/diy/ - Do It Yourself

'sup I'm making a custom PCB to house a raspberry pi zero in a Gameboy's chassis. Do you think this can work?
https://files.catbox.moe/zwty8j.pdf
It has an Atmega, linked to the pi with I2C, for input and power management.
I've like 2 battery protection circuits in daisy chain, is that ok? This is my first project btw
feedback? pic unrelated

I have a broken 4PST foot switch in a guitar pedal, is there any way I can use a different kind of readily available, cheaper, more reliable foot switch?

>>2761482

>>2761594
Well if it's already "broken" you could try drilling out the rivets and try to fix it.

>>2761476
are there any sub $50 cheap options or there for step+direction+encoder input dc servo drives? the higher the voltage and amperage, the better

So I'm trying to design a very low power induction heater with a ZVS driver and I'd like some advice on how to size the coil and the capacitor. I'm aiming for at least 200-300kHz.
Also here's what I've came up with so far, in case anyone spots some obvious mistake my dumb arse couldn't.

I'm looking for a type of switch mechanism that would allow me to build a control that looks similar to this. Basically I want to be able to push buttons/leaves left/right/up/down, as well as click in

Google is failing me for "5 way rocker switch"

I could accomplish the same thing with 5 momentary microswitches, but that puts a lot of leads close together in 3D space so I'm trying to find something that contains all of this already.

Any suggestions?

>>2761642
That switch looks skookum as frig.

>>2761717
yeah, get a 3d printer

>>2761792
I have a 3D printer
I'm not talking about building the paddles, I need an actual switch mechanism for them to click

I could wire up 5 microswitches all right near each other, like I said, but I wanna see if there's a better solution available

I have an old Pincor generator that I'm trying to repair. Originally the problem was a broken connection on the rectifier square, but now after putting in a new rectifier it still doesn't work. I saw these two other things wired to the rectifier and thought they were capacitors, but they ended up being more diodes. One of them had no continuity, but my question is what is the purpose of the extra diodes? Would that faulty diode make the generator not work?

Is there an easy way to interface with a +12V/-12V (slow) RS232 bus without a translator chip? I'm wondering if something can be done with a couple of transistors if I already have a +12V supply.

>>2761701
Is this some kind of oscillator? Can you explain it to me?

>>2761816
>but they ended up being more diodes

bullshit.
if they were diodes, one would be shorting the AC input and the other would be shorting the DC output.
so we can conclude that (1) they're not diodes, and (2) you have no clue, so leave the thing alone before you kill it.

>>2761925
>easy way to interface

one simple way is to use a resistor (in series) feeding a 5V zener (across).
this eliminates the negative section and limits the positive section so it can feed a gate directly.
this will work just fine 90% of the time.
fact is, lots of RS232 stuff breaks the rules in one way or another.
and it'll still work as long as the other gadget more or less conforms to the rules.

>>2761925
I heard the original ibm pcs had a compliant rs232 serial port. I expected some transistors, but they used some driver chips.

>>2761945
Also note the complete lack of 90 degree traces, because IBM doesn’t know anything about PCB design, or weren’t using altium.

>>2761933
I'm interfacing with a device that has a 5V to +-12V translator on the other side (MAX3243EIPWR), so I presume this translator will not really care. I just don't feel like adding that same translator to my breadboarding setup.

That said, I'm not exactly sure what you mean

>>2761953
nvm I understand now

>>2761717
There is a keyboardproject called lalboard that uses these switches. There is a Discord called Datahand/Lalboard/Svalboard Enthusiasts where everyone helps each other build those.

>>2761932
Nah they're definitely diodes. Literally has a diode symbol on it. A theory I've been considering is that someone nigger rigged a new rectifier system to an old rectifier system and it by some miracle worked until now.

>>2761926
Yes, more specifically it's a ZVS driver.
Voltage is applied to the gates of both MOSFETs at once (I forgot to finish the drawing, but they're IRFP260). Since no fet is equal, one will open slightly sooner than the other and the current flowing through it. That energises the coil but eventually it reaches it's peak and no current can flow, at which point the diode discharges the gate of that fet. The LC then tries to keep current going and in doing so activates the other fet and the cycle repeats.
The additional resistors and zeners are there to protect the fets.

>>2762000
>Literally has a diode symbol on it

nope.
that's the logo for International Rectifier.
they make thousands of different products, besides diodes.
could be a TVS, transient voltage suppressor.
one of those would normally show as an open circuit.

any ideas why this is so unstable as it moves through a range of frequencies?
it changes in 3 stages from a very poor sine wave to double beat heart attack, to regular ecg pattern

btw i was suprised to find that the circuit above still runs with both HT and heater supply under 6v. with a speaker connected to a 7 turn winding on the coil, audio output stops when the power rail hit 4.4v. i was always under the impression that tubes needed at least 12v to get any gain, but this proves that wrong

how do i increase the range on one of these things?

>>2762068
What frequency does it run at?

• RF amplifier -and/or-
• 50 ohm coax of length multiple of λ connected to an antenna -and/or-
• A directional antenna (e.g. yagi uda)

>>2761772
>made in italy
>skookum
They are shit, SHIT.

>>2762070
it says it runs at 433mhz
would changing the receiver antenna be a good idea? if so, what (omnidirectional) antenna should i change it to?

>>2762147
gift from pol

>>2762148
KEK
I'd ride it around town. Who needs cars when you have a pachyderm?

>>2762148
I eat it.

one bite at a time.

I installed a ceiling light today and I'm terrified of it being a fire hazard or something. I used one of those twisting things to join the black wires from both sides, and the white ones to the white ones. I took the ground wire from the fixture, ran it under the screw and then twisted it to the ground wire in the ceiling. Is there any actual fire hazard there? There wasnt anything in the blue thing to burn.
I just worry. I don't normally do electric stuff and can't sleep. It feels like it was too easy and I must have missed something.

>>2762223
yeah I don't trust those retarded caps worth a damn lol. but as long as the wires stay twisted and don't get yanked it'll last until the next sucker buys the place.
if you're really concerned, you can zip tie the wires together as a form of strain relief, so you aren't relying on sketchy chinkshit

>>2762225
The wires aren't under any strain so the zip ties wouldn't do much and I don't ever plan on selling this place. I'm just worried I'm gonna burn to death in my sleep. I can't think of anything I could have done wrong though. Worst case, if one of those caps fails the light just goes out right?

>>2762223
>>2762227
If there's no exposed wiring then you're fine anon. Wire nuts are used all over the place. If you're really worried about it, tug on the wire nuts. If they come off you gotta redo it. If they don't budge then you're fine. Odds are, if there is a problem, the wire nut will melt and the breaker would flip before your house caught fire.

>>2762148
I'd put Mr.Hands to shame.

>>2762223
>I don't normally do electric stuff and can't sleep

you're right to be so very worried.
suggest you find a person born with actual testicles to double-check your work.

Design review day today lads

Design review today lads

I'm building a 3x3 macro pad. Where can I find a good tutorial/guide that uses a raspberry pi pico microcontroller?

>>2762286
Damn, I knew 3d printed parts were terrible quality, but that on looks like it’s literally made from dried rice noodles painted black.

>>2762275
> never use the same connector twice
Atta boy… you’ll do well in big corporations.

>>2762235
One if Boeing’s greatest losses.

>>2762304
>looks like it’s literally made from dried rice noodles painted black.
That's because it is. Lol, jk.

I printed it on a raft and that's the support/separation layer, maybe I should have done brim instead. The outside looks better but I haven't dialed in my printer, as well as im using a scraped and worn tip. I'm about to upgrade my ender 3 v2 fans to noctua ones and I'll swap out my tip when I do. I didn't swap out my old tip because I'm saving the steel tip that came with my spider hotend for carbon fiber filled nylon.

I know my way around a PWM dimmer, but I've always just handled a bunch of LED on a breadboard. I now have the need to dim an incandescent light, possibly controlled by PID, so I guess I'd either have to switch mains voltage or step everything down to more sensible voltages and accept the increase in current draw. Suggestions?

Can I use a chink boost converter to transmit a higher voltage over a thinner wire with a buck converter at the far end to avoid voltage drop on cheaper thin wire?
What are the pitfalls?

>>2762352
Yes, that's how high voltage lines work. Do be aware that your chinkshit boost converter may be very noisy, and the long wires you're attaching to it are just like antennas. Doubt the FCC feds will come looking for you though.
You will also have some losses in the converters, so depending on the distance you want to transmit this power across, it may not be worth it to deal with converters.

>>2761796
Laser beams from cheap 25mW red laser modules could work, but you'd have to AC couple the receivers and modulate the red lasers so your whole setup doesn't get affected by ambient light.
I've done this kind of stuff before but with IR LED's and photodiodes, 555 sent a 1kHz signal and with a 1kHz bandpass filter on the receiving end I eliminated most natural disturbances. The only problems could be dust getting between the laser diodes and the receivers and the need for power.

>>2762305
I do not trust the assembly techs

>>2762399
>the long wires you're attaching to it are just like antennas

shielded twisted pair will reduce that.

I've made an over current protection circuit that measures low side but cuts high side. The idea is that when the comparator goes low the M1 Pfet turns on for a bit and charges the C2 cap to a higher voltage than the reference to give it some backoff time while dischargeing the cap, but it doesn't reset and also oscillates. I've built another variant of pic related without C1 that just latches and has a reset curcuit and it works well, but it feels like it could be simplified. Anybody have any ideas? Nevermind the high-side pfet driver and the specific components. I just threw it up and i couldn't be bothered to put in the proper circuit that i allready have.

Could this part be used somehow as a pwm controller?

>>2762286
Why would you go ahead and buy a 3D printer, key switches and a pico if you've got no clue whatsoever about electronics, then ask on /ohm/, not /mcg/?
>https://www.instructables.com/Raspberry-Pi-Pico-As-HID-Mouse/

>>2762463
do you mean the simulation oscillates or the actual device? i don't have any ideas but i just wanted to say that simulation may behave oddly with generic diodes. I'd assign some real parts to D1 and D2.

>>2761476
>makes new thread when it was on like page 5
>doesn't tell anyone in the old thread
>two concurrent threads have existed for 3 days now
YOU EXCUSE FOR A LIFEFORM

>>2762550
This is typically punished on fast boards but I guess /diy/ is not even moderated until someone complains about something.

>>2761476
>frog OP

>>2762560
It was you who did this. The frog hater trying to make frogs look bad. I knew you were a scorpion when I saw a hole where your benis should be.

>>2761482
For the DW01 redundancy, just leave those parts as unpopulated, and add a solder jumper across the FS8205A.

Any reason not to use Rpi GPIOs for buttons? Feels redundant, though I've heard Rpi GPIOs are a pain. You don't seem to have wired anything up to the GPIO header, but I assume that's coming later. Hope you're using a Pi Zero.
>composite video
lol, lmao

Also 16MHz may well be too fast if the 328P is running at the minimum lithium-ion voltage, consider using a slower crystal (or just dividing the clock in the registers/fuses). And add a diode on the ATmega reset line to stop it from being charge-pumped below the 0V rail or above the Vbat rail.

>>2761701
The voltage source needs to be able to actually source current into the inductor, but there's no link in your schematic between bat+ and the inductor besides indirectly via the gate drive circuitry. Most people use a pair of additional inductors, look at some more circuits. As for the values, I can only suggest that you simulate it in spice. The oscillation frequency of the main pair L1/C1 is pretty easy to calculate 2πf = 1/√(LC), but the other inductors for inrush current limiting, and the effectiveness of the gate drive circuitry, these are harder to analytically determine.

>>2761925
Yes. A resistor + BJT + diode circuit on the RX line will invert the signal and clamp it to your voltage rail of choice. A diode and voltage divider or just a zener+resistor works fine if you can invert it in software (or with a CH340G's R232 pin). As for TX, usually you'll want to boost the signal, but in some cases the receiving IC will have low enough and offset thresholds that it will treat a 0V and 5V as unambiguous logical 1 and 0 respectively. As was the case with my oscilloscope's MC145406.

>>2762275
>black solder mask
>QFPs not at 45°
>no curvy traces solely for aesthetic reasons
>no plating to reinforce through-holes
>XT-series connectors instead of EC5/EC3
>PowerPak SO-8 fets instead of based TO-220s
>programming pins not labelled in silkscreen
6/10
what's the design for? robot vacuum? where do the connectors lead? also that led better not blue

>>2762338
The common method is to use a mains-synchronised TRIAC dimmer. Usually this involves an opto-triac driver and an optocoupler-based zero-crossing detector. It can be used for phase-fired control, but in the case of an incandescent you won't notice flicker as much so you may be able to just skip cycles. Even more so if you're measuring temperature as a part of your feedback loop. The opto-isolation means you can have metal parts that are human-touchable like buttons, pots, thermistor wires, that sort of thing exiting your chassis without worrying that they'll be live. But if that isn't an issue anyhow, you may be able to shove a high-voltage MOSFET

>>2762576
fug posted too soon
>you may be able to shove a high-voltage MOSFET
shove a high-voltage MOSFET across a bridge rectifier's + and - pins, where the bridge rectifier's ~ pins are in series with the load. The driver circuit will have to use the - pin of the bridge rectifier as its 0V node, hence the whole thing will be referenced to mains. With a MOSFET your driving circuit will be able to PWM it at significantly higher frequencies than mains, and power-factor will be better than phase-fired if you use a bit of filtration. Also it's a simpler circuit.
Either way though, you need to care about a power supply to drive the circuit. Be this a step-down transformer, or an AC-DC SMPS board/module.

In a lot of ways, stepping down to a lower voltage for the lamp will be easier, but whether it will be cheaper depends on how much power will be drawn.

For the PID loop you can either get an off the shelf PID controller or program it into an MCU. Those PID controllers often have the PSU and switching element inside them, but are kinda expensive and I don't know how easy it is to customise them.

>>2762576
>>2762578
Thanks. Yes, I'll use a thermocouple for my feedback loop. I was thinking of going full analog with simple opamps for the PID part and trimpots to tune it, as I have them on hand, I really hate coding, and if I were to learn how to deal with MCU I'd never get this thing done. Power draw of the bulb is 25W, so it's not unreasonable to go straight to DC, I should even have a wall wart capable of spitting 2A at 12V DC in my junk pile.

>>2762584
Yeah I'd probably go DC, only go mains if you have to. If you use 12V, you'll be able to use the same PSU for the bulb as for the PID loop as for the gate drive, and automotive incandescent bulbs are readily available with a variety of power outputs. I'd probably use a quad op-amp for P, I, D, and summing, then a dual comparator for making a "triangle" oscillator and the final PWM comparator. Then I'd use an active pull-up circuit (https://circuitden.com/blog/11)) to give some extra drive to the MOSFET gate that a straight pull-up resistor wouldn't give, because it's not that much extra work.

>>2762463
This is the simplest overcurrent latch of that topology I could design. You should be able to reset the latch by putting a switch in series with D2. It requires a ground-sensing comparator, and the low-level output voltage of the comparator needs to be low enough to turn M2 off.

Personally I prefer not having an oscillation, more tactful would be to detect when the load impedance increases. But it should be possible to add an oscillation by putting Rs and Cs somewhere tactful.

>>2762148
circus train him for EZ cash.

>>2762550
who cares

>>2762569
I've since added those additional inductors, though I'm having trouble finding them (I have them, I just don't know where)...
What are some good splice simulators? I know kicad has one but I've never touched it.

>>2762609
All the people who spent the last 3 days not knowing there was a new thread, that's who. Not everyone spends every waking hour browsing the catalogue. The only people who don't post links in the old thread are fags from extremely fast boards like /b/ or /pol/. Oh hey, what's this >>2762148?

For what it's worth I'd go all John C Lilly and get a cute girl to try to communicate with it. While assisting from upstairs using telepathy.

>>2762644
LTspice. Watch a tutorial or two, you basically only ever need the ".tran 0 x" command, where x is the time period of the simulation. There's also the .ac command and multi-parameter testing and parts that need an extra model or command to run, but they're not necessary for a new user. In your case you may need to make a transformer, which itself is a trivial command. I'd also add "startup" after the main spice command, it helps semi-stable circuits like oscillators get off the ground.

>>2762659
>Oh hey, what's this >>2762148?
>>>/pol/459419120

New SiC transistors can commutate 3.3 kV and 400 amps, totaling up to 1.9 MW.
With really short turn on time, 0.3 mS or so. In small package of 8x10 cm.
Now, couple this with AED capacitor, or couple... And a coil

Man, finally freedom, outside of the US.... I mean eco-friendly alternative to conventional concrete nailer.

>>2762659
>All the people who spent the last 3 days not knowing there was a new thread, that's who.
how is it physically possible to be this clueless ahahaha
how do you find /ohm/ in the first place if not through the 'log? are you like a lost duckling if you ever close this tab?

>>2762486
>Why would you go ahead and buy a 3D printer, key switches
Ive owned these from 3d printing guns and building computers
>if you've got no clue whatsoever about electronics,then ask on /ohm/, not /mcg/?
Rude, everyone starts somewhere

>link
That looks gay, what would I even use that for.

>>2762716
Well at least it looks like SiCs come in good packages, unlike GaN FETs. Pic related.

>>2762751
I don't browse threads other than /ohm/, maybe a few of the other generals. I keep a tab of the catalogue open, but I barely ever browse it. I refresh the catalogue once a day or so, but don't exactly scour top 3 pages worth of threads for frogposters. Pretty easy to miss a single new thread if you're not terminally online. When I make the new general at a sensible page 8-10 I'll search for "ohm" and "electronics" just in case someone already made a thread, but why would I do that when there's still a week left in the thread?

>>2762275
Disregarding design choices (no clue about that topic), you might wanna clean up the area above the BRUSH OUT connector. The IC seems to have a bit of flux left on it, as well removing the solder mask from the vias above the capacitor array.
I'd also reflow C9, seems dangerously close to C10. Don't know what consequences a bridge would have but better be safe than sorry.

>>2762576
>also that led better not blue
>tfw I bought a surplus reel of of 3000 0805 blue LEDs and don't know what to do with it

>>2762757
>literally spoonfeeding
>"that looks gay"
Dude
If you want the pico to act as an input device, you need to set it up as such. What you want is an HID framework to flash onto the pico.
https://en.wikipedia.org/wiki/Human_interface_device
You only need to change the inputs the pico is sending to the computer, so instead of moving the mouse, you want it to send keyboard button presses in a specific order.
You wanted a "tutorial/guide that uses a raspberry pi pico microcontroller", this is literally it.

Also
>Rude, everyone starts somewhere
/mcg/ is literally a thread for everything microcontroller. If you have questions about microcontrollers, start there.

>>2762576
>Black solder mask
Boss' decision, but it does look cool
>XT connectors instead of EC5
ECs don't come in PCB mountable format like XTs do
>PowerPak SO-8 instead of TO-220
These Vishay SIR FETs just can't be beat on price/performance desu
>Programming pins not labelled
They're on the back due to the position in the machine ;)
It's going in a floor cleaning machine btw. Going to be making 10k units per year.

>>2762865
Do you want to sell floor cleaners for the rest of your life or come with me and change the world?

I'm fixing this tractor hydraulics computer and found a burned power resistor but I don't know what to replace it with.

>>2762874
This is the burned one. Checked nearby capacitors and other components haven't found shorts.

>>2762874
>>2762875
LOB-3
That's not a zero in LOB.

>>2762874
>>2762875
https://www.ttelectronics.com/products/passive-components/resistors/lob3/

>>2762795
>Well at least it looks like SiCs come in good packages, unlike GaN FETs. Pic related.
GaN is for microwave home security device.
SiC is for projectile home security device.

I'll be going to the IT junkyard in my college in a couple days with a friend to scavenge a couple of stuff. I'm personally looking for inductors, motors, transformers and power supplies. Is there anything else I should keep an eye out for? I'm expecting mostly old junk. Is there any chip that is particularly useful that I'm not aware of?

>>2762874
>burned power resistor but I don't know what to replace it with

0.08 ohms, 1%, size looks about 3-5 watts

>>2762913
For use or resale?
Resale, I'd be looking for anything big and old. Tubes come to mind, as well as transistors, maybe non-standard incandescent bulbs and CRTs (especially monochrome). Caps deteriorate over time, so I wouldn't waste my time on them.
Also ICs of everything related to video/audio. Old home computers, ISA video/sound cards, professional recording/production equipment etc. Some of those old chips are out of production and hard to replace because the manufacturer never released the specs.
Use? Idk, what are you into?
I'd be looking for the oldest device I can find and try to restore it just for fun.

>>2762913
Anything hifi and electrical. Mechanical hifi (cassette players, jukeboxes, etc.) are too prone to irreparable damage, though if you have a 3D printer you could give it a shot. Amplifiers and radio receivers are the best thing out there to fix up and sell, because they’re simple and easy to troubleshoot. Bonus points for remodelling the face-plate to suit a more modern aesthetic, if matte black and veneer don’t suit someone’s beige living room. You can also add Bluetooth or whatever to it.

>>2763042
For use, I doubt I'll find anything that valuable there. The good shit must have all been taken by their own techies. Still, if I can make some change from this, it'd be cool.
Do all caps degrade or only electrolitic ones? My stock is pretty low, but I can often buy them for cheap so I don't scrap. If the other types store well I just might...
Will keep an eye out for chips.
>what are you into
Lately, high frequency stuff and radios.
>>2763049
I do have access to a 3d printer, but my sense of aesthetics is non existent. I don't think I ever made a housing for anything, I'll happy with loose wires and switching things on by touching wires instead of properly putting on a switch.

>>2763056
Ok let me give you a rundown on zoomer hi-fi aesthetics:
>concrete, shiny or matte
>shiny black plastic, cannot be matte
>anodised aluminium, especially if it's coloured
>zero sense of depth, nothing protrudes from the faceplate
>edge-lit clear plastic that sticks out too far to the sides, shiny or matte
>throbbing RGBs (especially garishly bright blue lights)
>no visible wires at all
>capacitive touch buttons (capacitive sliders if you're fancy)
>no UI elements on the device at all, it's all controlled from your phone
>having all front contrast solely derived from LEDs, so when you turn the appliance off it looks like a featureless slab (with a massive logo for brand and model recognition)
>having no way to turn it off at all, it just idles in standby mode
>high fidelity sound effects each time you change literally any setting, flashing lights too
>high fidelity bass-boosted voice speaks to you "you're about to experience ______ like never before" when you turn the appliance on
>OLED screen displaying a meaningless swirly audio visualiser
>dark mode
>low latency gaming mode
>integration with phillips hue
>integration with alexa or siri (never both)
>plain matte black or white external power brick
>true wireless mode where it will connect to an identical version of itself to double the number of independant audio channels
>hdmi audio-only input (no hdmi passthrough)
>3.5mm jack instead of RCA (no mini-toslink)
>no front-facing aux input or heaphone output
>no fan or grates, terrible passive cooling via top of chassis
>screws are covered with feet or stickers, or replaced entirely with plastic clips
>rear panel is just galvanised pressed sheet metal

>>2763167
Hahaha…
Do alpha, gen-x and boomer æsthetics!

>>2762990
So it's not a power resistor but a current sensing resistor.

>>2762597
Well i would like the reset to be automatic with a delay time of a second or so. I already have working circut that resets by gernerating a periodic pulse that briefly lifts the reference voltage above the latches voltage on the negative input but it feels like it could be done with just an RC net somewhere, but i can't get it riggt without it oscillating after triggering

>>2763056
Electrolytic are the most commonly used, specifically aluminum. If you can find rather new devices, it might be worth it, just keep in mind that the bigger they are, the more expensive. Tried to resurrect an old matrix printer lately, had to pay 7 eurobucks for a 10,000uF/63V. So if they look like they were manufactured within the last ~10 years, you might be in luck.
Just keep this in mind: https://en.wikipedia.org/wiki/Capacitor_plague
Other types, like tantalum electrolyte and ceramic, are less common but also less prone to failure.
Don't know what that IT junkyard is holding, but if you're in luck, you might find really old hardware, like 1960s and before. In that case, keep an eye out for discontinued cap types, such as Mica, air and glass caps, for the same reason as the ICs. If you can find old radio equipment, they might have variable capacitors for tuning, which can get rather large and thus expensive.
https://en.wikipedia.org/wiki/Capacitor_types#Types_and_styles

tl;dr bigger + less common = more expensive

Also >>2763049 probably meant less of the front plate redesign and more of "small plastic pieces that degraded over time", like small latches, hinges, gears etc.
In the 1980s and 1990s CAD wasn't really there yet, so milled and injected pieces were rather simple and way less complex than nowadays. They'd rather use more pieces and glue/screw them together than waste three moths on a single injection mold. Such pieces are 99% puprose-made and impossible to get spares for, so 3D printing them is the second best thing to actually making them from metal stock.
On that note, if you want to get into repairing old hifi, you should think about getting into metal fabrication. I'd start with a small lathe and go from there. There's plenty of videos on how to get these chinkshit lathes to work properly:
https://www.youtube.com/watch?v=V3mu9NLjkR4
https://www.youtube.com/watch?v=05vUCdzhoe4
Once you got it right get a mill and rotary table.

>>2763167
>>2763172
How about gen-x?
>brushed aluminum or steel, silver or black
>big fat knobs for volume, input selection and menu navigation
>dimmed vfd or lcd
>mess of cables behind the stack, amp serves both as audio center and video switch for a dozen consoles
>UI is text as it should be
>controlled with proprietary 52-button IR remote
>front places focus on the most important buttons and knobs, display shows clock when turned off
>on/off button activates a physical relay shutting down the amp, leaving only RTC and display active
>no sound effects when pushing buttons, just the clicking of relays when switching input
>no voice effects at all
>no virtualizer
>low latency mode is your friends shouting "git gud, faggot" when you fall off rainbow boulevard
>integration via proprietary system link (Kenwood SL16 rules supreme)
>internal passively cooled psu
>hdmi, scart, component, composite, s-video, toslink with passthrough
>analog video and audio channels can be freely assigned to each channel
>rear aux input via RCA audio
>front 1/4" headphone out
>top and sides are ~70% cut out for air flow
>feet sit on long bolts for balancing on uneven surfaces
>rear panel is the same as the rest of the chassis, connectors are logically grouped for ease of use, even blind
T. Sony/Kenwood fanboy

>>2763260
>Hifi cabinet with glass door and magnetic latch
>barrel jacks with center-nigative polarity
>spring-loaded speaker terminals
>hard power switches (none of that standby voltage stuff)
>9000-foot AM radio dipole antennas
>8x D-cell batteries
>flashing 12:00
>cordless phone (landline) RFI
>gratuitous RF shielding (soldered to PCB)

>>2763269
Patrician taste

>>2763271
>modem handshake noises in 1200bps

How much thicker/wider should make my traces?

>>2763360
if only there were standards in place to calculate this... oh wait
https://www-eng.lbl.gov/~shuman/NEXT/CURRENT_DESIGN/TP/MATERIALS/IPC-2221A(L).pdf

>>2762329
Well, fuck me.
Didn't expect it to work.
There's still some mechanical problems and I need a new ribbon, but this is a good start.

>>2763363
Nice.
>CAUTION
Don't burn yourself on that edge.

>>2763385
That's the printhead, contains 24 needles and probably just as many solenoids. Gets hot with usage.

>>2763392
Make a tattoo gun with it and give a fentanyl zombie a hot forehead stamp. Use the old Print Shop RSVP font.

>>2763362
The toner/gap widths follow some distribution, but nearby widths are correlated. Maybe the variogram says that widths 5 mm apart are not correlated. Then I am only making a small error to replace the continuous 1 m with 200 independent samples. As the width increases it becomes less likely that a single sample will be too thin.
IPC 2221 seems to assume that widths outside of the +/- range are impossible.

>>2763360
You already failed when you steel wooled it.
You don’t need to steel wool, sand, grind, or abrade it like that and everyone that does ends up with with an amateurish looking pcb.
Steel wool is especially bad because it puts tiny steel nibs everywhere which react with the copper and form little pits when it reacts.
The steel wool meme probably came from jlcpcb to help prevent people from msking their own PCBs

Does anyone know why 2-layer boards typically only have a GND plane on the bottom and no Power plane on top and power on the top signal layer is instead delivered by tracks or smallish polygons even though most 2-layer PCBs still have a top polygon pour that is either connected to GND or left floating so it is just a cosmetic feature.

>>2763400
Once it fully works, I'll probably set up a Win98 or ReactOS machine as a typewriter for shits and giggles.

huh, these things actually work. I had a very shitty experience trying to solder something and decided to give it a try and the lead was sticking very well after this

>>2763460
Do it on the Commodore 64 so you have an excuse to build a Centronics adapter.

>>2763464
Already got a 25pin to centronics and a usb adapter and working C64 are expensive

How difficult would it be to build a discrete RISC cpu out of ICs? I have no practical experience with electronics outside of school.

>>2763459
Shielding.

>>2763572
Easy.

>>2763463
Buy an ad

>>2763459
Because you don’t want to capacitive couple your signal lines to a noisy voltage rail. Also all the signal traces will end up breaking up that plane, defeating the purpose of it. I usually have a mostly complete ground plane on the bottom, and a partial ground plane on the top, with plenty of vias to link them together. Though for a high-power circuit, noise matters less and I have less signal traces, so a power plane is something I’ll use. Often around FETs.

>>2763572
See Ben Eater’s breadboard computer.

>>2763585
>you don’t want to capacitive couple your signal lines to a noisy voltage rail.
I though capacitive coupling is a good thing. There is this grounding guru Rick Hartley who always talks about this: bring the signal plane as close as possible to either a ground plane or a power plane to reduce inductance and increase capacitance to keep the EM field constrained in a small dielectric layer / which would reduce EMI. Which is why he absolute hates the Signal - Ground - Power - Signal layout even though it is recommended in IPC-2221 for 4 layer PCBs. Go figure. I don't really understand this shit. It seems to be important only for really high frequencies and for compliance only in commercial products, not for signal integrity.

>Also all the signal traces will end up breaking up that plane, defeating the purpose of it
Yeah, I thought about that too. It doesn't make sense for two-layer boards but Mr Rick highly recommends this approach for multi-layer PCBs. For example:
Signal/Components/Power pour
Ground
Core
Ground
Signal/Components/Power pour

and a lot of via stitching.

>>2763615
Capacitive coupling to the from a signal to a ground plane is good, because the ground plane is your noise-free reference. Capacitive coupling from a power plane/trace to a low-impedance ground plane is fine, because you have capacitors from your power rail to your ground rail anyhow. But capacitive coupling from a signal trace to a power plane will impose power ripple atop your signal, bypassing whatever PSRR you have. Doesn't matter much for digital signals unless you have really noisy rails (e.g. motor drivers) but it definitely matters for analogue signals, the higher the signal impedance the worse the problem will be. Voltage references and regulators will only have capacitors to ground, not to the power rail, unless that power rail is really clean.

In Mr Rick's layup, there's only edge-to-edge coupling from traces to power fills, no face-to-face coupling. Only face-to-face coupling is between ground and power/signal. If I need the extra low power plane impedance, I'll replace one of the ground pours with the power plane for a section of the board, and keep my signals away from it. Low-impedance signals like from a current sense resistor are probably fine.

>>2763638
> there's only edge-to-edge coupling from traces to power fills
Do you mean on the same layer? How is it different than what you said above:
>But capacitive coupling from a signal trace to a power plane will impose power ripple atop your signal,

> face-to-face coupling
What do you mean by that? Plane-to-plane, i.e. from one layer to another one underneath it?

>>2763642
>Do you mean on the same layer?
Yeah on the same layer.
>How is it different than what you said
The cross-section area is far smaller, so the capacitance is a lot lower. Though for sensitive signal traces you may still want to add grounding guard traces on either side. Doubly so for transmission lines.
>Plane-to-plane, i.e. from one layer to another one underneath it?
Yes.

Can you wire low power leds of different voltages in series? How would I find out what value the current limiting resistor would be?

>>2763673
Yes, so long as they all want the same forward current value. Though it isn't usually advised since different colours will appear to have different brightnesses at the same current value.

The equation for a current-limiting resistor in series with LEDs is:
>R = (V_supply - V_LED) / I_LED
V_LED can be a sum of multiple series LEDs, and these forward voltages are dependant on colour. For example, a 2V red LED, a 2.5V green LED, and a 3V blue LED on a 9V battery will get:
>V_supply - V_LED = 9 - (2 + 2.5 + 3) = 9 - 7.5 = 1.5V
If the forward current is a standard 20mA, then:
>R = 1.5V / 0.02A = 75Ω
That's an easy E24 value already, if there isn't an easy size then just round up to the next one, like 45Ω going up to 47Ω. Additionally, the power rating is:
>P = R*I^2 = 75*0.02^2 = 0.03W
So you wouldn't need to worry about having to use a high-power resistor in this case.

>>2763675
Thanks for the good info, I never really thought about what would happen with different leds in the same circuit until now. Even if the brightness is different, it won't really matter, since it's just accent lighting on an rc car.

>>2763675
> E47 ohms
Is metric/european standard? Why make crazy values like 47 ohms, it would be better to make 10 ohms, 100 ohms, 1000 ohms, like that.
I noticed a lot of caps have stupid values like 4.7 uF as well. Bizarre.

>>2763693
It's a logarithmic scale. There are more values between 1 and 4.7 than between 4.7 and 10. It makes it easier to get resistor ratios of your choosing, which matters because voltage dividers exist. Additive relations matter much less in comparison.

>>2763693
>Why make crazy values like 47 ohms

using just 4 values per decade, you could make any numeric value by putting resistors in series (or caps in parallel.)
these are 1,2,3,6
so, using only 70 resistor values you could cover every single whole number from 1 ohm to 10 meg.

>>2763750
>using only 70 resistor values

wtf, i meant 28, not 70.
seeing double fingers from alcohol.

>>2763363
So the next problem is the carriage. Sometimes it goes too far to the right or left. Since it only checks the left dead point via a micro switch and only on startup, it keeps hitting the left and right edge of the rail it's gliding on. The motor moving the carriage has more wires coming from it than motors usually have, so I suspect some form of rotary encoder in the motor housing.
Two possible scenarios:
1. The encoder or the wires coming from it are broken.
2. The logic board has a problem with reading the encoder, either because of a cold/broken solder joint or because the IC decoding it was fried when the caps on the logic board blew.
Since both the motor and the IC are nigh impossible to get, I really hope it's just a solder joint.

Also, picrel is my jury-rigged PSU.

>>2763809
>jury-rigged PSU

current official nomenclature as per NIST IOS/TC276 is ''nigger-rigged''

>>2763811
Because niggers stands in front of juries so often?

>>2763809
Maybe the IR LED in the encoder died, or it's dirty and blocking the pickup.

>>2763750
>these are 1,2,3,6
what's the connection to his example of a 47 ohm?

>>2763809
>>2763818
The motor is a 7.5 degree stepper motor, which is still somewhat available, but I'd have to modify it, since mine has an additional plate with a protruding shaft for a gear on the top, which the ones I can find on ebay don't have. Seems spot-welded, at least I can't find any screws.
At least that explains the wiring, so most likely no rotary encoder.
Additionally, I found that the motor connector on the logic board goes to a group of darlingtons, then further to logic level inverters and finally to an IC I can't really find much about. It's marked as D65022GF116, made by NEC, the manufacturer of the printer itself, so I guess it's some form of ASIC.
There's a broken trace close to one of the darlingtons. I can't rule out that this has blown one of them, but since this trace goes to the positive power supply pin at least one of them is not powered right now. I'll fix the trace, if that doesn't work, I'll have to replace the darlington itself.

>>2763819

30+10+6+1

I want to make a few dimmers for some power tools and an electric fan that I got with no speed control. In the internet I found this schematic which is pretty simple so I like it for that, problem is I have no idea what's the max power this can handle. How can I figure that out and what components would I have to beef up if I want to increase it?

>>2763831
The triac is the one that's handling the load, but keep in mind that the circuit you posted kinda sucks and generates a lot of EM noise and heat, especially at higher power. Ideally you want something zero-switching or a proper VFD.

>>2763834
>VFD
Aren't those for induction motors? Those tools are all brushed and so is the fan.
Where can I find better schematics for this stuff? I'm not knowledgeable enough to draw my own.

>>2763831
>figure that out and what components

check current rating of your motors.
double it or triple it. then google a triac with that rating.
get some heatsinks. the bigger the better.
the heat sink will likely become live by contact with the triac.
there's numerous ways to isolate it, so Bing it.

>>2763860
Noice, thanks a lot.
What about what the other anon said, should I worry about the noise?

>>2763864
>should I worry about the noise?

absolutely not.
unless you're using million dollar liquid-cooled gadgetry to listen to faraway galaxies.

>>2763694
If it were useful you’d probably see it anywhere else, but you don’t. e.g. pipe sizes, box sizes, shim sets, light bulb wattage, fuse/breaker amperage, wire size, wine bottle sizes, fan speeds, thermocouple values, bolt/nut/screw/thread sizes, tape widths, weight, counterweight sets, etc, etc, ad nauseam.
I’m guessing the department of silly walks had nothing to do that day, so the dreamt this up.

Who’s not using 1% resistors for everything nowadays? Manufacturing is so accurate (sub-1%) they have to intentionally randomize the 5, 10%, and 20% resistors on the laser trimmer so boomers will have a chance to sift through their box of 4.7 million resistors and find the exact value they want with their simpson ohmmeter.

>>2763876
>so boomers will have a chance to sift through their box of 4.7 million resistors and find the exact value they want with their simpson ohmmeter.

It's funny how stupid old people are. LMFAO. And it's actually just that one generation. Everyone else will be smart in their old age.

>>2763831
My dremel’s speed control kicked the bucket almost immediately, so I use a light dimmer on it now. Works exactly as it did before, and I put a big “full blast” warning label on the end of the mains cable.

Same with my soldering iron—I just creep up on the wattage I need as I go.

I don’t spend several hours computing the thermal mass, trace thicknesses and widths, substate thermal coefficients, solder alloy type and girth minus the hollow flux core, time allotted, component lead thermal conductivity and heat dissipation capability, ambient temperature and wind chill factor from my fixed-speed flux fume extractor. If you have a variable speed fume extractor, you should link it to your microcontroller in your soldering iron using a proprietary fiber optic link, or perhaps bluetooth otherwise it will be impossible to solder anything.

>>2763864
> line noise, don’t worry?
I can tell if my neighbour is using is microwave oven, electric drill, or watching 8k porn just by looking at my bandwith meter. I wish I was joking.

Please look into an EMI filter.

>build dim bulb tester
>nervous about plugging it into wall
>plug it into wall
>doesn't explode in a shower of sparks
feels good man

>>2763828
I don't get it. 47 ohm is a standard value. one single resistor.
why would you combine 30+10+6+1, thats four resistors.

>>2763895
>47 ohm is a standard value

it's only standard coz some dumb-dumb made it so.
the 1,2,3,6 method would have made ANY value easy to obtain.
and, of course, people would design FOR those values, instead of the satan-inspired ones we have now.

>>2763899
cluttering the board with 4x more resistors is not fun. single values are convenient.

>>2763902

thx for missing the entire point.

If you really need very specific values, why not make them yourself?
SMD resistors are so cheap nowadays, it'd be easy to chain a few together, put legs on either end and pot them in epoxy. Wouldn't be much larger than normal 1/4W thru-hole resistors.
Or use picrel for small values.

>>2763970
Like this. Those are 10k each, making 8.3k. They're 1206, which is actually large enough to solder with an iron. 10k thru-hole for size comparison.

NEW THREAD >>2761476
>>2755510 NEW THREAD
NEW THREAD >>2761476

What are some good software for making gerber files?

>>2763876
Oh yeah with 1% resistors it makes little sense. It’s just a holdover from the years before cheap and good metal film resistors.

Also there is no analogue to division for those mechanical standard sizes you listed, unlike the voltage dividers with resistors. Nobody ever wanted one machine screw to have a certain ratio of diameters to another machine screw.
It’s also just generally more convenient to have finer spacings between sizes that are smaller. Metric screws for example, go down to M2.5, which is more of a standard value than M7. Nobody ever asked for a 4” and 7/32 diameter pipe.

>>2763878
Tetraethyllead does wonders

>>2764010
KiCAD, it can be scripted, and the board files themselves are relatively easy to interpret.

>>2764006
>after 152 posts

>>2764006
>>2764038
Wait a fucking second, that's the old thread. Which is at 388 posts, mind you.

Need to make a discrete oscillator (discrete as in out of transistors and not ICs, so no 555 timer, schmitt triggers, etc.)
Requirements:
1. Tunable. Can have it's frequency tuned with a potentiometer
2. No inductor
3. Only 1 capacitor
4. Only NMOS or NPN transistors.
5. Preferably as little component count as possible, but obviously with those requirements that might not be easy.
>why would you want that
I have my reasons.

>>2764060
The fuck do you think 555s and schmitt triggers are, numbnuts?

>>2764060
Square, sine, triangle, or sawtooth?

>>2764060
What duty cycle are we talking about here? Does that have to be adjustable?

>>2764060
look up how to turn a bjt or mosfet into an inverter with hysteresis, then use that with a low-pass RC filter feeding into its input. You may need three of these inverters in series to make it stable. You may find the phase offset from three mosfets is enough to get it to oscillate without any hysteresis, but I'd be skeptical about relying on it for low frequencies (e.g. capacitor >> fet gate capacitance).

>>2764026
> a certain ratio of diameters
Gears
> M2.5
Yeah, the world can’t continue to function with only M2 and M3. We need something not on the left side, not on the right side, but right in the middle.
Sounds like M2.5 was invented by apple’s malicious compliance with yet another absolutely bogus EU targeted directive to hinder their market penetration in europe in favour of Nokia 20-year old technology flip phones they want to sell in the EU again instead of the only people that will buy them—india in exchange for indian immigration of 50% of current eu population year over year.

>>2764060
Does it need to be guaranteed self-starting?

>>2764066
doesn't matter, take any of them. If you got for all 3 post them, I don't mind experimenting with all of them.

>>2764069
If it has an adjustable duty cycle, its great, if not, it's not the end of the world.
>>2764077
I already tried that, the three inverter in series is not enough to get the circuit to oscillate, at least not with the setup i'm using. Having more than one filter defeats requirement #3.
>>2764080
Yes, I'd like it to start oscillating all the time on power up.

>>2764060
> NMOS
How do you know if a transistor is NMOS or not? Is the 2N2222 “nmos” or is are all BJTs “nmos”?..

>>2764097
There are different types of transistors. Two commonly used ones are MOSFETs and BJTs. There are two types for each of them, for mosfets you have P-channel (PMOS) and N-channel (NMOS) and for BJTs you have NPN and PNP. 2n2222 is a bjt of npn type

>>2763902
Can’t your pick and place machine grab a few other values of resistors if it doesn’t have the exact right one, and twist them together and stuff those in the hole?

https://www.makershed.com/collections/beginner/products/make-electronics-ultimate-components-pack-3
how legit is this kit + book for learning actual fundamentals for electronics?
truthfully long-term the only electronic related projects i'm going to do is mess around with are keyboards/amps/dacs and was wondering if it would help to actually go through learning theory with the kit above or just learn it piece meal
or if that kit is a retarded idea for the projects i listed above just let me know

>>2764202
kek never mind found someone who priced everything from a vendor and it's borderline a rip off to buy the kit directly
but my other question still stands as to whether or not anyone thinks that book + projects it suggests is good for learning the actual fundamentals

>>2764060
>Preferably as little component count as possible

easy to do if you have a UJT instead.
or an IUD if a 28-day cycle is acceptable.

>>2764202
>>2764208
it's available on anna's archive if you want to download it and check it out (slow download server):
https://nrzr.li/d3/y/1708961212/200/e/lgrsnf/1619000/9805ad8baa32265bf3baa0393e664102~/OhvqOnoduKo_pRQCBKlAXg/Make%3A%20Electronics%20--%20Charles%20Platt%20--%202009%20--%20O%E2%80%99Reilly%20Media%20--%209781449378660%20--%209805ad8baa32265bf3baa0393e664102%20--%20Anna%E2%80%99s%20Archive.pdf
Or just look it up yourself if you don't feel like trusting that link

Anyways, I don't think it's really worth the money. I would recommend something like Forrest Mim's "Getting Started in Electronics" followed by some reading about the specific areas that you're interested in. Learn to solder and make a cheap IC amp even if you're doing it based on a youtube video or instructables lmao.

Why does my soldering look like actual shit (ie. turds)? When I lift my iron away, the solder clings onto it and it forms stiff peaks when I remove it.

>>2764208
>learning the actual fundamentals

nope.
it's a toy thing for making flashing things and beeping things.
fundamentals = all the gay math shit in Horowitz and Hill - The Art of Electronics, 3rd Edition
available at a fine torrent site near you.

>>2764215
>soldering look like actual shit

buy actual Kester solder instead of chink stuff made from tin cans and cockroach shells.

>>2764090
Pic related. The circuit can also be made with NPN BJTs, if you desire.

You can also make a reverse-biased BJT oscillator which takes even less components, but that's a bit of a hack.

>>2764215
You need to either be quicker at soldering to the point that the flux in your flux-core solder hasn't burnt away, or just add extra flux. Having a high-power iron helps with making quick solder joints, you ideally shouldn't need to use external flux with THT soldering if your tools and technique are decent. And your solder, but any old hardware-store flux-core leaded solder should work fine.

>>2764215
As the other anon said, the longer you heat the solder, the more you're burning up the flux. The flux is there to make sure the solder behaves like a liquid and flows into the right spot/doesn't cling to your iron.
Soldering is not a science, it's an art.
Ideally, you would somewhat heat up both ends of the connection you're trying to make to help with the bonding, then put in the solder. If need be, pre-tin the ends, though most vias and part leads are already chemically tinned.
At the same time, you need to be careful not to overheat anything, as parts can die from heat stress and pcb traces can lift up.
As said, it's an art. The only way to get there is experience.

>>2764101
What's the practical difference between them? I know they all behave like switches, closing the circuit if you apply a voltage at the gate but then why so many types? I think one of the mosfet types opens the circuit instead when you apply voltage but that's it.
Also what about triacs? I know they behave differently but they look the same, are there different types of those as well?

>>2764216
Why is math so gay?

>>2763809
if there's an encoder wheel, check if it cracked or slipping on its axis .

>>2764232
BJT needs a certain current ratio to be maintained between base and collector. MOSFETs do not have this problem. Also MOSFETs switch faster.

>>2764212
>>2764216
Thanks fellas

>>2763827
I've tried to find the original parts, namely UPA1436H and UPA1437H. They're out of production and the only shop I could find that still sells them has a minimum order value of US$250.
Since each of them has two separate darlingtons in one package, my next idea would be to replace them with discrete parts with the same (or very similar) characteristics.
They're well documented, alldatasheets.com has docs on both the NEC-made version used in my printer, as well as a Renesas-made variant with the same part number and values.

Anybody here with experience on transistors who can tell me about possible pitfalls going this route?

>>2764256
There most likely isn't, see >>2763827

>>2761594
Mind giving a bit more information about the part? Any part number, pinout or at least some dimensions?
Chances are they didn't spend a million bucks to engineer a switch for a foot pedal, so it's most likely an off-the-shelf part.

>>2764284
I used Octopart to find a darlington in a TO-220 package with very similar characteristics. Could I theoretically replace the 1436 with four of these?

Check craigslist for needs, variac, oscilloscope, soldering iron, etc. Old working quality gear tends to keep working. Project parts too. Guys that bought up 30 years ago now selling stuff off.

>>2764104
that's horrible. some people even use "calculators": "3 values in series plus a couple of them in parallel" will get you your exact value. instead of buying that odd valued resistor. or just rounding off to some standard close enough value. insane. resistors are cheap. just buy every single value and one day you will need all of them.

how exactly are you supposed to use hot air to remove components like these? I've been sticking tweezers under the chip and gently prying it off with the heat gun which is apparently the completely wrong way to do this. I don't want to heat the entire component and board until every individual leg is free because I'm scared of damaging other components on the board.

>>2764315
it is my impression that craigslist went to shit in the last few years. i used to buy used power transformers, even toroidal cores. since new laminations are super expensive, you could buy some of those commie to freedom voltage transformers really cheap. many were crappy auto transformers with aluminum wire, but i once bought a couple of 2000W ones from a Korean dude with some exotic moon runes on the case. all 10-12 gauge copper. Dirt cheap. So awesome. not anymore. nothing of the kind.

>>2764336
You use a hot air station (or IR if you have the $$$) and Kapton tape to secure surrounding components. You don't need flux to remove stuff even though people do it all the time. All that does is make a lot of smoke.

>>2764336
For larger boards with a heavy ground plane you might need to use a bottom pre-heater before heating the top side component with hot air.

>>2762275
i know that this is not relevant to your question but what is that going to be used for? is it a esc or what

>>2764315
hoarder mentality

>>2764215
Flux, baby. Flux.

Part warpage reduces the workpiece-nozzle gap height, which in turn causes more problems.

The upwards force on the nozzle depends on the height and width of the gap into which the filament is extruded. Moretti 2022 measured forces around 0.1 to 1.2N with a load cell. The signal is mostly (0.8N) due to gap width, and they didn't separate the height component. Whereas picrel is not a sensor, it is suggestive of active sonar or nc-AFM.

US8393009B2 seems to use the same piezo electrodes to both create and measure vibration. The usual AFM uses a PLL, but I don't understand how that applies when vibration sensor is also the vibration actuator. Is the drive circuit left out of the patent because it's obvious, or is half of the invention still secret?

>>2762049
Thank you very much anon. Assuming they were tvs diodes, I removed the one with continuity and used a dc supply to put power to the brushes, and now it works just fine. Next question is how necessary was that tvs diode? Should I bother replacing it or is the hazard of overvoltage minimal?

I like this watch, a Bulova Computron. There's only one problem, you need to press a button on the side for it to display the time. This is a battery saving measure. How could I modify one so that the time is always displayed? Any drawbacks, with the exception of reduced battery life?

>>2764391
>how necessary was that tvs diode?

it's a personal choice.
personally, i like to live dangerously.
but if it's gonna bother/worry you in the future, then replacing it will put your mind at ease.
problem is you likely dont know the specs, so you're never gonna get the full satisfaction you seek.
as foretold by the Rolling Stones in 1965.

>>2764284
I'd make a daughter-board with discrete parts on it.

>>2764336
Slide a desoldering shim underneath the pins as they melt. Gotta be thin, not too thermally conductive, and bad at adhering to solder.

>>2764404
holy shit it looks like shit
you don't need one
problem solved

>>2764404
It reminds me of those chink escalators that eat the bugs at the end of the ride.

>>2764221
>but that's a bit of a hack
Yes, I know about that one, not a fan at all, especially since it requires a pretty high voltage compared to your typical logic level circuits.

>>2764232
A BJT is turned on by injecting a current into the base. A mosfet is turned on by injecting a voltage into the gate (mosfet version of bjt's base). They have many differences, but to make it short:

The mosfet usually doesn't take any current into the gate to turn and remain on. This means it's very power efficient, produces less heat, etc. This is obviously not entirely true, since a mosfet gate has some capacitance, which needs to be charged/discharged that could be seen as "current", especially if that switching is at very high frequencies, as well as leakage, but when used as a switch, you can think of a mosfet as taking no gate current. In other words, mosfets have a high input impedance, which is usually desirable. That is much better than a bjt which requires constant current to remain on, but comes with some downside, first of all, since the input has a capacitance, removing the voltage is not enough, you need to discharge it with the use of pull up/down resistors. Second, frequency is usually limited on discrete mosfets compared to discrete bjts. Another downside is that the voltage required to turn the mosfet on can be quite high, especially on old types of mosfets. This means that they might not be suitable for 3v3 logic or even 5v in some cases, unless special care is taken in selecting the mosfet.

The BJT uses current into the gate to turn and remain on. This means it's terrible in power efficiency, and produces a lot of heat. To make things worse, BJTs have a positive temperature coefficient, that makes them unstable. More heat makes the base draw more current, which produces more heat.... boom. The advantages of BJTs though, are usually higher gain, lower voltage required to turn the BJT on (usually around 0.7V, vs mosfet that can be from 1V to 10V easy for some types of mosfets) and usually higher power rating and current rating. They also operate usually at higher frequencies than discrete mosfets (not true for integrated though).
1/2

>>2764232
In the case of NPN.PNP and Nmos/Pmos, the difference is mainly the requriements for the transistor to turn on/off

In NPN for instance, you usually need to give 0.7V above the emitter voltage to turn the transistor on. In the case of PNP, the reverse is true, you must provide 0.7 LESS at the base than the emitter to turn it on. Why is it useful? Because this allows you to create a circuit where the switch is on the HIGH side, called HIGH SIDE SWITCHING. The NPN is for LOW side, (LOW SIDE SWITCHING).

The NMOS and PMOS is the equivalent of the npn and pnp for the BJT, where in NMOS you need a voltage higher than the threshold voltage between gate and source to turn the mosfet on, while in PMOS you need a voltage lower than the threshold voltage between gate/source. Once again, allows you to do the switching on the high side or low side. There are many reason to do the switching on one side or the other, but usually you could design around them.

>I think one of the mosfet types opens
This is not related to the polarity (p/nmos) but operation type: Enhancement mode vs depletion mode. In enhancement mode, you need to provide higher voltage than gate to turn ON the transistor. In it's initial mode, the transistor is off. In depletion mode you need to provide a lower voltage than gate to turn OFF the transistor. In it's initial mode, the transistor is on.

>triacs
Those are similar, but instead of turning on/off a mosfet, you turn on/off a DIODE. You can think of it as two diodes placed in parallel but opposite direction, and they behave like an open circuit when no voltage applied, or like a diode when voltage is applied. In the case of a triac, since there are two of them, you can apply a negative voltage to turn on the first diode, and keep the second closed, and apply a positive voltage to turn on the second diode and keep the first one closed. It is very similar to SCRs which have only 1 diode instead of two, and only need positive voltage to turn on
2/2

>>2764221
Weird, I designed this circuit specifically previously and it would not oscillate (different resistor values though). Could it be Rf or R2 being wrong values? It seems to never oscillate and just reach SS. Maybe I got chinked by the mosfets.

I have on my hobby study a soldering station, air gun, hotplate, and a DIY reflow oven (lidl mini oven + raspy with pico reflow). But sometimes I spend long periods living in hotels due to my work. For now, I will be in the office for a long while I think, and I'd like to start getting busy again with electronics, but I've been thinking of doing something like a portable electronics lab next time I have such projects.

I find soldering fun and relaxing but I thought maybe I should build something like pic related for my portable lab. Can you somehow use it for two-layer PCBs? Any ideas? It´s just for simple hobby stuff so fine details (brittleness, tension, mechanical stress, etc) are not really relevant to me, I just want a portable method to solder SMD components on two layered PCBs quickly while traveling.

>fatguy_frying_whole_chicken_in_hotel_bathroom.webm

Well now I really feel like a retard. The generator had a schematic the entire fucking time, I just didn't see it because it was covered in dust.

>>2764428
>while traveling
You shouldn't solder & drive, Anon.
* unless it's lead-free solder.

>>2764432
>Captain's log: I can't FTFM.
>one fortnight later
>I finally FTFM. I had to CTFM to FTFM so I could RTFM. Spock is gay.

>>2764434

kek

OLD THREAD >>2755510
FRESHLY ARCHIVED >>2755510
READ ALL ABOUT IT >>2755510

>>2764432
What are “surge supressors” anyway. Symbol looks like a diac.

>>2764506
varistors

>>2764428
PCB-based reflow plates are not recommended, they get too hot for their adhesives to not slowly delaminate. You’re better off with either one of those miniware-style reflow plates, or a mains reflow plate with an STC1000 or PID controller or some custom circuit.

>>2764506
A diac is basically the same as a bidirectional TVS diode, which itself is like two TVS diodes in anti-series. Not sure if the conducting voltage drops lower than the threshold voltage for either case though.

>>2764511
>>2764512
I thought surge suppressors looked like blu ceramic capacitors.
Well, it’s no wonder we had trouble figuring what what those were.

>>2764603
That's a MOV (metal oxide varistor) and it's one of the components used to suppress transient voltage spikes (surges).

>>2764427
Yeah it's somewhat dependant on resistor values in my simulation. The schmitt trigger isn't entirely nonlinear, but you could probably fix this by having two schmitt triggers and the inverter between them.

>>2764603
MOVs are less precise voltage-wise, but they can absorb more energy. TVS diodes are often better suited for lower voltage cases, where you can be sure that it will trigger above the 12V nominal supply, and clamp to below a 16V maximum rating of a component, as an example. TVS diodes can sometimes also have foldback, where they trigger at X voltage, but begin conducting at a somewhat lower voltage. This is also classified as having a "negative resistance region" of its I/V curve. This is good when you want to limit the time period that a spike exists. Gas discharge tubes and spark-gaps do this to an extreme. DIACs too, though for a different reason.

>>2764427
oscillators sometimes need startup transients to get going

Is the Pinecil worth a shit?

>>2764671
yeah i like it a lot and its way cheaper than most alternatives, havent used it for super long so cant speak for its durability though
I use it with the official charger after a chinkshit ugreen 60w charger i was using nearly exploded with it

>>2764675
Are you in the US? Wondering about how long shipping takes.

>>2764678
Europe, was about 2 weeks

>>2764675
Just ordered one. Setting up an electronics bench today, and it looks like I left my old soldering station in another state.

>>2764681
That's not too bad.

>>2764671
No, tips are expensive and they evaporate on you within a few soldering sessions.
Mine's been sitting unused for years because it's more economically viable to buy a new chinkshit iron every year and make my own tips out of thick copper wire.

>>2764408
Good idea, but the execution is not my main focus right now. I first need to know if I can just swap it against another part with similar specs, what specs are most important to match, whether something could break etc.
I'm so close to getting it fixed, I don't want to fuck it up now.

>>2764404
It's an LED watch, they obviously draw more power than LCD.
https://www.youtube.com/watch?v=9E65KfMwTqY
You could, of course, find out how exactly the button works. If it stays on while holding it down, it's an easy mod, just hardwire the button closed or replace it with a latching button or a micro switch (think DIP switch). If the button starts a short time span in which the watch lights up, it does so either in logic or some form of delay timer (called TOF), which could be done with a few discrete parts.
https://www.youtube.com/watch?v=wgvo4aTATsU
If it's a discrete delay, again, bridge the button.
If it's in logic and holding it down doesn't work whatsoever, there's two possibilities. Either the button simply closes a contact on the common pin of all LEDs or it closes each individual contact on the LEDs that are supposed to light up at a given moment. In the former case, you could bridge that common pin to its destination. Be careful though, there's no standard on how they're wired. The common pin can be either cathode or anode.
https://thisvsthat.io/common-anode-vs-common-cathode
If it's the latter case, however, and the logic individually turns on every pin, you'd have to replace the entire logic i.e. you're shit outta luck.

>>2764434
So that's where leadfoot comes from?
Dropping leaded solder onto your foot while driving?

i'm trying to fix an older (1990s/early 2000s) bench SMPS and one of the decoupling capacitors on the output current sense op amp is blown up. the decoupling capacitor is a through hole 10uF, 25V tantalum cap on the +15V positive voltage rail of an OP07 op amp.

is there any reason why they would choose to use a tantalum over ceramic? the SMPS is all through hole, but they have numerous other ceramic capacitors used in the design. whole reason i ask is because i have a handful of ebay mystery chink branded 10uF, 50V rated ceramic TH capacitors on hand... and i don't feel like ordering tantalum bullshit if i can get away with ceramic.

>>2764736
Tantalum caps are more resilient to voltage and temperature fluctuations and don't drift much in value over time like other electrolytics.

>>2764215
>>2764372
As in use more flux

>>2764736
>choose to use a tantalum over ceramic?

in that era, regular ceramics only existed up to around 0.1uF.
MLCCs existed but they were NASA-like expensive.

Just dropping by to say this general has the best OP image

>>2764761
The elephant fume extractor made its rounds on /pol/ the other day.

Just dropping by to say this general has the worst OP image

>>2764773

>>2761476
Anyone have the apu who tied up the elephant to use as solder fume ventilation

>>2764693
>chinkshit
ive seen a bunch of replacement tips on aliexpress for the ts100 tips that should fit on pinecil, have you tried any of them?

>>2764753
Tants and those mlcc caps are very fragile though.
The dielectric in a tant is an oxide layer around sintered tantalum dust—thats how they get big capacitances and tiny voltage ratings.
If diy, I always over-rate them, make sure they stay cool, and absolutely prevent any physical stress on them (including thermal stresses)
In commercial products, please continue to use the Absolute Maximum Ratings table to spec your nominal working parameters so I can get more free shit that was expensive and easily fixable, like with the continuing electrolytic capacitor debacle.

>>2761476
What's the name for a miniaturized switch that can be toggled electronically and retains its state without consuming power?

>>2764889
Magnetic latching relay

>>2764891
Those look big. I don't have that many mm's depth to work with, I'm looking for something to put into (completely custom, variable resistance) flooring, so ideally less than 2mm thick. I'm hoping for something with an SMD form factor, doesn't matter how small.

>>2764891
Yeah I found this:
https://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Data+Sheet%7FV23086-X0000-A005%7FF%7Fpdf%7FEnglish%7FENG_DS_V23086-X0000-A005_F.pdf%7F8-1416000-9
Supposedly these are some of the smallest magnetic latching relays.
But even if I file off the terminals, that leaves a 1cm height and ~12x12mm footprint. That's pretty big holes I'd have to make in the top layer flooring.

>>2764889
Unironically a 6-8 pin microcontroller with built-in EEPROM. Because you can't buy discrete floating gate transistors, flash/eeprom memory cells, or feram memory cells.

Storing a bit with a few microfarads of very low-leakage capacitance across the gate of a mosfet might work fine for months or even years, so long as you're careful to prevent leakage currents in/out of the gate while it isn't powered on, pic related.

>>2764902
Right, I was trying to avoid that (maintaining state by storing a charge), but... it's a grounding system, there is some voltage to leech.
My electronics reading is basic and rusty, the... tripple thingies with the arrows, they're "N-channel, enhancement mode, MOSFETs"?

>>2764893
Nah. They make tiny reed latching relays for switching RF signals and whatnot.

>>2764915
>N-channel, enhancement mode, MOSFETs
Yes. But the normal 1µA drain-source zero-gate-voltage current and 100nA gate-source leakage current will mean you only get a few minutes of life from a 10µF capacitor. You'd need much lower leakage currents, and maybe some sort of supercapacitor or battery. A coin cell lasts a watch for years, would that be acceptable? If not, I think an MCU with EEPROM is the easiest way to go.

>>2764919
>reed latching relays for switching RF signals
Don't RF signals just capacitively couple across relay contacts?

>>2764919
Do they latch also if there's almost no current passing through them? I'm making a regulable resistance flooring system, there won't be more current involved than in regular ESD flooring.

why not make a high voltage distribution system completely floating with no such thing as a ground connection? just hot and neutral. everything is in the air and floating. sounds perfectly safe. why do we need earth at all??
just two wires, you can grab hot wire and nothing will happen.

>>2764945
>what if instead of having two wires, where one is shorted to ground and thus at the same potential as your body...
>...we had two wires, both of which are at whatever potential they are and thus hot in relation to ground
I don't know, anon.

>>2764947
> hot in relation to ground
but i just said what if there is no relation to ground. both wires are floating with respect to earth and are running all the way from the power plant to your house.

>>2764945
>why do we need earth at all??

we dont.
80K people saved yearly simply doesnt justify extra expenditures in copper.
also it lowers society's intelligence by preventing many Darwin Award bestowments.

>>2764950
Your home and your body and everything touching the ground is at the same electrical potential as the ground. That doesn't mean it is zero, we call it zero because we can't measure objective zero and needed a reference. Neutral is only zero because it's shorted to ground, if it wasn't there would a difference in potential between that and ground and then it would be hot in relation to you.
What does exist if you live in bumfuck no fucking where (quite rare nowadays) is single-wire earth return. Where you only have one wire (hot) and you make a neutral in your house by grounding some piece of metal and that's what you use to close the circuit.

>>2764960
I don't get it. Why would there be a potential difference with regard to earth (dirt) if neither of the wires is connected to it. Why would touching the hot wire complete the circuit? Why would electrons flow from the hot wire through my body to earth, and then where? what completes the circuit?

Or for example what if I connect a bunch of car batteries in series and then touch either terminal. Nothing will happen because I won't complete the circuit and there is no reference to earth even though I am standing on earth. That is called floating connection. Why not apply the same idea to AC power lines.

>>2764969
>what completes the circuit?
A difference in electrical potential.
Electrons go from higher electrical potential areas to lower electrical potential areas. If there is a difference in potential and some connection between two areas, then you have electrons flowing (aka current) or, if both are at equilibrium (same electrical potential) then the electrons are just there chilling.
Potential is how much electrical charge you have in a body. Electrical ground is a body which can give or take electrons and not have its electrical potential change. This is ideal and doesn't really exist but if the body is real large (like yo momma) then the charges can disperse within that body and not cause an appreciable change in potential. The Earth is that very large body.
We cannot know what's the electrical potential of the earth, we can only measure differences in potential, not potential itself. So we called earth's potential zero to have some standard.
If you touch the hot wire and you're touching earth, then earth can act as either positive or negative (depending on the phase of the AC current going through the wire) and current will flow because there's a difference in potential. Neutral however is shorted to earth so that one shouldn't have a difference in potential and thus you could (but shouldn't) touch it and not get zapped. If neutral wasn't shorted to ground, it would be at whatever potential it wanted to be and even if it was still at the same 110V difference from hot, it could (and probably would) have a non zero potential difference in relation to ground.

>>2764958
Darwin has nothing to do with a screw getting loose in your metallic appliance and moving a live wire to the body of the appliance.

>>2764981
>Electrons go from higher electrical potential areas to lower electrical
careful not to mix conventional/real current.

>>2764978
It would complete it, but you wouldn't feel that because batteries are DC and your skin is pretty damn good at isolating yourself from that and you can only feel currents past a certain point.
Emphasis on "skin", some dumb burger died because he tried to measure his internal body resistance with a 9V multimeter.

>>2764987
Sorry anon it doesn't make any sense. So the current would flow from the + terminal (conventional current flow!) through my body, into the earth underneath me, and then where? How will the electrons magically find the way to the negative terminal?

>>2764987
One probe in his mouth and the other in the pooper?

Anyway, here is what the wiki says:

>The ground conductor is usually grounded (earthed) at the top of the supporting structure, to minimize the likelihood of direct lightning strikes to the phase conductors

That's the ONLY reason.

So in a world without lightnings, there would be no need for grounding, bonding, earthing, GFCI, stray voltage wouldn't be an issue, ground fault wouldn't be an issue, etc. All because we are afraid of lightnings.

I thought you guys knew your shit. I've never read so much nonsense in a few posts.

>>2764990
The earth itself would act as the negative terminal (or positive if that were the case, earth can do both).
Remember that earth is our standard for zero potential. Whenever you have some non zero potential, it means it is at a different potential in relation to earth.
>>2764986
Why do we even need conventional current anyway?
>>2764991
>The sailor took a probe in each hand to measure his bodily resistance from thumb to thumb. But the probes had sharp tips, and in his excitement he pressed his thumbs hard enough against the probes to break the skin. Once the salty conducting fluid known as blood was available, the current from the multimeter travelled right across the sailor's heart, disrupting the electrical regulation of his heartbeat. He died before he could record his Ohms.
Source:
>https://darwinawards.com/darwin/darwin1999-50.html
First time I came across this tale, the report said he intentionally poked holes on his thumbs to try test it after a lecture on electrical safety which involved measuring the skin resistance. Can't find that one.

How to not burn the leds when soldering? I am trying to make led lamps that will last a long time to replace my car bulbs

>>2764999

>>2764998
>the sailor tale
so do you actually believe it?

>>2765009
I find it funny so I want to believe.

>>2764999
- Use hotplate
- Use hot air
- Use higher hot air temp and shorter contact time
- Use lower hot air temp and cover the top of LED with kapton

Usually using hotplate (old clothes iron) gives me the best results (that is LED package not melted).

>>2764996
RCDs require a ground connection to work, not the other way around. If you're working around electricity and water (e.g. a bathroom, spa pool, outdoor cable trench, etc.) it's pretty hard to be absolutely sure that a wire didn't touch the water somewhere. If that water is already grounded via pipes or fittings or whatever, then that instantly trips the RCD and you're safe. If not, you get a situation like that disney world pond. If you ground every thing that could accidentally be connected to live and protect it with an RCD, you make it safer.

These guys saying the floating wires would cause current to flow just because there's a potential difference don't know shit. Even if there is a difference, there's such a high impedance that minimal current will flow. Significant current would only flow if there was already a low impedance path established. Like someone jamming a spade through a cable conduit.

>>2765001
Good workholding is important, as is a good tip geometry that can put heat into the joint quickly. For those I'd probably use a knife edge tip or 2.4mm screwdriver tip. Recently got myself a ~5mm screwdriver tip, it's great for DPAKs and such, but might be a tad too large for those SMD LEDs. If you're soldering them to old pads, you want to remove as much of the solder bump of one of the pads as possible, in order to have the LED sit flat. Then reflow the other bump (with fresh flux) as you push the LED down into it. This process should take a second or two. Then you wait for it to solidify, maybe wait for it to cool down for a bit, then add fresh solder and heat into the other joint. Replacing all lead-free solder with leaded solder will make your job easier.

For desoldering without delaminating pads, a really wide tip that can reflow two joints at once can be handy, otherwise you may just have to cut the component off. Or use hot air I guess.

If I need a certain voltage DC, can I just rectify mains voltage and pass that through a buck/boost converter? Seems too simple a solution, what's the catch?

>>2765094
>what's the catch?
That if you want to touch any piece of metal on it, it will be referenced to mains. This is fine if the entire electronics is going to be fully insulated from anything external, but if it has plugs or wires or anything leading out, you'd be safer having an isolated switching topology, like a flyback or forward converter. This is what common AC-DC wall adapters use. There are dedicated ICs for this purpose that alleviate some of the issues you commonly run into with mains powered SMPSs, especially the high input voltage. This voltage is usually far too high to power a switching controller directly, so these ICs are usually initially powered by some sort of zener-like regulator that charges up a capacitor, followed by a seperate winding of the main transformer that provides a lower voltage rail for the IC's operation. They also often use a winding for isolated voltage feedback, though something like a TL431 on an optocoupler is pretty common too.

what kind of cap would you add to a 12v ac generator to stabilize the output

>>2765127
Not him, but wouldn't be possible to connect 30 diodes in series to drop to the required voltage? And it sounds pretty safe, since the voltage is reduced gradually.

>>2764671
>>2764693

>>2765141
Yeah. Christmas tree lights do that all the time, dropping, say, all of the 120 V mains.

>>2765139
Depends on the load, and the available number of phases, etc.
High loads need bigger caps.

>>2765094
Virtually every SMPS does this, yes, and then smoothes it with a 400 V electrolytic.

>>2765141
You mean like a zener regulator? Normal silicon diodes will take like 140 of them to drop a hundred volts.
If it's for the actual power output, then it will have like 2% efficiency, and so will dissipate an absurd amount of heat for any reasonable load. Capacitive droppers are more useful, though they have their own problems.
>safe
Still mains referenced. If you use a single diode rectifier and can be sure that the 0V rail is connected to the neutral and never the live, it's kinda ok, but still sketchy since it bypasses any GFCI.

If it's for a power rail for powering a switching converter, then yeah sure but you'd need to keep its current consumption pretty low. At 170VDC (rectified and filtered 120V mains), a 10mA draw is 2 watts, and many switching controllers will be drawing more than that via their gate drive circuits.

>>2765139
There's an equation relating current draw, frequency, voltage drop, and capacitance, for a filter cap after a rectifier. See if you can derive it from I = C*dV/dt. Hint: the frequency after a full-wave-rectifier is twice that of the input frequency.

>>2765165
i'm not rectifying

>>2765170
then what are you stabilising? is the input noisy, if so how much noise? does the input frequency or amplitude vary, if so then by how much?

>>2765173
lets just say a hand crank dynamo driving a bulb i don't want it to dim immediately if i slow down for a few secs
it's probably freq that changes right because amplitude is probably set into the design of the generator? i have no way to measure how much

>>2765176
Well, yes. That and the speed at which the coil moves relative to the magnet is also a factor. Higher speed = higher voltage.

Also consider a flywheel instead of a (super)capacitor for storage.

>>2765176
So basically you want to stabilise the output voltage/current/power for a given load. All the same thing really, though if you have a dynamic load it may change. There's a few ways to go about this:
>massive flywheel - will mechanically buffer the speed of the dynamo, which is proportional to the output voltage, you can calculate how large it will need to be via conservation of energy and power
>rectifier and capacitors - will electrically buffer the voltage of the dynamo, if you wanted to make it AC again you'd need an inverter circuit (but why would you do that for a bulb), this is the only option that could also keep your output frequency constant
>some sort of switching regulator that can actively change the output voltage throughout an AC cycle - basically a more complicated PFC circuit, not recommended unless you have a hard-on for keeping your power-factor high
>a variable transformer connected to a motor - not that efficient but kinda fun if you're into mechanical shit i guess
Option B is going to be the easiest and cheapest.

What's a cheap, abundant source of powerful 24-72V DC motors? Motors for things like golf carts and forklifts are ideal but there are none available near me.

>>2765198
yeah but what kinda cap? any non polarizing one rated for 12v?

>>2765207
Lime Scooters are free. They can be found laying discarded on sidewalks all around many towns

>>2765207
Treadmill motors aren't as big as forklift or golf cart motors, but they're still kinda beefy.

>>2765217
A capacitor won't do shit to AC besides filtering out high-frequency noise. It won't buffer voltage on a timespan of 10 seconds, because the AC voltage is changing over a timespan of 0.1 seconds or less. The only way a capacitor will buffer voltage for this system is if it's put after a rectifier. Since it's after a rectifier, you can (and should) be using polar aluminium electrolytic capacitors. Anything else will be less cost efficient.

>>2765154
Whut

i really do not like designing single sided boards
spent like 4 hours on this bullshit
gonna spend another 4 reworking it tomorrow because the ground fill is ass and there's no room for a programming header
and that's after coopting unused pads on the mcu to shuttle ground and vcc about the place
makes me want to bend the pins on some of these parts to be upside-down instead
think i'll give up and add a bunch of jumper wires for signals so the power traces can be more direct

gonna spend the rest of the day watching nisemonogatari

>>2765044
Can you put a tip on a soldering gun?

>>2765233
> programming header
Whats that for? Don’t you just take the mcu, put it in the ZIF socket on the programmer, then put it in the socket on your board?

Jumpers! Chad move.

>>2761476
I see a techtronic tds 1002b oscilloscope and a 3101 function generator bundled together for $750 cad. The function generator is the same model ive used at my college. Is this oscilloscope a reliable model? I've been told i shouldnt spend less than 1k usd on a oscilloscope due to noise, since these are depreciated models I cant find prices.

>>2765094
buck/boost converters are non-isolated and usually fail short circuit. will everything downstream tolerate being fed 170V DC?

>>2765225
>A capacitor won't do shit to AC besides filtering out high-frequency noise
damm mind = blown

>>2764900
>>2764902
ATTiny10
>16 8bit general purpose registers
>1KiB program memory
>32B SRAM
>1 16bitTimer/Counter w/ prescaler and 2 PWM channels
>watchdog w/ on-chip oscillator, can produce internal interrupt
>4 I/O, each w/ 8bit ADC if VCC>2.5V
>in-system programmable
>VCC 1.8-5.5V
>selectable clock in power-of-2-multiples of 128KHz up to 10MHz(TS8R)/12MHz(TSHR)
>active 200uA @1MHz/1.8V
>idle 25uA @1MHz/1.8V
>power-down <0,1uA @1.8V
>6pSOT/8pUDFN
>literally too cheap to fake, cheaper on mouser ($0.438) than aliexpress
https://ww1.microchip.com/downloads/aemDocuments/documents/OTH/ProductDocuments/DataSheets/ATtiny4-5-9-10-Data-Sheet-DS40002060A.pdf
https://www.youtube.com/watch?v=71K1L3ZJR8g

>>2765249
function generator looks nice but that's an outdated as hell oscilloscope.

>>2764900
>>2765301
Totally forgot

>>2764902
>you can't buy [...] eeprom memory cells
Not the easy way at least.
There are small eeprom modules but they require SPI or I2C, such as the AT93 and AT24 series in UDFN, TSSOP or even VFBGA
You could go even smaller with the AT21, a 1KiB 3-lead SOT23 with pseudo-I2C adressing.
That is, if you want your brain to melt.

>>2765310
>extending the memory address space by abusing I2C addresses
Used that before. I "love" compatibility.

>>2764998
>9V DC
>stopping a heart
A quick search brought up this study:
https://www.sciencedirect.com/science/article/pii/S0006497120663030
Depending on a myriad of factorss, the resistance of blood is anywhere between ~400 and tens of thousands of Ohm. Disregarding the fact that it needs pulsed current (defibrillator) or plain AC to disrupt the sinoatrial node, 9V at a maximum of 23mA and as low as in the uA range are barely enough to make your tongue tingle.

https://www.youtube.com/watch?v=COkIXaMbtZU

>>2765312
Tell me 'bout it.
I2C is already a clusterfuck if implemented half-assed.
>buy QMC5883 magnetometer module
>check example code
>works fine
>tweak for a couple of hours to do exactly what you want
>clean up code
>stops working
>fml
>try different module
>try different mcu
>spend 9001 hours to find out what went wrong
>completely rewrite the code from scratch
>find error
>furiously curb-stomp the damn thing
>magnetometer has 6 registers, lower and upper 8bit each for x/y/z
>only updates registers with new readings if ALL have been read from in the previous cycle

Lads, I'm stuck and need some advice.
I need to make a through hole slip ring that's got a 70mm ID and a height of < 15mm.
It needs 3 channels and must carry occasional loads of 500mA for a lightly used servo. Rotation speed will be max 30RPM.
I've looked around and my best bet seems to be to make a PCB slipring. It seems something approachable, I just need to know where to buy the brushes. Any ideas what they're called or anywhere that sells them?
I would struggle to make the brushes by myself so any advice is appreciated. The commercial ones seem to use gold plated or a brass/copper piece of bent plate. I've tried searching for switch contacts or brushes on RS/mouser and haven't found anything.

>>2765317
Nickel-plated steel strip cut and bent to shape. Use a form to bend/press them so they're all identical.

>>2765317
Might be able to salvage them from other parts. They're sometimes used to connect daughterboards, or maybe pull apart a potentiometer or an RJ-style socket. Maybe some of these spring pins used to connect rechargeable batteries (picrel) but they might be a bit too stiff and eat away the trace rather quickly.

>>2765317
>contacts or brushes on RS/mouser and haven't found anything

old pots have dual springs you can cut out.
some dollar-store spinning light toys may have all you need already made.

>>2765302
For features i think its just missing compared to the keysight DSOX1204G: wavegen, digital multimeter, and bode plot. I think my main concern would be reading waveforms without extra noise, whats the issue with the scope?

>>2764215
Use way more flux and don't use a shit solder

>>2765094
The catch as people pointed out is isolation. For safety, you need the output to be galvanically isolated from the input, otherwise the entire device must be insulated from the user. Instead of using a buck/boost try a flyback topology, those are isolated, but transformer selection is a bit more complex.

Hello frens, I'm new to soldering and absolutely suck at it, or at least my gear does. I've been using cheap soldering irons for some time and I've noticed that my tips get destroyed faster than they should even for a chinesium iron. I'm using 99% Sn solder and non corrosive flux (it's not rosin) so the chemicals aren't supposed to be so destructive.
Is it possible to make soldering tips out of thick wire used for tig welding stainless steel? It's 316SS and comes in 1/16, 3/32 and 1/8 diameters

>>2765404
>99% Sn solder and non corrosive flux (it's not rosin)
It may be worth trying leaded rosin core solder, at least until you're confident in your abilities. Lead free solder can be temperamental even for experienced people.

>>2765404
Do what other anon said, try a more forgiving solder (I use leaded shit, it's cheap and I'll die before it gives me cancer anyway).
But also check out some tutorials on how to care for your tips, they should be lasting quite a few months without much effort if you're properly cleaning and tinning them.
>making tips
Probably will work, but if you keep treating them the same as the ones you have you'll kill them just the same, possibly faster since those aren't purpose built for being used that way. Tips are cheap and a last a long time if you care for them, making your own is more effort than it's worth.

>>2765301
You forgot the minor detail in that the ATtiny10 doesn't have any EEPROM. Would you just use progmem instead? 10000 cycle writes is ok I guess, and you can spread a bit or two throughout a fair few bytes to lengthen that even more. But they're a pain to program. I'd rather use a tiny202, but that's since I already made myself a UPDI programmer, a tiny13 is more universally easy to program.

Can't believe Ben Heck made space invaders on a tiny10.

>>2765310
Personally I'd rather get one of those onewire TI TO-92 memory chips. But either way you'd need a microcontroller or something to read/write the memory.
It would be an option to make a nonvolatile memory cell with a pair of transistors and a magnetic core, but unless you can find the right powder and epoxy together your own tiny magnetic core, you'll probably find it larger than a latching relay.

>>2765404
That lead-free solder is significantly worse for tip longevity, and for two reasons. Firstly, it requires a higher temperature to melt, and secondly it just dissolves your tip more readily. That's why you should buy tips/irons/stations that specifically say they're designed for lead-free soldering. No-clean flux, idk it's probably fine for the tip, but for hand-soldering there's not really a reason not to use rosin.
Stainless steel is both bad at wetting solder, and bad at thermal conductivity. It would make for terrible soldering iron tips. Proper tips are solid copper underneath since it conducts heat well, with a coating of iron or nickel or something special that wets decently well without dissolving.

>>2765435
>>2765439
>>2765441
Thanks, I'll try to find some leaded solder if they even sell it anymore, the one I was using is pretty common around here so it made so it made no sense for it to be that destructive, nobody would use it.
As for stainless, I think the poor heat conduction can be mitigated with using a smaller wire as a tip inside a more powerful iron, probably wouldn't be as easy to use but tips would be cheaper

>>2765441
Yeah, overlooked that.
Theoretically you could piggyback an AT21 onto the 10, since they both have the SOT23 outline. Both have GND on pin 2 and the AT21 SI/O line lies on pin 4, which is one of the I/O pins on the 10. Just snap off pin 6 on the AT21.

What are some good and easy to use circuit simulators that include all sorts of ICs?

>>2765458
Seconded. There is like over a dozen different ICs out there now, I need a simulator that has every one.

For driving piezoelectronic tranducers for a ultrassonic cleaner, do I need a square wave signal or a sine wave?

>the unit is called an Ohm
>the symbol is an omega
it shouldn't have taken me this long to get it

>>2765490
How long did it take?

>>2765458
You'd be best of just figuring out a good workflow for downloading and installing spice files for ICs you want. I use LTspice for analogue stuff, where it doesn't particularly matter which op-amp you use (within reason), and where you have the ability to use a .wav file as a signal source. For digital stuff I'll use Logisim Evolution, or the idealised logic blocks in LTspice.

do you tent your vias? both sides?

>>2765526
he said tent

>>2765526
Yes.

>>2765506
Seen a lot of mongoloids but she a cute.
Funnily enough, a similar picture was in the brochure where I started my apprenticeship (German airforce)

>>2765509
Is KiCad a good option?

>>2765549
KiCad is an EDA not a simulator. It may have a built in simulator but it is not its main purpose.

>>2765549
Never used it for simulation, but apparently the simulation aspect has been getting better in KiCAD 8. There are release candidates for it already out, you could give it a try, but I don't think it's as quick to use as LTspice. It has a wide variety of parts in its default libraries, not sure how many come with spice models.

>>2765509
Where can I find splice files for ICs?

>>2765585
From the manufacturer's own website if they aren't just being assholes to sell you their own validation software (or want to force you to use their excel spreadsheet calculator). Octopart and Digi-Key and such may also have them. Popular chips may have third party models made and posted on forums and the like, usually only to make them more realistic (e.g. making the 555 output voltage range not just the full voltage rails).

>>2765458
Keysight ADS, and it's free if you know where to download it from

I have one of those inline fans that I use with my grow tent. I want to control it with a microcontroller instead of the pot that comes with these. My question is if I can just replace the pot output with a pwm signal from an arduino or if this is one of those traic things for ac voltages that.

>>2765733
Since it doesn't go down to 0V, I'd guess it to be a variable output ac-to-dc smps, but there's a chance it's a 12V brick with a seperate PWM circuit or buck converter. You can probably replace the pot with some sort of MCU circuit, but you'd have to reverse engineer it in the first place.

It would be easier to just use an MCU-driven PWM MOSFET (don't forget the freewheel diode) after a normal 12V brick.

I had a small aneng multimeter which could measure DC frequency correctly, sadly it's broken and all other multimeters I have can only measure frequency of AC. Does anyone have a reliable solution for this? I like my new UNI-T UT139C, it works great except for this, measuring Hz in DC is useful

>>2765745
>measuring Hz in DC is useful

DC, by definition, is 0 Hz.
if you wanna measure ripple on top of DC, you just add a cap in series, and measure ripple voltage and frequency on AC range.

I found this tiny stepper motor and it says it can be directly driven by microcontroller pins, is this actually the case? https://thepihut.com/products/18-micro-stepper-motor
> This 10mm micro stepper motor offers 18° per step and 2-phase 4-wire. No special driver board is required, the stepper motor can directly work with microcontrollers such as the Arduino, ESP32 and others. The motor has female headers ready to plug into the pin headers on your development board.
at 3.3V with 20ohms coil resistance that would be 165mA and 0.54W or is there some coil fuckery I'm neglecting?
either way I would expect that I need some kind of driver so if there is one does anyone know any?

>>2765778
>can directly work with microcontrollers

this is marketing speak
marketing's job is to get more sales by making it appear simple and wonderful
you job is to cry about it on the web

>>2765764
Ok, thanks
Is the capacitance important?

>>2765801
>important?

kinda
for accuracy, you choose a big enough value that the impedance of the cap is less than 1% of your meter's input impedance.

>>2765344
>>2765324
>>2765320
Thanks for the suggestions lads
I think I'll go for the nickel plated steel and try to make a stamp to press out the shape. Can't be that hard if the vietnamese can do it.
I would salvage but I may need lots of them

>>2765733
That control knob most likely sits on a potentiometer, which can't just be replaced with a PWM signal.
If you're deadset on PWM from an Arduino, your best bet would be either replacing the poti with a PWM-controlled DAC or to connect its shaft to a servo motor. They usually run on 5V PWM an can be controlled without a separate driver.
Small ones run you around 3 bucks on aliexpress, throw in a 12V/DC relais for shutting off the power supply completely and you're golden.

>>2765865
>but I may need lots of them
May make sense to 3D print a jig then instead of folding them all by hand.
Somewhat on-topic:
https://www.youtube.com/watch?v=BHODpAE0lC8
https://www.youtube.com/watch?v=bsGUJMZ4QbE

>>2765865
>>2765887
Maybe should've read your post first.
Don't judge me, I'm tired af, been up for like 48 hours.

The contactor for one of the motors in the lab failed today. I didn't have my tools but I had a tester screwdriver and it does get power and even turn on if I manually push the contacts in. I'm assuming the problem is the solenoid coil that burnt out. Is there any other possibility for contactor failures that match this condition?

>>2766001
>Is there any other possibility
The transformer that powers the contactor might be dead. Test for voltage on the low side when the motor is in a power-on state.

>>2766001
>any other possibility for contactor failures

some have overload protectors that cut power to the contactor coils.
you need to be reset the overload if it's been tripped.
in pic, terminals 95 and 96 would be in series with contactor coil, preventing it from operating.
except if you manually depress it, like a big dummy.

>>2766008
>Test for voltage on the low side
of the contactor.
Check the power switch/motor controller too.

>>2765865
You'd be better off going for hardened spring steel, as opposed to mild steel.

>>2766001
Could be a thermal fuse inside the windings burnt out. Or a fuse or something elsewhere, if you didn't test the coil for voltage.

I know this question might be extremely entry-level, but is it safe to connect a 10k ohm resistor with a 0.5W rating to a 12V battery?

tell me about op-amps
i tied + to ground yet it lights a led

>>2766203
P = V2/R
P = 122/10000
P = 0.0144W < 0.5W

>>2766224
It’s the difference between + and -
With… what?… a gain of 20000?

>>2766224
are you using a ground-sensing op-amp or rrio op-amp? at what voltage is the - input relative to ground? what are your voltage rails?

>>2766267
I tied both to ground and it lit pretty brightly
then i tied - to power with 1Mohm and it's sparking when i connect the battery but off otherwise
>>2766277
2 different opamps, a 741 and a ne5532. I have no idea what a ground sensing or rrio op-amp is. It's on a breadboard with a 9v

>>2766224
weird shit happens when you run them outside of spec. don't tie inputs to ground unless it's specifically rated for it

>>2766224
>i tied + to ground yet it lights a led

if you run an op-amp in open loop, the output is either gonna be high or low.
when it's low it'll light an LED tied high, and vice versa.
so one of the LEDs in pic is always gonna be on.
no mystery there.

>>2766285
the 741's input voltage range is ±13V nominal (±12V worst case) when the supply voltage is ±15V. In other words, you probably shouldn't run it with an input that's within 2 or 3 volts of its power rails. If you do, it will not behave as expected. Read the datasheet.

>>2766285
Hello, is this your circuit? >>2766318
If so you must be very careful about open-drain vs push/pull output types. In an open drain, both LEDs are turned on when the output is high while only the blue one is open when output is low. If you want only one or the other, you need a push/pull type output,

Can anon recommend a good desoldering gun/station?

I've been recommended to get a HAKKO FR-301.. Fine, but I'm resistant to spending $250+ on a soldering gun if there are decent alternatives, or possibly even something better. Please advise me.

>>2766551
Get solder wick and a hot air station instead.

>>2766555
Aside from being cheaper, why? I have like hundreds of caps from old electronics to remove

>>2766559
Then it makes more sense to get the FR-301 if you're doing repairs with lots of THTs. I used Aoyue 474A once and it was okay, but the Hakko feels better to use and is portable.

>>2766551
How much desoldering do you do? Even at my peak, the need for desoldering wasn’t that great that I needed a vacuum gun.

>>2766555
> use wick
Yeah, if you have to remove a 128-pin header or something like that, yeah, wick is your best bet. You have to wick it after you suck it half the time anyway.

>>2766568
I've only ever used wick, its all I've ever had. But now I've got several old computers and consoles that I want to get brand new caps into, and between how wick can be finicky and not wanting to damage the boards at all, I wanted to get a desolder gun after watching videos on them.

Only concern I really have is that I've heard that sometimes they can suck off the solder mask and damage boards, but I don't know if thats true or not.

>>2766318
>>2766488
Nah my circuit is just a led and a resistor into ground with positive facing the op-amp

>>2766573
And what exactly is the issue you're facing? Try to connect it from Vcc -> Resistor -> LED -> op amp output

>>2766575
i was just trying to figure out why it lights as much as it does when both inputs are tied to the same voltage. But i realize now that it's probably a spec issue as the other anon said

>>2766581
Ah yes, this could be due to the op amp internal designs being non-ideal, but it could also be due to noise in your system. Even when you tie both inputs to the same potential, you could have some noise that turns on/off the output very quickly, you'll need an oscilloscope to check. If your LED seems to flicker quickly (sometimes it flickers so quickly that you cannot see it with your eyes), that could be it. What you can do is tie one of the inputs to a voltage divider (say to 2V) and have the other input go from 0V to 5V or something. That way, whenever you want the LED on, you provide 5V and go way above the 2V threshold, and then back to 0V way below the 2V to turn off.

>>2766570
> replacing all caps
There’s lots of people (myself included) that don’t believe in this practice. If there are really bad caps when you test them, then replace them. I’ve got lots of caps from the 60s 70s and 80s that are still just fine.

The electrolytics I see fail are ones that were under spec’d and/or kept in a ridiculously hot enclosure; it’s really like 1 in hundreds you see fail in practice.

I also don’t believe that chinese industrial espionage story about the dielectric formula. Chinese caps are still bad.

The now-banned PCB oils made great, long-lasting caps back in the day. Just like leaded solder was great, and so was lead paint. All banned.

>>2766551
I bought a Yihua desoldering iron. Instead of a vacuum pump it just has a solder sucker built into it. A bit tedious, and the temperature control is just passive, but it works fine. There may be more upmarket chinky options with temperature control and/or vacuum pumping, you may even just be able to buy a hollow tip for your current iron with a hose barb on the side. Then you’d just need to buy a vacuum pump and build a trap for any gunk.

>>2766555
Hot air is for SMDs, hollow sucking soldering tips are for THTs. Hot air is less precise, and more likely to damage plastic components like connectors.

>>2766589
An addendum to this is there are known bad batches of caps. Xbox clock capacitors are a very well known case. I’d check the date and manufacturer and pre-emptively replace caps that are specifically likely to fail.

>>2766605
>Hot air is less precise, and more likely to damage plastic components like connectors.
If you know what you're doing it's easy. Some Kapton tape and aluminum tape/foil helps.

what's the pmos version of the A2SHB (chinks' favourite SOT-23 nmos)?

>>2766659
Embrace the AO3400/3401

I'm looking to build a Tesla coil. I will stand on the top load naked in an attempt to shower with only electricity.

What's the highest voltage I could try without emitting UV?

Are Uni-T oscilloscope probes any good? I want to get a good set of probes to go with my "new" oscilloscope instead of using random noisy junk probes that I have lying around.

Do CRT monitors actually hold a charge after you turn them off? I keep hearing the you need to discharge them every time you work on them but then I've also read that most CRTs are designed to automatically discharge to ground every time you turn them off, which is why they make that crackling noise you hear when you turn them off. So which one of these is true?

>>2766946
Both are true and the same goes for microwaves.
They are designed to do that, but do you really trust chang that much to be willing to put your arse on the line?

>>2766946
No, you need to discharge them. ALWAYS.

>wire up dim bulb tester
>works and doesnt burst into flames

>>2766958

cool, but i prefer the non-gay version.