Twinklr: notebook

Earlier today: where things are.

There are some rough edges, but a lot of things are working. When it’s good, it’s very good.

One wired-up back panel. As, slightly, expected, the amplifier is massively rubbish, but at least it’s there - the line-out works nicely.

In today’s post was a PCB.

Normally, a PCB would be nicely silkscreened and covered in soldermask. This it not; it’s a very hasty prototype ordered from PCBTrain in order to confirm everything works.

Because there’s no soldermask, I decided to leave off a ground plane and instead do a point-to-point ground net. This would be enough to wire up the board without risk.

The board does a few things: it acts as a power switch; it hosts all the electronics that go on to talk to the Raspberry Pi; and it hosts a 1W amplifier. If you plug a jack into the lineout, the amplifier is bypassed and you connect straight to the Pi’s line out.

This was also useful to confirm if the back panel had been cut correctly. So far: looks like it might be. I’m waiting on one last component - a volume pot - to arrive. And then we can wire the whole thing together.

If that’s a go: well, then it’s time to build the thing for real.

Good progress with the case. I think this is the final design.

It’s made of 6mm ply and 3mm acrylic. All the acrylic can be cut from a single 600x400mm sheet; the ply is about half a 600x400mm sheet. It’s designed to be bolted together with 14mm M3 bolts. A lot of measuring twice and cutting once has paid off.

I am probably going to take a second pass at the engraving - the legend on the DC jack to indicate whether the tip is positive or negative failed to come out - and I do wonder if I should attempt an acrylic infill.

Here’s one use of Twinklr.

I’m running the software on my computer, rather than the final box, and using a remote encoder. I’m spitting MIDI out of Twinklr and into a MIDI-CV adaptor, which is then playing notes on the modular synthesizer. Two playheads are moving in opposite directions; as I play, I tweak the attack of the envelope.

By the end of last Friday, this is where we were.

This is a full-size version of the case; I think it’s going to be this shape. The sides are 6mm ply; the rest of it is 3mm acrylic. The electronics all hang from the screen mount; the bezel on top covers that up.

I built this in clear acrylic because a) it’s useful for testing and b) it’s what I had lying around. The final thing will not be clear. Also, the final thing will be bolted together - this friction-fit was just to test sizing, and dial in the settings on the laser cutter.

Tom Armitage » Weeks 173-174

(via OP-SCHOOL 08: Sketch - YouTube)

“There are a few factors that have contributed to natural, human-feeling grooves in my drum machines. In order of importance.:..”

And then, Roger Linn goes on to describe what he thinks the core instrumentiness of a Linn drum machine - be it a Linndrum, a MPC, or a Tempest - are. I really like how clear and plainspoken this is, and entirely in service of the performer or musician, rather than of a specific ideal.

Today: drawing experiments with form.

(via Wintergatan - Marble Machine (music instrument using 2000 marbles) - YouTube)

Boom.

This is huge progress - albeit perhaps a tad behind schedule. I’ve just about sussed the CAD tool now, and am going to be much faster working with it in future - and, for instance, correcting that gap at the right of the case. Otherwise, we’ve gone from a 3D mockup to a well-fitting 2D maquette.

I definitely think the case might be a tad larger than this - certainly more raised - but I’ve now got a good idea how much space for electronics there is, and definitely want it to feel a bit softer.

It helped that we fired up the more powerful lasercutter today - this case took about five minutes of cutting, as opposed the the 100 minutes it was going to take on the Little Laser Cutter That Could...

I’m thinking a decent amount about the enclosure for Twinklr.

To that end, I’m playing a lot in Fusion360 to sketch out a frame - and to wrap my head around CAD software. CAD’s all a bit new for me - I’ve done a lot of Sketchup in my time, but nothing quite so carefully constructed as this. The goal is to make a case that can be lasercut, probably. And so I’m sat, building a case out of 3mm sheets.

The above is where I got to last by last night: a correctly modelled touchscreen, along with a suitably modelled Pi and Touchscreen PCB. I’ve made a hole in the “middle” layer to take everything that has to pass through it. Today, I’ve built up the rest of the case, adding tabs for support, removing lots of overhangs.

“...there’s something to a real music instrument. It’s not a piece of software, it’s not something where you - well, perfect example - you have a little keyboard with a row of unlabelled knobs. Well, what do those knobs do? Well, that depends on what software you have in, so it has this plugged in, it might do one thing, it’s this it’s something else. And that software might stop working in five years if you don’t update it every year. It’s just not a musical instrument. Something like this [the Prophet 6] if you grab this knob, it’ll do the same thing today, tomorrow, thirty years for now. It’s actually kinda nice - I could say you could keep this for 30 years. Because people have Prophet-5s from 35 years ago and are still using them (and in fact they cost more now than they did when we sold them).” - 1:30ish in An Evening With Dave Smith Part I

I built some new legs for the Raspberry Pi Touchscreen.

The original legs are desinged to hold it at about 60º, and at tabletop level. This moves it to about 75º, and lifts it up. The right leg has a large hole in front of the USB ports.

My digital calipers had run out of batteries, but were clearly accurate enough in their analogue calibration to get things right: a short while in Fusion360 (which I’m definitely getting faster with) and I had some legs ready to go.

Because the original legs were 6mm, and I only had 3mm acrylic, I had to make four legs. But everything fitted together in the end: the wobbliness of the design is exacerbated, but I could find a simple way to brace them I’m sure.

I’m not sure this is the final angle of a Twinklr screen, but at least I’ve now got something to play with. It goes quite well with the little wooden crankhandle stand. So now I’ve got some sample hardware to tinkle with.