he used the same method to produce a remarkable effective Kapton-tape heating element, which could have tons of applications.
Here’s hoping that [Zach]’s quick and easy graphene method inspires further experimentation. To get you started, check out our deep-dive into Kapton and how not every miracle material lives up to its promise.
One of the biggest challenges for wireless sensor networks is that of power. Solar panels usually produce less power than you hoped, especially small ones, and designing super low power circuits is tricky. [Strange.rand] has dropped into the low-power rabbit hole, and is designing a low-cost wireless sensor node that runs on solar power and a supercapacitor.
The main components of the sensor node is an ATMega 328P microcontroller running at 4Mhz, RFM69 radio transceiver, I2C temperature/humidity sensor, 1F supercapacitor, and a small solar panel. The radio, MCU, and sensor all run on 1.5-3.6V, but the supercap and solar panel combination can go up to 5.5V. To regulate the power to lower voltage components a low-drop voltage regulator might seem like the simplest solution, but [strange.rand] found that the 3.3V regulator was consuming an additional 20uA or more when the voltage dropped below 3.3V. Instead, he opted to eliminate the LDO, and limit the charging voltage of the capacitor to 3.6V with a comparator-based overvoltage protection circuit. Using this configuration, the circuit was able to run for 42 hours on a single charge, transmitting data once per minute while above 2.7V, and once every three minutes below that.
Another challenge was undervoltage protection. [strange.rand] discovered that the ATmega consumes an undocumented 3-5 mA when it goes into brown-out below 1.8V. The small solar panel only produces 1 mA, so the MCU would prevent the supercapacitor from charging again. He solved this with another comparator circuit to cut power to the other components.
Battery technology is the talk of the town right now, as it’s the main bottleneck holding up progress on many facets of renewable energy. There are other technologies available for energy storage, though, and while they might seem like drop-in replacements for batteries they can have some peculiar behaviors. Supercapacitors, for example, have a completely different set of requirements for charging compared to batteries, and behave in peculiar ways compared to batteries.
This project from [sciencedude1990] shows off some of the quirks of supercapacitors by showing one method of rapidly charging one. One of the most critical differences between batteries and supercapacitors is that supercapacitors’ charge state can be easily related to voltage, and they will discharge effectively all the way to zero volts without damage. This behavior has to be accounted for in the charging circuit. The charging circuit here uses an ATtiny13A and a MP18021 half-bridge gate driver to charge the capacitor, and also is programmed in a way that allows for three steps for charging the capacitor. This helps mitigate the its peculiar behavior compared to a battery, and also allows the 450 farad capacitor to charge from 0.7V to 2.8V in about three minutes.
You might think the smell of an electrolytic capacitor boiling out is bad, but if scientists from the University of Sydney have their way, that might be nothing. They’ve devised an ultracapacitor — that uses biomass from the stinky durian fruit along with jackfruit. We assume the capacitors don’t stink in normal use, but we wouldn’t want to overload one and let the smoke out.
One of the things we found interesting about this is that the process seemed like something you might be able to reproduce in a garage. Sure, there were a few exotic steps like using a vacuum oven and a furnace with nitrogen, and you’d need some ability to handle chemicals like vinylidene fluoride. However, the hacker community has found ways to create lots of things with common tools, and we would imagine creating aerogels from some fruit ought not be out of reach.