heh, implying anything short of thermite can truly fix a printer
Mon Apr 01 19:51:17 +0000 2019no comment required on this blessed day
Mon Apr 01 19:55:09 +0000 2019I'm supposed to be doing my Russian homework so let's take apart one of these reciprocating-crank-LED-lamps things for the equivalent of one buck
Tue Apr 02 16:18:22 +0000 2019Replying to @nabijaczleweli
Right off the bat: two switch interfaces on the side, the one further forward toggles battery output, other one unused.
Tue Apr 02 16:21:34 +0000 2019Replying to @nabijaczleweli
Two sides of the cover no longer held together by PH1 screws, revealing the juicy flesh.
Tue Apr 02 16:23:31 +0000 2019Replying to @nabijaczleweli
Trash lens comparison, battery source
Tue Apr 02 16:24:24 +0000 2019Replying to @nabijaczleweli
Triple-parallel LED assembly, 1P2T switch, common Bat+, no external resistor
Tue Apr 02 16:27:26 +0000 2019Replying to @nabijaczleweli
Three LR44 cells, and the holder, which was quite smart actually, gotta give it that
Tue Apr 02 16:28:50 +0000 2019Replying to @nabijaczleweli
Close-up of the triple-parallel LED assembly, the two strands coming from under the spinny stack look suspiciously like magnet wire
Tue Apr 02 16:30:24 +0000 2019Replying to @nabijaczleweli
The crank action (neutral, full-in, extracted)
Tue Apr 02 16:31:15 +0000 2019Replying to @nabijaczleweli
Crank handle held in by steel pin, spring reset action provided by, well, spring. All white plastics feel like a nylon.
Tue Apr 02 16:36:54 +0000 2019Replying to @nabijaczleweli
Primary wheel removal. Small wheel interfaces directly with crank, big wheel interfaces with clutch wheel. Also: oily goop!
Tue Apr 02 16:39:18 +0000 2019Replying to @nabijaczleweli
Clutch close-up: engaged, disengaged.
Engages centrifugally when crank pushed and rotates CCW, disengages on CW spring return.
Tue Apr 02 16:41:04 +0000 2019Replying to @nabijaczleweli
Clutch stack removed, the black ring on the underside both (a) reeks of oil and (b) is magnetic, with roughly 8/10 poles?
Tue Apr 02 16:43:44 +0000 2019Replying to @nabijaczleweli
All that remains is the coil, which is held in through some weird clamp action on the housing; nothing a bit of Reasonable Force and aforementioned screwdriver couldn't fix, though!
Tue Apr 02 16:45:52 +0000 2019Replying to @nabijaczleweli
Coil, upside-down, notice both the reeky oil and the remnant of the housing inside.
Oh, and also: the strands soldered directly to the LEDs that looked suspiciously like magnet wire? They are magnet wire, and come directly off the coil (bridge would be Too Expensive™/lossy?).
Tue Apr 02 16:49:50 +0000 2019Replying to @nabijaczleweli
Oh, and the clutch just sits loosely on top of the magnet housing.
Tue Apr 02 16:50:42 +0000 2019Replying to @nabijaczleweli
This concludes The Dollarstore Adventures (a.k.a. I always wondered how cost-optimised these lamps are) for tonight!
Tue Apr 02 16:53:24 +0000 2019Replying to @violent_meals
wouldn't it be Юровская (Yurovskaya?) tho
Wed Apr 03 18:20:22 +0000 2019Replying to @violent_meals
well on one hand sure but on t'other Юровский is the m version and Юровская is the f version ya feel
Wed Apr 03 18:42:08 +0000 2019Step One: get a permanent waterproof marker, some laminate and a hacksaw (not pictured); attempt not to cut off your thumb nor your calf
Sat Apr 06 18:55:34 +0000 2019Replying to @nabijaczleweli
Step Two: attach one (1) piece of trash wire recovered from a car installation to the laminate cut-off
Sat Apr 06 18:58:44 +0000 2019Replying to @nabijaczleweli
Step Three: insert one wooden thingy and two "that's probably okay"s of water (not pictured) into some Indian Polypropylene
Sat Apr 06 18:59:39 +0000 2019Replying to @nabijaczleweli
Step Four: mount trash to Indian with springy
Sat Apr 06 19:00:29 +0000 2019Replying to @nabijaczleweli
Step Five: assemble death trap (yes it's still the same PSU from last time, I'm trash)
Sat Apr 06 19:01:22 +0000 2019Replying to @nabijaczleweli
Step Six: attach carbon rod from a battery to the negative and trash to the 12V rail, short enable to ground with tweezers, look at bubbles
Sat Apr 06 19:03:44 +0000 2019Replying to @nabijaczleweli
oooh look at the copper-blue wisps, neat
Sat Apr 06 19:06:45 +0000 2019Replying to @nabijaczleweli
Step Seven: assess what an absolute state this is
Sat Apr 06 19:07:16 +0000 2019Replying to @nabijaczleweli
Step Eight: wipe the laminate with a brush and decide that this is not going anywhere anymore
Sat Apr 06 19:08:09 +0000 2019Replying to @nabijaczleweli
Step Nine: reach the conclusion that you're too trash for electroetching and would probably need a better process to get anywhere with it
Sat Apr 06 19:09:58 +0000 2019Replying to @nabijaczleweli
This would probably've worked better with a more active etchant but the only other thing I have rn is conc. sulfuric, which would turn this from "unproductive electroetch attempt" to "local moron dies from breathing hydrochloric"
Sat Apr 06 20:23:28 +0000 2019Replying to @nabijaczleweli
If I had more than a quarter of a fucking brain I would probably have designed the board with the packing considerations in mind, but now it's built and doesn't fit into the socket holes anywhichway so it seems like a visit to bodge town is in order
Sat Apr 06 23:21:26 +0000 2019Replying to @nabijaczleweli
So now there's a problem to be solved: where in the ever-loving hell is this thing gonna go? I can see two options: (1) on the top plane, over the output cabling, near the wobbly AC in, or (2,3) on the side, over the mains caps.
Sat Apr 06 23:26:10 +0000 2019Replying to @nabijaczleweli
However, this means I'll have to go through the cover (and prolly replace keyed headers with flat ones). Now, the side location doesn't look all that bad (1), but the only place with enough clearance on the top is also sadly where the main out vent is (2), so over the HV it goes!
Sat Apr 06 23:30:21 +0000 2019Replying to @6b766e
like "каш" :v
Tue Apr 09 12:32:39 +0000 2019hash-tag transformer thursday or sth
Thu Apr 11 19:34:08 +0000 2019Replying to @nabijaczleweli
7 to roughly 250 turns, original core and primary from some PSU probably, secondary from the previously featured LED crank lamp https://twitter.com/nabijaczleweli/status/1113121372010577921
Thu Apr 11 19:37:28 +0000 2019Replying to @nabijaczleweli
yes I have a lot of russian homework rn how did you kn o w
Thu Apr 11 19:41:25 +0000 2019You know it's been a productive evening when the main takeaway is "maybe I'll not lose my right eye" (hopefully)
Sat Apr 13 22:41:56 +0000 2019Replying to @nabijaczleweli
Also: this trash https://twitter.com/nabijaczleweli/status/1116424208907669505? yep, it's trash (probably, I don't have a cap rated high enough to check, but it's def. not enough as a dump inductor).
Sat Apr 13 22:59:53 +0000 2019Replying to @nabijaczleweli
OTOH: a 240/12V@500mA transformer ran at 3A? Less trash, managed to get A Spark across the flash electrodes (and another one across the primary switch but hey it's Progress, going up to the 20(?)mm).
Granted, it did go up to the best part of fifty degrees, but what can ya do.
Sat Apr 13 23:05:34 +0000 2019Replying to @nabijaczleweli
Today in doing Very Smart Things (yes that's a hole right through the clip, yes I blew two breakers).
In unrelated news: driving that transformer from mains did not work for some reason, with the rectified voltage at the output shooting up to ~100V then discharging 🤔
Sun Apr 14 13:54:34 +0000 2019Replying to @__vlqc
🤔 https://crates.io/crates/lovecraft
Mon Apr 15 21:56:24 +0000 2019Replying to @thecoshman and @tweetsbi
it's almost as if individual employee satisfaction which yields to higher capital gains in the long-term is oft^W always thwarted by capitalists in order to acquire higher short-term gains, even as a means to a destructive end 🤔
Tue Apr 16 09:07:08 +0000 2019Replying to @thecoshman and @tweetsbi
Introduce artificial scarcity (a.k.a. KPIs/bonus systems or w/e) and it gets a whole lot cheaper /and/ works just as well, if not better (spite's the best motivator, innit).
Tue Apr 16 10:27:50 +0000 2019Replying to @tweetsbi and @thecoshman
Companies with a set goal (that isn't "generate capital") also survive longer, thereby generating more income in the long term, but :v
Tue Apr 16 12:13:19 +0000 2019Replying to @nabijaczleweli
Nope, not sketchy at all!
Tue Apr 16 19:45:00 +0000 2019Replying to @nabijaczleweli
Aaaaand back to where we were before I fucking smoked it (not pictured: 3 months' worth of dust)
Tue Apr 16 19:50:01 +0000 2019After a day-long soak in extraction gas with acetone, desoldering the b l o b b, and applying Unreasonable Force™ I finally managed to remove the bulb from this dead LED lamp!
Now if only the wires were longer than (2) shows this might be workable! fuck
Tue Apr 16 20:09:32 +0000 2019Replying to @nabijaczleweli
deliberately not including any photos of the setup I used to drill those because it'd give anyone with an ounce of self-preservation in their blood a heart attack
Maybe I'll just go buy a autopencil before I try do something even dumber trying to drill through this 1cm clearance
Tue Apr 16 20:22:37 +0000 2019Replying to @nabijaczleweli
In a way I'm extremely lucky I don't have an access to a dremel because I would /definitely/ be missing a finger, or a bunch.
Tue Apr 16 20:24:40 +0000 2019Replying to @nabijaczleweli
Well okay that's an exaggeration, I delammed a couple pads and the new components are a bit squint but it ~works so… progress?
Tue Apr 16 20:29:05 +0000 2019Replying to @nabijaczleweli
At least one of these holes should probably line up maybe
Wed Apr 17 14:25:42 +0000 2019Replying to @nabijaczleweli
Remember to learn from your mistakes, kids! Today, to talk to you about fixing your trash designs, we've invited R11, the forthleading expert in not blowing transistors!
Wed Apr 17 16:24:04 +0000 2019Replying to @nabijaczleweli
Finally found some scissors (harder than expected tbh), cut the wires, and oh god it's even worse inside, what kind of resin(?) is this, it looks like rotting flesh help
Wed Apr 17 16:32:45 +0000 2019Replying to @ClarisRobyn
Thu Apr 18 13:30:40 +0000 2019Replying to @nabijaczleweli
A further bit of drilling and not stabbing my hand later, we have:
* screws, going through a
* silicone tube because that's what I'd, meeting with the
* circuit board, and topped off with
* some cut-offs from the Quite Dodgy Drawer https://twitter.com/nabijaczleweli/status/1046888590721126401, laying 'round
Thu Apr 18 23:02:22 +0000 2019Replying to @nabijaczleweli
Now! This all looks cool and good and all, but there's literally zero damned clearance between the fan connector and the case, which is, indeed, very cool
Starting to think my "drill first, ask questions never" approach might be flawed
Thu Apr 18 23:07:54 +0000 2019Replying to @nabijaczleweli
on the other hand I just realised I can resolder the connector to the copper side and so what's the point of designing anything if you can just fix it in production ?
Thu Apr 18 23:10:50 +0000 2019Replying to @nabijaczleweli
an another absolute revelation that just hit me is that I could've just used normal panel-mountable potentiometer if I'd more than two active brain cells but
Fri Apr 19 09:31:10 +0000 2019Replying to @nabijaczleweli
And so, mind triumphs over being a dumbass once more!
Fri Apr 19 19:33:21 +0000 2019Replying to @nabijaczleweli
The timestamps say 4:55am so I believe them, despite failure to recall any part of the process
Sat Apr 20 14:18:04 +0000 2019Replying to @PancakeIdentity and @Vazkii
I mean . . that's kinda the goal innit
Sat Apr 20 18:44:47 +0000 2019how the hell ?..
Sun Apr 21 01:09:04 +0000 2019came across this old belt when cleaning and hmmmmmmmmmm
Mon Apr 22 20:52:53 +0000 2019Replying to @nabijaczleweli
More aleae were iacted, output delay sexion in place (this transistor is also shorted internally for what I assume is some very good reason)
Tue Apr 23 18:33:40 +0000 2019Replying to @nabijaczleweli
Includes a hand-filed potentiometer, because these never fit anywhere and are a general pain in the arse
Tue Apr 23 18:44:57 +0000 2019Replying to @nabijaczleweli
Snubber network for the pump, pulled back a little afterward, so the motor shaft doesn't grind through the diode package hopefully
Tue Apr 23 18:51:06 +0000 2019Not sure if it's the camera or Soviet technical thought but I'm having a real big-time trouble focusing down my old детский микроскоп nowadays :/
Tue Apr 23 19:13:45 +0000 2019Replying to @greenTetrah
sweet spot
Wed Apr 24 15:40:09 +0000 2019Replying to @nabijaczleweli
That went much easier than I'd thought!
Thu Apr 25 11:52:18 +0000 2019Replying to @nabijaczleweli
dunno what the professional way of doing this is but my soldering iron's gay now, so that's a win in-of itself
Thu Apr 25 19:46:51 +0000 2019Replying to @nabijaczleweli
к р ё м п
Thu Apr 25 19:48:23 +0000 2019Replying to @nabijaczleweli
This tape is both narrower (an inch vs 3cm) and thinner, so I'm only getting a relatively good grip when doubled-back, but still
yeee-haw, pardner
Thu Apr 25 19:52:58 +0000 2019fifteen minutes in ?
Thu Apr 25 20:44:14 +0000 2019Replying to @nabijaczleweli
another ten, goodness knows what this is, but it's edible! so
Thu Apr 25 20:56:43 +0000 2019Technically finished high school?
Fri Apr 26 14:28:03 +0000 2019Replying to @nabijaczleweli
l o m g friend !
Fri Apr 26 17:31:27 +0000 2019Replying to @PLT_cheater
an overengineered attempt at A Thing that only turns on the case fan of this ATX PSU when the heatsink reaches a pre-set temperature because it's too noisy to've it running all the time !
Fri Apr 26 18:22:48 +0000 2019Replying to @PLT_cheater
sounds expensive and quite susceptible to parallax, what with measuring the temperature of the fan vs that of the heatsink (also thermocouples scare me, as a concept), as opposed to what is essentially a 1-2 dollar-equivalents' worth of jellybeans tbh
Fri Apr 26 22:46:58 +0000 2019the lad down at the shop said he tried to bring this absolute beast of a pardner (en theorie holds 125.625 nameplates, says it's limited to 100) to life, but also that he couldn't get it to chooch, so here it is, on my floor, asking for… 115 to 230V at 45 or 66 Hertz? hmmmm
Fri Apr 26 22:53:24 +0000 2019Replying to @PLT_cheater
two euros… on top of a fourty euro (cheapest I could find in the right form factor from papst @ a local distributor, not incl. shipping) fan… would make it cost eighteen times as much as the PSU in-of itself…
Sat Apr 27 01:01:56 +0000 2019Replying to @PLT_cheater
I don't think I /have/ fourty-two euros plus shipping to my name rn ?
Sat Apr 27 01:08:04 +0000 2019Replying to @PLT_cheater
from my tests so far, running it on temperature trip for a good dozen seconds is ~good enough~, so vive l'ingenierie
Sat Apr 27 18:39:24 +0000 2019Replying to @UsualHotDog
mmmhmmm 🍞 bringin' 🍞🍞 dem 🍞🍞🍞 loaves 🍞🍞🍞🍞 baby
Sat Apr 27 18:53:37 +0000 2019Replying to @nabijaczleweli
Perhaps the most curious thing about this is the rubber grommet on the molex bundle, which is to say I'm yet to've seen an application which would accommodate it (unsurprising, as this is 5 years older than me).
Hopefully the socket is the standard one and not some abomination!
Sat Apr 27 19:32:54 +0000 2019Replying to @nabijaczleweli
having a very normal 11:29pm
Sun Apr 28 21:43:53 +0000 2019Replying to @nabijaczleweli
this fucking labcoat still reeks of goddamn sulfur; how long's it been? three god-forsaken months? ho w
Sun Apr 28 21:48:58 +0000 2019Replying to @nabijaczleweli
The Process turned out to be much less clean than I was led to believe, but we've achieved separation with zero deaths, so an overall positive result.
Sun Apr 28 21:57:41 +0000 2019Replying to @nabijaczleweli
The good: I've finally managed to print the AT{,X} connector pinouts.
The bad: it took a good couple hours, printers are cursed, printing software more-so still somehow.
The ugly: this is decidedly *not* an AT PSU, as opposed to what I've previously thought.
Fun!
Mon Apr 29 17:37:43 +0000 2019Replying to @nabijaczleweli
Absolutely *delightful* two-slide axion from the 1mm(!) pressed steel case.
The (retrospectively un)surprisingly-high-quality fan leads are ziptied together.
Similarly, all outgoing cables are crimped before going through the board.
This may sound idiotic, but I feel… unworthy
Mon Apr 29 20:56:24 +0000 2019Replying to @whitequark
eh, we've all been there
Tue Apr 30 13:09:36 +0000 2019The duality of mira
Tue Apr 30 19:35:46 +0000 2019Replying to @nabijaczleweli
Label says 1995, week 34.
Board says 1995, week 28. And that's already version 5.
That's five and a quarter years older than me.
It's served its function for over twenty four years.
I have not functioned even remotely correctly, ever, over my eighteen years.
i n a d e q u a c y
Tue Apr 30 21:07:39 +0000 2019