OK Let's do part 2 of my mega thread on hobbies, motivation & keeping things fun! we rejoin the quest for plinky in nov 2019. I'd forgotten why I started - to reignite my hobby with something simple & instead fallen down a rabbit hole of complexity.
worse, I'd gone on a side quest to learn about STM chips, an interesting intellectual journey but I had taken for granted that capacitive touch sensing was 'easy'. (part 1 was here: if you want to catch up)
fast STM processors don't have a captouch feature built-in, so I was planning to use an old trick that arduino's and even teensy2.0 era projects used: you connect the touch plate thru a huge resistor, and time how long it takes to charge up/down.
I had got this working on an STM prototype board, but not really checked that the results were sensitive & useful musically. oops. once I actually got hours and hours into the project, and actually tried to hook it to sound, it was awful.
it was very sensitive to 50hz noise, to the point that it would need so much software filtering that it wouldn't be fun to play. you'd think I'd have checked this, but here's the thing about hobbies: you do whatever is most rewarding, immediately.
so I'd not wanted to find out if it was a good idea. I just wanted to make things, not have to actually think too much. and now I was burned. I whinged to my gf, and contemplated binning the whole project, or returning to teensy.
but none of them felt 'fun', and they felt like a step backwards. ugh. I was so depressed. I went on a brief side quest - the physical size of my programming connector was too big, I couldn't program plinky while it was assembled.
this was the moment when I discovered how fast & cheap JLCPCB are. This is gonna be relevant in a mo I promise :)
at this point, @scolar off-handedly pointed out that I could use a secondary CPU with captouch to do the front panel. and lo and behold, smaller STM chips *do* have a captouch.
it suddenly dawned on me that I had hated using an LED driver chip anyway, and I could replace that driver chip with a CPU, and at the same time solve my captouch woes. MOTIVATION RETURNED instantly.
I tried it out on an STM protoboard, and sure enough a single resistor and single cap was enough to get really solid captouch readings for a single channel.
I quickly designed my front panel to have its own STM cpu, without thinking too much or prototyping anything. LETS DO THIS. its the only way out of my hole, so why wait? (why indeed...) after an excrutiating wait, it arrived! baking time
its alive!
...kinda. my mood was lifting. once again, I was on the up!
got the leds working ok. it felt so cool to obliterate an annoying expensive LED driver and replace it with a CPU.
OMG working captouch on stm! omg omg omg I am gonna take over the universe I love hobbies right
I was so high, I just had to go back to my mass production side quest. ITS GONNA BE REAL MAN
but once again I was getting ahead of myself. I carefully read the outputs across 16 touch channels, and some are better than others. with identical circuits. wtf. I still dont know why this is.
I decide Im going to ignore the slight-weirdness on some captouch builds, and build the test jig instead. its more fun that writing the software
and finally, I actually get some sound out of a complete plinky on stm. I should have done this long ago. captouch is still a bit dodge but... progress?
so at this point I made YET another big mistake. I was so driven to get plinky out on stm come hell or... bugs... I just kept ploughing forward, even as the problems piled up.
I think this is quite common in creative projects. it's like, 'oh I have all this debt and stuff is not going well but I JUST GOTTA PROGRESS otherwise I'll slide into depression'. don't do this to yourselves, people.
so despite my happiness being quite low - I knew I didnt understand the captouch dodginess, and I needed to write a *lot* of software to use all the expensive features of STM plinky - I carried on.
carried on how? why, going into mass production of course! <HEAD DESK>. I sent off an assembly order for 10 super-expensive plinkys. these things are monsters, right?
they have the top spec STM32H7 processor, AND the little one on the front, AND USB-C (untested), AND 64 megs of fast DRAM (barely tested, and unused by my software) AND an SD card reader (mostly untested) and... and... and... but it's boring having to test all thats stuff!LETSGO!
and so I clicked 'go' on the first assembly order. and duly the thing got made! china fabs are amazing.
and then, a few weeks lter, the assembled boards - 10 - arrive! nearly a grand. yikes.
...and the first one seems to work? kinda? <cautious optimism>
but then, the captouch is all glitchy! insane! my mood crashes. I think at this point my girlfriend wants me to get into the sea with all my mood swings.
I finish assembly (headers & thruhole) some more of the 10 boards I have, feeling awful and sick that I've wasted money. each one I build, fails in a different way. WTF. everything is awful. I dont have time to properly debug, this is happening 30 mins at a time in the evenings.
so just taking stock, at this point I have 1 barely working prototype, I've lost touch with the fun of the project, Ive made it super complicated, and I've spent loads of money. the guilt man. holy shit.
every time I look at my hobby table (all of this is happening in my bedroom on a tressel table, lol) I see the 10 manufactured boards taunting me. and I hate them, I hate everything about them. everything is awful. (I wonder if I'm borderline bipolar. or just a child...)
so I quit my hobby. or so I thought.
but.. Im not very good at quitting. especially because my electronics twitter feed is full of lovely cheerleaders
at this point I happened to have a week off work, and I was sitting stewing at home, when I made an offhand tweet about a tiny idea that had appeared in my head.
and I was excited again! here was an idea, with no baggage, that felt small and achievable. and I got reactions like this:
over the course of that night, I did a whole design, layout, etc. BBO (byte beat oscillator) was born! the hobby-joy was back!
I'd discovered the existence of JLC's new assembly service - which was amazing, very cheap, but required choosing all parts from a limited stock. constraints are so good! I was hooked. I had a new joyous project!
over the next few days, the design evolved a bit, sprouting a USB keyboard port (risky....)
tom @musicthing tried once more to save my from myself, but no! I am a maximalist who never listens :)
a front panel design followed rapidly,
...and then a PCB design
I sent it off! and then... chinese new year hit. all the fabs close for a week. DOH! while my order sat in limbo, I decided to build a prototype. I know right! sensible and shit!
I struggled a bit more than I expected, but luckily everything in the design sat in china was lining up to work. PHEW. hobby was fun again, plinky was forgotten...
I even got so excited I went back to software development, making sure that I would be ready with awesome firmware when the boards showd up
and then.... COVID-19 hit. I just can't imagine what it must be like to be in Wuhan right now. :( BBO was a 4 layer board, so it cannot currently be manufactured there.
So, I was forced to reflect. (can you tell!) where had all this journey got me? what was I doing? what about plinky? did I still believe in it?
I decided that yes, I did. but that I needed to return to that original, original simplicity. that first standalone demo with isla
so I redesigned an STM plinky around a single board, that can be assembled cheaply using JLC parts, that has no eurorack jacks, that is under 5mm thick. a pocket operator thing.
I was excited about plinky again! yes, I have wasted a grand. an iphone's worth. but I was fed up with beating myself up about that. I *needed* to get plinky back into something small enough to actually ship
constraints are good. I keep forgetting. so I tried to make plinky-cheap be one board; a small chip (80mhz!) with captouch; standalone. the one thing I let myself add was a battery.
I even had fun designing a case for it ...
and that, unsatisfyingly, is where we are today. China is in lockdown, BBO is stuck in the middle, and I have a new, untested, cheap plinky design that I'll fail on in some way I have yet to discover.
but I've also had a chance to reflect on my crazy rollercoaster journey, and what is worth while persuing, and what isn't. It's a WIP, so I'll let you draw your own conclusions!. End... for now :)
ps its been pointed out to me that i made a very bad joke about suffering from bipolar disorder. just want to apologise for that - sorry! it’s no laughing matter and i shouldn’t have equated my ups & downs to a serious illness that is quite debilitating for those affected. sorry!
