/diy/ — /ohm/ - Electronics General

>>2989599 (OP)
>suspiciously placed huge walnut with an eye in the middle

I saw a guy 3D print normal plastic with the traces of a PCB layout embossed into it. He then stuck down copper foil/tape and sanded off the high-spots, leaving just the embossed traces. Soldering it would definitely soften the plastic, but if you do it quickly it shouldn’t be an issue even with PLA.

But the best way is definitely a fibre laser. They can punch holes through the FR-4 that are charred, allowing you to electroplate them with copper. I think dual-colour lasers can also make un-charred holes and slots and edge cuts using an ultraviolet or blue laser, but I’m not sure. Blue is definitely better for cutting through copper than IR.

>>2989609
>huge walnut
looks regular-sized to me.
wtf are pogo pins for?
also, anyone noticed a fuckup?

>>2989623
Pogo pins connect to the antenna, the thin metallic thing stuck to the inside of the shell. Obviously not a big deal considering it worked without the lid on.

Maybe some smoke started quickly offgassing when the last of the batts were soldered on? Hard to say.

>>2989624
Shorted the batteries with his tweezers turned the tip red hot.

>>2989610
>even with PLA
Doubt, but maybe with a lot of flux and high temperature one can do it.

>>2989674
It takes like 5 seconds max to make a joint to a non-power component. Even if the plastic under the component melts, if you don't deform it it won't be an issue, and it won't heat up enough to impact the structure of the whole board. Unless you're making 10x10mm mini boards. Making your board like 3mm+ thick probably helps.

You can always anneal your PLA.

File: BabbysFirstCircuit.jpg (1.69 MB, 3264x2448)

>>2989599 (OP)
>Where to get started?
Babby's first circuit.

File: images.jpg (6 KB, 220x229)

>>2989690
>Babby's first circuit.

>>2989676
Pla is half-step up from hot melt glue… it melts in your car.
The whole premise is ridiculous.
Real PCB material is probably cheaper than the pla.

File: 1749581539951388.jpg (63 KB, 832x803)

>>2989624
>antenna
ahhhh, right! ty.
>>2989628
>Shorted the batteries with his tweezers turned the tip red hot.
yep. space is so cramped that when he released the pressure on the tweezers, he ended up shorting positive to the ground.

wtf do you even find lipo cells this small? in vapes or some shti like that? I think I've seen a vape battery in a video someone did and they looked bigger tho...

>>2989706
If you anneal it, the heat deflection temperature goes up past ABS or even polycarbonate, to 150C or so. To anneal something flat, just turn your bed up to 100C after it’s printed and chuck a towel over it for an hour. If you’ve ever tried to plastically weld with a soldering iron, you’ll have a good idea of how poor a thermal conductor plastics are, only a few millimetres around your iron tip gets molten. If thats beneath a layer of copper, then it will hold its shape.

Fibreglass FR4 can’t be 3D printed, so its cost is moot when discussing this method of prototyping. You could print a negative then apply a thermoset resin like epoxy, maybe with fibreglass, and do the same technique on that instead. But epoxy is definitely more expensive per kilogram than PLA.

>>2989720
>watching some man from africa read a chatgpt script
ishygddt

>>2989728
sounds like you watched it, senpai

>>2989720
nigger

>>2989725
> PLA
Is literal waste from making greek yogurt, they didn’t know what to do with all the lactic acid from the processes until 3D printing came along.

>>2989745
Pretty sure it’s corn/maize derived. Biodegradable if you own a 50C+ bioreactor!

>>2989745
>Is literal waste from making greek yogurt
Hey! They could mix it with some fluoride (waste from steel making) and put it in our toothpaste.

>>2989747
> Biodegradable if you own a 50C+ bioreactor!
No, 50 deg C won’t do it any longer… I’ve been tempering it my PLA. Over and over. So now it’s harder than a diamond, denser than osmium, and melts 5 times higher than tungsten.

Actually, I’m thinking I can make PCBs from Turds if I temper them….

I can’t afford a $2 phenolic FR2 PCB after I bought a $3000 printer and a $1500 PLA drier, and a PLA subscription service. I’m glad I can eat the PLA though… thanks for letting me know.

File: Screenshot 2026-05-07 230647.png (52 KB, 791x514)

Do I recall correctly that surface mount electrolytic capacitors are difficult to hand-solder? I made my layout with thru-hole ones instead. I need a 1 uF bipolar capacitor to remove the DC offset from an audio signal and I've heard that ceramics have "microphonics".
Also does anyone know a good SMT equivalent of the 1N4148 diode? I found one with a SOD-523 package but I think that might be a bit on the small side, all my other components are 0805.

>>2989730
10 seconds of it, Big Chaks.

>>2989757
>capacitors
What's the cap doing? Doesn't look like it's a cap to ground, for any series caps in an audio circuit I'd want to use film caps. Though if it's under constant DC bias, an electrolytic isn't the end of the world, they can actually be pretty low leakage if you overspec their voltage rating. Polymer electrolyte caps should be better than conventional liquid electrolyte caps by any relevant metric.

>a good SMT equivalent of the 1N4148 diode?
1N4148W is what I'd go with. The WS is a small and somewhat delicate SOD-323 package (still larger than a SOD-523), but the W is in a larger SOD-123. There's also always MELFs, which are bigger still. Pic related shows MELF, 123, 323, 523.

File: retard moment.png (143 KB, 1114x1608)

>>2989780

File: 1747473417981695.webm (3.88 MB, 720x1280)

Question for /ohm/: is there any money in electronics repair? Do people even look to repair stuff in 2026? Can you even make a living wage with it?

>>2989790
Yes, but it seems tough. I know a guy that specialises in television and home audio, he does both home-call installation and repair. So I guess he's already effectively filtering for whate repairs are and aren't worth attempting. Because of the labour cost, you can only really effectively repair things that are sufficiently valuable. If that's new electronics, then it's often difficult to repair thanks to tiny SMDs, custom programmed MCUs and ASICs, and proprietary software tools. You might have a chance with those for items popular enough to heve been reverse-engineered, like smartphones, but microsoldering doesn't seem very appealing.

The market I'd aim for is analogue audio, especially vintage stuff but not only. Recapping amps and crossovers is easy on the eyes and the hands, while you're at it you can sell them bluetooth receivers and pre-terminated speaker cables. The harder fixes are still relatively easy to troubleshoot, by measuring Vbe and bias currents and such, and with repeating audio signals it's pretty easy to use even the shittiest scope. Even newer stuff can be pretty easy to work on, unless it's class-D or some sort of fancy multimedia HDMI amp.

I volunteer locally with a "repair cafe" once a month, there's definitely a lot of people out there looking to get broken things fixed. But not many people willing to pay a lot for that. If you can replace a micro-USB socket in a bluetooth speaker in ten minutes, maybe it's worth it. But more often than not it takes half an hour just to take it apart.

Is easyeda stealing all my data?

File: circuit.png (158 KB, 2422x1953)

>>2989780
>>2989781
Thanks, the MELF package looks like what I want, in fact I had some in a smt "practice kit" but didn't know what they were called.
The circuit has 3 caps in series at various places to remove DC bias that is introduced from OTAs, or to centre my waveform around 0 V after applying a DC offset. I think it's important to remove the DC around the filter feedback loop.
To be honest, this is not high-fidelity audio, just square and sawtooth waves, and I wonder if film caps would be overkill. I'm just using basic bipolar aluminium electrolytics on my breadboard which are about 1/8th the cost of the equivalent film cap. Easiest of all would be a smt ceramic cap but I've heard that it's "conventional wisdom" not to use them to decouple audio.

>>2989797
>The market I'd aim for is analogue audio, especially vintage stuff but not only
Interesting. Thanks! I guess that makes sense.
I was hoping for some computer stuff since I know a lot more about it.
>But not many people willing to pay a lot for that. If you can replace a micro-USB socket in a bluetooth speaker in ten minutes, maybe it's worth it. But more often than not it takes half an hour just to take it apart.
Yeah, 90% of things are unprofitable since the replacement is cheaper than the time it takes to open and diagnose something.
Items that have nostalgia and sentimental value are the only things people are willing to pay to have them repaired.

>>2989812
Yeah I’d just use electrolytic caps, probably 50V rated or more. Though in your case it looks like biases will be going positive and negative, so you’ll want to consider bipolar electrolytic caps.

File: Screenshot 2026-05-08 174705.png (459 KB, 1019x596)

>>2989856
Out of interest I tried swapping my 1 uF bipolar aluminium electrolytics with 220 nF polypropylene film capacitors. Now my square wave slopes downwards, but it still sounds the same. I'll probably stick with the electrolytics.

File: Screenshot 2026-05-08 175207.png (267 KB, 1012x601)

>>2989867
this is with the 1 uF
resonance is looking nice :^)

>>2989720
Youtube is so cancerous right now, how they manage to get people to listen is beyond me.

File: Screenshot 2026-05-08 212937.png (274 KB, 1643x1152)

I aligned everything to the grid this time (sort of) and used a consistent trace width. I feel like the layout could be more compact, but it still comfortably fits within the 10x10cm square that seems to be some price break point at JLCPCB

>>2989890
Gonna cram in some pin headers for input/output connectivity? Both for power and signals.

Also looks like a few of those blue traces could be a bit shorter, and some of the pads and vias could be spaced a bit more from nearby traces for clearance reasons. I also like to ensure vias of different nets (and any traces or footprints on the reverse layer really) are spaced apart from one another, such that the ground fill can get between them. You could also thicken up the power traces if you want, but nots not a big deal for this kind of circuit.

You can also use Mitxela’s PCB trace melting KiCAD plug-in if you want to make your boards look cooler.

File: Screenshot 2026-05-08 221422.png (72 KB, 907x796)

>>2989894
Yes this is about 95% done, I'll connect the pin header tomorrow. I know I should try and fit as much as possible on the top plane, so is it fair to say that the *number* of vias isn't so important, but I should optimize to minimize the length of traces on the lower layer?
Once I am satisfied with the connectivity, I'll turn grid snapping off to space out the traces a bit more nicely. Good pointer about the ground fill.

I'm a bit annoyed I have to waste an entire half of a dual flip-flop IC to make my frequency divider, but the single flip flop I ordered turned out to have much faster edge transition requirements than the rest of my circuit can provide. Maybe I could wire it up as a second frequency divider, to provide a sub-sub-octave....

>>2989900
>Maybe I could wire it up as a second frequency divider, to provide a sub-sub-octave....
I'd just wire it up with pads to solder wires onto, and a pull-down resistor or jumper on each input if needed. That way you can wire it up to whatever you want if you ever need to.

File: schematic.png (16 KB, 318x345)

>>2989599 (OP)
i want to build a dtmf tone generator to revive a rotary dial phone without janky ass microcontrollers doing shenanigans and use a tp5088n tone generator chip, i got everything figured out except interfacing the chip with the actual phone line, the datasheet gives this very simplistic schematic, would this actually work? wouldnt the dc current of the phone line fry the transformer? for the ringing ac im not concerned since i want to place this thing after the on/off hook switch

>>2989936
Maybe the transformer has a high enough ESR to handle continuous DC?

>>2989599 (OP)
why haven't IC manufacturers capitalized on photovoltaic inverter ICs the way they have with SMPS ICs? you can find hundreds of different ICs for buck converters, boost converters, flyback converters, full bridge converters, etc. but it seems solar is always done with an MCU.

File: secretary phone.jpg (88 KB, 815x739)

>>2989936
>wouldnt the dc current of the phone line fry the transformer?

of course not
you'd use a 600:600 ohm transformer which is the load that a typical phone puts on the line
however, you def need a hook switch or else the phone is forever busy
like every asian tranny hooker

but i'd suggest stuffing in a portable miniature secretary phone instead, as it has all the circuitry and buttons already

>>2989941
ill order the thing from aliexpress and hook it up, at worst i'll fry my modem

>>2989938
Because a proper MPPT algorithm is a looping process. Measure voltage and current, multiply them together, and vary the duty-cycle until the product is maximised. Even without the seeking logic, multiplication like that is a pain to do analog.

I’ve seen some cheap mock-MPPT ICs, where you program in a maximum power voltage via resistive divider, and its feedback loop attempts to keep that constant. Might be 5% less efficient than a proper MPPT algorithm, but it’s much simpler and cheaper. It was a Chinese brand chip like this that I used.

File: Screenshot 2026-05-09 184852.png (297 KB, 1335x848)

finished routing and attached the edge connector. I think this is looking better than my previous attempts. Tried to keep the jumps on the bottom layer as short as possible, it will otherwise be ground, +/- 12 V and +5 V

File: Screenshot 2026-05-09 204159.png (456 KB, 1364x1005)

>>2989999
I more or less didn't bother with ground fill on the top layer because there are so many traces that I end up with a lot of isolated areas. The back ground plane is fairly unbroken, though
thanks for reading my blog

>>2990009
Stitching vias. Needs more stitching vias to cover the long traces going to the pin header.

File: 5v perimeter.png (455 KB, 1364x1005)

>>2990009
>>2990017
Yeah I think either a bunch of stitching vias, or at least go out of your way to push the power traces to the very perimeter, like I've done here with the 5V rail. Such that no ground-referenced pads are outside them, ideally no signal traces either. R12, R67, etc. have a pretty roundabout return path.

>>2990009
Also C25 is backwards, if it's polarised.

File: Illustration2.png (1.40 MB, 2734x1995)

>>2990017
>>2990041
>>2990042
thanks anons, it's useful to be able to run these by people who know what they are doing. I tried harder to route the power traces and I think they break up the ground plane a lot less now?

File: IMG_5944.jpg (386 KB, 1832x1716)

What is this mystery pin?
I spotted a shiny thing on the carpet. It's not from any project of mine. I don't recognize this type of pin at all. It must have fell out of a device or appliance. I have no clue what though. Nothing around the house appears to be malfunctioning.

File: A_plug[1].jpg (484 KB, 2112x1188)

>>2990128
NEMA connector?

>>2990131
Far too small.
0.4mm thick
13.88mm length
2.54mm at the wider part
1.41mm at the thinner part
Not magnetic at all.

>>2990128
>>2990136
USB A contact?

>>2990138
Looking closer, 2.54mm seems way to wide for that. It still looks like a connector from some sort of plug. Especially given the plating, wear, being non-ferrous, and the indent on the smaller side.

>>2990062
Yeah that looks good. C29 looks a bit isolated, but it doesn't matter since the top-side ground-plane connects it to C24. It's not necessary anymore, but if you wanted to, you could run the traces from the pin-header to the THT caps on the top-side of the board, then run bottom-side traces from the THT caps to the rest of the board.

>>2990142
I hope it won't matter too much because I just ordered the board, lel. At that point in the board the top layer ground plane leads right back to a ground, pin so I'm hoping it should be fine.
The boards were about 50 cents each for 10. I don't know how JLCPCB do it. I suppose they're subsidized by their government as part of a plan to achieve global dominance or something.

Got hired by my friend to do a smart meter lora project for him for several buildings. This is great. I love doing shit like that, plus i can do it literally 10x cheaper than the cheapest commercial solution.
All those chinkshit parts i impulsepurchased from ali over the years and never used will finally come in handy. I have such a stock of them i don't even need to buy anything else kek

>>2990179
Check the current prices for the things you need to replenish and include/mark up that cost in your invoice.

File: led1.png (21 KB, 518x362)

>>2989599 (OP)
Which button do you need to press to turn on the LED?