shinyhappyrainbows

The actuator that I've purchased for building a single stick prototype of Wrist Bot has 16mm travel. So, working on rough lever lengths, here are images of how the wrist flick a human drummer uses actually works. The first image is stick at rest, the second image is stick flicked, showing the stick at point of contact with the drum skin. By pivoting the stick around its balance point, the stick rotates beyond the angle the actuator pulls its cradle, hits the skin and rebounds back to the cradle. By holding the actuator in the pulled position, the stick can bounce after rebound to come back for a second hit, potentially allowing for "crushed rolls" and other rebound effects a real drummer plays. Quickly snapping the actuator back to rest after the drum is hit prevents double taps. Image 1: stick at rest Image 2: stick flicked to skin contact point Note the increased angle of the stick at full contact, after being flicked. The dual-pivot approch, while trickier to set

3D printers are an amazing tool and, the thing that tantalises me the most, is the potential to 3D print a 3D printer that can print its own "sister," that can print its sister, that can print its own sister... The key to being able to print a printer that can print exact copies of itsel is only in the design and materials choice, really. These machines are only a metal printing advance away. Or are they already here? Take a look at this proof of concept. Exploded view of a concept printer with a standard 225mm x 225mm print bed that could print its plastic parts. OK, obviously you can't print the motors, hotend or control board. But even that might be doable if a "TIG" welder and a printer were somehow combined. We already have 3D printers that can print in 2 or more colours. It's theoretically only a small step to deposit a layer of copper on a layer of carbon-fibre impregnated PETG or nylon, even plating holes through a board and doing complex, mult

This is a gallant attempt at immortality by creating a drummer-bot with a wrist action, for use with a drum machine, to create more realistic sounds and feels than the drum machine's in-built samples. Essentially, using MIDI velocity data, mapped to Arduino PWM, applies a light voltage for soft hits and near full voltage for hard hits to pull an actuator to make the stick (or beater) hit the drum. On MIDI note-off or timeout, apply a light return/hold voltage to quickly restore the stick (or beater) to ready position for the next MIDI note. One control unit per drum or cymbal, up to 4 actuators per drum or cymbal, depending on that drum or cymbal's complexity. For example, 2 sticks for snare, plus a mechanism to drop the right-stick height on MIDI "sidestick" note for rimshots requires 3 actuators. Hi-hat requires 2 stick actuators and 2 top cymbal actuators for half closed and fully closed. (PWM alone would not be effective enough.) Toms, percussion and straight

Like all of my projects, my robo-drummer needed unique branding. Meet WristBot, named for the drumming action I want to emulate, the "wrist-flick." Every robot drummer project I've seen online uses a linear pull action to swing a drumstick down in a fixed arc. No rebound, no ability to motor-tweak the stick angle to extract feel from MIDI velocity, no ability on snare to raise or lower the stick height on the sidestick MIDI drum note so that both the rim and the skin are hit. Drum machines have almost as much dynamic sound control as human drummers. My Alesis SR18 does. Even the HR16 I used to use in the 80s and 90s as bassist for the Hobart, Tasmania, blues duo, Bottleneck, had feel. "Way 'too much' feel! Don't get a real drummer!" in the words of one audience member. My aim with WristBot is to emulate that balanced stick, wrist flick grip, that allows the stick to bounce away from the skin, the instant the skin is struck. Even kick drum has &

(Authors Note: I'm actually going to use a digital servo motor to raise and lower the hi-hat, but I've been looking at a range of ways to close, half open and open the "hats" and there's still validity to this article.) No, they really do! As a weird example, I've been "researching" Arduino-based, robotic drum machines - Heath-Robinson affairs that allow a drum machine to play a real drum kit, instead of stored sounds. How do bicycles make this better? Bowden tubes. "Bowden tube" is the engineering name for the outer sheath of a bicycle's mechanical brake cable. You might call it "the hose" or just the "outer." The whole mechanical brake cable has not been used on one, single robotic drum system I've seen, yet, for the hi-hat, the upper and lower cymbals that a drummer operates with a pedal, as well as hitting with sticks, all have these complex linkages and gadgets and firmware code that gives the effect of

No, they really do! As a weird example, I've been "researching" Arduino-based, robotic drum machines - Heath-Robinson affairs that allow a drum machine to play a real drum kit, instead of stored sounds. How do bicycles make this better? Bowden tubes. "Bowden tube" is the engineering name for the outer sheath of a bicycle's mechanical brake cable. You might call it "the hose" or just the "outer." The whole mechanical brake cable has not been used on one, single robotic drum system I've seen, yet, for the hi-hat, the upper and lower cymbals that a drummer operates with a pedal, as well as hitting with sticks, all have these complex linkages and gadgets and firmware code that gives the effect of half-pedal, rather than an actual half pedal. They're using car door lock actuators (solenoids) to do this, these things are on or off, they don't have a half position. Or do they? Spoiler alert, they do, but the solution is mechanic

While detailing my logic controlled 2x2 matrix switch, in my previous post from a few hours ago, I referenced making a matrix switch out of toggle switches and a product called Make-A-Bracket. I wondered whether it might do the job... It actually might, except it's probably no less expensive than doing it with logic control! Maybe a bit less (holds up 2 fingers in a pich gesture), but as I detailed in the previous post, no EEPROM to store and retrieve settings, either. That's a sketch of the mid-century modern version up there . The main panel is the legendary, Australian (possibly Bunnings-only) hardware product that can get you a kludged project panel, housing or bracket without a 3D printer. It comes in various sheet and bent varieties, all predrilled to a 20mm x 25mm rectangular tesselation. This stuff holds bits of my cargo bike together (including a very stylish diacomp BMX brake to classic side-pull mount brake adapter that actually fucking works! Yeah, I was surp

Matrix (or Grid) switchers have fascinated me since I was a kid. You have a panel full of holes, inputs up the left, outputs along the top. (Or right and bottom, or both.) By sticking a pin, or shorted jack plug or throwing a switch, you can parallel, mix, match and repatch a synth or a mixing desk, etc, almost any way you want it configured. A friend in senior secondary school built an ETI 5600 from discrete parts and the pin-patchbay amazed me! Anyway, cut to the 21st century, and I want one that's logic controlled. And flexible. Like a Lego matrix switcher. I have a Zoom L-12 mixer/recorder and the monitor busses seem to me to be just perfect for mix groups. With this I want a switcher that can select my regular mic and instrument inputs and patch them through to my 12 lin-ins, or select my B, C, D and E monitor outs (via appropriate headphone to line matching), to give my studio greater flexibility. I could make a matrix switcher out of SPDT toggle switches, but that'd