How The Roland 808 Cowbell Worked

May 12, 2022 by Jenny List 9 Comments

Every generation has an instrument which defines its sound, and for those whose formative musical years lie in the 1980s, a very strong contender to the crown is the Roland TR-808 percussion synthesizer. Its sounds can be recognized across a slew of hits from that era and every decade since, and though the original instrument wasn’t a commercial success it remains accessible through sample packs, emulations, and clones. The 808 was an all-analogue device that didn’t use samples, thus [Mark Longstaff-Tyrrell] has been able to reproduce its distinctive cowbell sound with reference to some of the original circuitry.

It shouldn’t come as too much of a surprise to find that the circuit is refreshingly simple. The trigger pulse is converted into an envelope which controls a pair of oscillators. The mixed output passes through a bandpass filter to create the distinctive sound on the output which you can hear in the video below the break. The circuit is recreated on a breadboard with the only concession to modernity being a microcontroller taking the place of the Schmitt trigger oscillators in the original.

Altogether it provides a fascinating insight into the synthesis behind a classic sound, and gives us an increased appreciation for the design skills of those Roland engineers who created it. We’ve looked at the 808 before a few times, including an explanation of the famous faulty transistors which contributed to its sound.

Continue reading “How The Roland 808 Cowbell Worked” →

With Rocket Lab’s Daring Midair Catch, Reusable Rockets Go Mainstream

May 12, 2022 by Tom Nardi 34 Comments

We’ve all marveled at the videos of SpaceX rockets returning to their point of origin and landing on their spindly deployable legs, looking for all the world like something pulled from a 1950s science fiction film. On countless occasions founder Elon Musk and president Gwynne Shotwell have extolled the virtues of reusable rockets, such as lower operating cost and the higher reliability that comes with each booster having a flight heritage. At this point, even NASA feels confident enough to fly their missions and astronauts on reused SpaceX hardware.

Even so, SpaceX’s reusability program has remained an outlier, as all other launch providers have stayed the course and continue to offer only expendable booster rockets. Competitors such as United Launch Alliance and Blue Origin have teased varying degrees of reusability for their future vehicles, but to date have nothing to show for it beyond some flashy computer-generated imagery. All the while SpaceX continues to streamline their process, reducing turnaround time and refurbishment costs with each successful reuse of a Falcon 9 booster.

But that changed earlier this month, when a helicopter successfully caught one of Rocket Lab’s Electron boosters in midair as it fell back down to Earth under a parachute. While calling the two companies outright competitors might be a stretch given the relative sizes and capabilities of their boosters, SpaceX finally has a sparing partner when it comes to the science of reusability. The Falcon 9 has already smashed the Space Shuttle’s record turnaround time, but perhaps Rocket Lab will be the first to achieve Elon Musk’s stated goal of re-flying a rocket within 24 hours of its recovery.

Continue reading “With Rocket Lab’s Daring Midair Catch, Reusable Rockets Go Mainstream” →

Section from the ESP8266 datasheet, showing maximum input voltage as 3.6V, but not mentioning ESD diodes to VCC and only talking about a snap-back circuit set to 6V.

Is ESP8266 5 V Tolerant? This Curve Tracer Says Yes!

May 12, 2022 by Arya Voronova 44 Comments

Some people state that ESP8266 is tolerant of 5 V logic levels on its GPIOs, while others vehemently disagree, pointing at the datasheet-stated 3.6 V maximum. Datasheets aren’t source code for compiling the chip, however, and aren’t universally correct and complete either. [Avian] decided to dig deeper into the claims, conduct an experiment with an actual ESP8266 chip, then share the results for all of us.

For the experiment, he used a curve tracer – a device capable of producing a wide range of voltages and measuring the current being consumed, then plotting the voltage-to-current relationship. This helps characterize all sorts of variables, from diode breakdown voltages to transistor characteristics. The curve tracer he uses is a capable and professional-looking DIY build of his, and arguably, deserves a separate write-up!

The reasoning behind [Avian]’s experiment is simple – if the pin, set to an input, starts consuming a higher amount of current at a certain voltage threshold, then there’s gotta be some chip-internal structure, intended or unintended, that would be damaged at this voltage. Curve tracer in hand, he set up an ESP-01 module to set a GPIO to input, and started increasing the voltage.

The tests have shown that, while there’s a reverse biased ESD diode from GPIO pins to ground, there don’t seem to be diodes from the GPIO pin to the VCC rail – and those are the primary concern for 5 V tolerance. There does seem to be something functionally akin to a 6 V Zener diode internally, which should clamp the voltage before it gets too way high for the chip to handle. None of that should be a problem for 5 V compatibility, and it seems fair to interpret this as a confirmation of 5 V tolerance until someone shows otherwise.

[Avian] didn’t want to destroy an ESP8266, so the experiment was conducted with a 1 K series resistor between the curve tracer and the input – which might have biased the results a bit. On the other hand, adding series resistors in front of your inputs is an overall underappreciated practice, 5 V or otherwise. He also points out that, while the pins don’t seem to be adversely impacted by the higher input voltage, the bootloader might set some of them to 3.3 V outputs on boot-up, shorting your 5 V source to your 3.3 V rail — worth keeping in mind!

[Avian]’s research journeys are fun to follow, and we recommend you check his blog out; last time, we covered his research of an innocent-looking 3.5 mm jack hiding a devious audio compensation circuit. Since we first covered the ESP8266 in 2014, we’ve been researching all the things it’s really capable of, and we brought up the topic of GPIO 5 V compatibility way back in 2016 – it’s reassuring to finally put this question to rest!

We thank [Adrian] for sharing this with us!

Making A Concrete Pinhole Camera

May 12, 2022 by Lewin Day 9 Comments

A pinhole camera is a simple device that can be built out of virtually any simple closed chamber, and is a great way to learn about the basic principles of photography. [amuu] has created a version that can be readily made out of concrete, of all materials!

The photos captured by the camera featured some artifacts from light leaks and grit, but the results are enjoyable for their lo-fi, homebrew aesthetic!

The build starts with the creation of a mold for the concrete, using laminated sheets of foam. The foam is assembled with cut-up pieces of a ballpoint pen serving as cores in the mold. This provides a space for the film winders in the final product. The concrete is then mixed and poured into the mold, and allowed to set. Tapping or vibrating the mold is key to getting all the air bubbles out of the mixture.

Once set, the foam is mechanically removed from the concrete and the camera can be finished off. The internals are given a lick of black paint to improve the camera’s light-tightness. The shutter, pinhole, and film winder are then also fitted to allow the camera to function.

[amuu]’s first attempt to take photos with the camera lead to some results that were pleasingly lo-fi. There are overscan issues on the film and some other artifacts, but overall, the results are esoteric and fun. If you’re not a fan of the concrete camera, though, you can always consider making a 3D-printed pinhole camera instead!

Pitch Sequencer Turns Tascam Tape Deck Into Instrument

May 11, 2022 by Lewin Day 4 Comments

The cool thing about magnetic tape is that by varying the speed at which you play it back, you can vary the pitch of the output. [Issac] decided to take advantage of this, executing a fancy digitally-controlled pitch mod on his Tascam Porta 02 tape deck.

The build uses a Raspberry Pi Pico, which employs PWM to control the speed of the tape drive’s motor. This is achieved with the use of an NPN transistor driven by the PWM output of the Pico. This allows accurate control of motor speed, and thus pitch.

With that sorted out, the project was fleshed out with an OLED screen and a rotary encoder. These allow various patches or scripts to be run on the Pico, controlling the motor speed of the tape player in various ways. With a bit of work, [Issac] was also able to create a function that converted MIDI note values into PWM values that determine various motor speeds.

The natural thing to do next was to put in a tape with a looping sample at a set pitch, and then vary it in a sequence controlled by the Pico. The 8 steps of the sequence can be manually set with the rotary control, and in future, [Issac] even plans to add a real MIDI input, allowing the system to act as a monophonic synth.

If you prefer other routes to pitch shifting shenanigans, check out this project. Video after the break.

Continue reading “Pitch Sequencer Turns Tascam Tape Deck Into Instrument” →

Easy Network Config For IoT Devices With RGBeacon

May 11, 2022 by Lewin Day 25 Comments

When you’re hooking up hardware to a network, it can sometimes be a pain to figure out what IP address the device has ended up with. [Bas Pijls] often saw this problem occurring in the classroom, and set about creating a simple method for small devices to communicate their IP address and other data with a minimum of fuss.

[Bas] specifically wanted a way to do this without adding a display to the hardware, as this would add a lot of complexity and expense to simple IoT devices. Instead, RGBeacon was created, wherin a microcontroller flashes out network information with the aid of a single RGB WS2812B LED.

In fact, all three colors of the RGB LED are used to send information to a computer via a webcam. The red channel flashes out a clock signal, the green channel represents the beginning of a byte, and the blue channel flashes to indicate bits that are high. With a little signal processing, a computer running a Javascript app in a web browser can receive information from a microcontroller flashing its LEDs via a webcam.

It’s a neat hack that should make setting up devices in [Bas]’s classes much easier. It needn’t be limited to network info, either; the code could be repurposed to let a microcontroller flash out other messages, too. It’s not dissimilar from the old Timex Datalink watches which used monitor flashes to communicate!

When Sticks Fly

May 11, 2022 by Dan Maloney 26 Comments

When it comes to hobby rotorcraft, it almost seems like the more rotors, the better. Quadcopters, hexacopters, and octocopters we’ve seen, and there’s probably a dodecacopter buzzing around out there somewhere. But what about going the other way? What about a rotorcraft with the minimum complement of rotors?

And thus we have this unique “flying stick” bicopter. [Paweł Spychalski]’s creation reminds us a little of a miniature version of the “Flying Bedstead” that NASA used to train the Apollo LM pilots to touch down on the Moon, and which [Neil Armstrong] famously ejected from after getting the craft into some of the attitudes this little machine found itself in. The bicopter is unique thanks to its fuselage of carbon fiber tube, about a meter in length, each end of which holds a rotor. The rotors rotate counter to each other for torque control, and each is mounted to a servo-controlled gimbal for thrust vectoring. The control electronics and battery are strategically mounted on the tube to place the center of gravity just about equidistant between the rotors.

But is it flyable? Yes, but just barely. The video below shows that it certainly gets off the ground, but does a lot of bouncing as it tries to find a stable attitude. [Paweł] seems to think that the gimballing servos aren’t fast enough to make the thrust-vectoring adjustments needed to keep a stick flying, and we’d have to agree.

This isn’t [Paweł]’s first foray into bicopters; he earned “Fail of the Week” honors back in 2018 for his coaxial dualcopter. The flying stick seems to do much better in general, and kudos to him for even managing to get it off the ground.

Continue reading “When Sticks Fly” →