• srry for the ts

    s.mariam • 07/23/2026 at 21:32 • 0 comments

    This project took about two years. On and off, with long stretches where it just sat there and I didn't touch it, and other stretches where it was basically all I thought about. Looking back, I think it shaped how I grew up more than I expected a school project to.

    There was a lot of not giving up involved. Not in some dramatic way, just the boring kind where something breaks or doesn't work, and you have to go back and try again, for the fifth or the fifteenth time, without really knowing if this attempt is the one that finally works. The cogging fight was like that. So was the measuring. Tbh, so was basically every part of this build.

    I'd learned the theory in class long before I touched any of this. Faraday's law, magnetic fields, all of it on paper. Actually building something and watching it either match the numbers or completely refuse to is a different kind of understanding. Formulas stop being abstract once you're the one who has to explain why your generator isn't doing what the page says it should.

    There were points where this project genuinely got to me. Going back and forth to school just to get one reading. Rebuilding the same part over and over. Staring at a number on a screen that made no sense and not being able to just ask someone why. I think dealing with all of that, slowly, actually made me more resistant to stress in general. Not immune to it, just better at sitting with a problem instead of panicking at it.

    Finishing this feels like a weight off my shoulders, honestly. But it also did something I didn't expect: it left me wanting more. More building, more measuring, more figuring things out the slow, annoying way instead of just looking up the answer. This project is done, but I don't think that feeling is.

    Also, sorry in advance for any typos across these logs. English isn't my first language, and some of this was written half asleep.

  • Measuring in the Dark

    s.mariam • 07/23/2026 at 21:24 • 0 comments

    I don't own an oscilloscope, and nothing at home is sensitive enough either. The voltages here are so small that a regular multimeter just doesn't see them. So every single measurement meant packing up and going back to school to use theirs.

    And every time I went, it showed something a little different. Not wildly different, but enough that I genuinely couldn't tell if I was making real progress or just chasing noise. For a while I was basically improving the design blind. Tweaking the gap, the winding, the blades, with no reliable way to check if any of it actually helped until the next school visit.

    The strangest moment was on the channel measuring after the diode bridge. The scope showed a current that honestly shouldn't have been possible, way more than that tiny voltage could ever push through closed diodes. I'm still not fully sure what happened there. It appears to be related to how that channel was set up rather than anything flowing through the bridge. Either way, it taught me that a weird reading is still data. You have to figure out what it's actually telling you, and sometimes that feels like trying to understand someone speaking a language you've never heard before. :((((

  • Chasing Cogging

    s.mariam • 07/23/2026 at 20:49 • 0 comments

    Spent way longer than I'd like to admit just fighting cogging — the magnets kept sticking to the ferrite rods instead of letting the rotor spin freely. Went through blade after blade, size after size, until I was honestly sick of it. What finally worked was a huge wire skeleton wrapped in duct tape and hot glue, with the base reinforced as much as possible. 

  • The Nozzle That Fought Back

    s.mariam • 07/23/2026 at 20:25 • 0 comments

    Before any of the electronics happened, this project started as a doodle: what if a wind turbine's blades lived inside a duct shaped like a de Laval nozzle — wide at the inlet, narrow in the throat — instead of sticking out in the open? I don't own a 3D printer, so step one wasn't design; it was logistics: finding somewhere that would print two oddly shaped parts for a hobby project without charging me a small fortune or asking a hundred questions about tolerances I didn't have answers for. Took a few tries and a few "sorry, minimum order" replies before one local shop agreed to just print what I sent them.

    I modeled everything in Blender, cylinder and funnels alike, and printed two identical white funnels first — one for the front, one for the back. Only after holding both in my hands did it click that they shouldn't be identical at all: the front one needs to be more rounded, almost ball-shaped, to actually work as an air inlet, while the back just closes things off.

    So I tried to model that rounder front piece myself. I got the curve wrong. Not subtly wrong — wrong enough that when it came back from the printer, it looked like a funnel that had given up halfway through becoming a sphere. I ended up taking a hacksaw to it, cutting away the parts that didn't belong, basically sculpting the shape I should have modeled in the first place.

    That fixed the curve, but now the piece was too small in every direction. My fix for that was somewhere between improvised and ridiculous: I cut strips out of soda cans and taped them on with reinforced duct tape, widening the smaller opening and stretching the whole piece longer, one strip at a time, until it finally matched the rest of the housing.

    It works, and structurally it's solid; but if I rebuild this, that front piece is the one part I'd model correctly from the start instead of eyeballing it and fixing it with hand tools afterward. Lesson noted for next time: model it more elongated and slightly oversized to begin with. The full write-up with dimensions and photos is linked in the sidebar, for anyone who wants to skip the hacksaw step entirely.