/diy/ - Do It Yourself

>>2693432
> op-amp
Build your own op-amp

>>2693432
Rolling for tesla coil

Stupid question:
If I want to run a Rpi (Zero 2) with a standard 3.7/4.2 V LiPo battery as power source, do I need a booster (DC/DC boost converter) to 5v or will the Pi be happy with the lower voltage.

In addition to the Pi Zero, the battery would power a tiny 1in SPI lcd and audio output (no amp). According to data sheets, the screen and audio chip both take 3.3V as operating voltage.
As the requirements are pretty light, I would disable HDMI, Wifi, Bluetooth, and probably throttle or disable some cores.

>>2693449
>do I need a booster

raspi starts to complain if voltage drops below 4.7V or so.
a lil bit lower and it goes wonky.

>>2693449
A quick google search will tell you that
>The Pi is engineered to work at 5 volts, plus or minus 5% (4.75 - 5.25 volts). If you supply less voltage than required, the Pi won't power on.
So you'd need something like a BMC board, like one of those HATs that are made exactly to power up your stuff from a battery.

>>2693432
Does anybody know a reasonably cheap logic analyzer with a bandwidth of at least 100Mbs, which also works on linux (preferably with pulseview)? I'm looking at DSLogic analyzers rn and they're the cheapest I can find around. Stuff like saleae, digilent discovery scopes, picoscopes and so on are all too expensive; I'm looking to stay <300$ if possible.
I've taken a look at the used market but in this part of Europe nobody seems to sell shit

>>2693440
Inductors? Kinda strange, I prefer the few sound-au.com designs.

>>2693449
IIRC SoC itself runs on 3.3V but some of the peripherals want 5V. If you dig into the schematic you may find it's not going to damage anything if you feed 3.3V to everything, though you won't be able to use the 5V peripherals, and you may need to disable a DC-DC converter or two. You'd maybe be able to run it off an ultra-low-dropout regulator.

The professional method is to use a boost converter to 5V.

>>2693456
>>2693458
>>2693466
I was afraid so. From searching online, it seemed that as long as you don't need to use the components that need 5V (HDMI, USB), you're good. I'll get the booster then. Thanks.

>>2693420
to the guy from last thread
I dont quite understand what you mean by ASCII decoder, the way I understand it MDA ist still a normal line-by-line video signal, just TTL level (and with split syncs), which is why they can get away with a normal vertical deflection chip.
Do you mean what generates the video, like a terminal or computer?

Also, what is a good supplier for PIC32 chips in europe, ebay is way overpriced and MC and digikey want 20 bucks in shipping. (or should I ask that in MCG?)

I have a 18v panel thats only putting out 9 volts. The other 4 panels i got with it put out fine power.

Is this something i can fix in the panel? Obviously half of the magic loop isnt there. Or should i just find a 6v battery?

>>2693504
is it shaded? Sometimes a little bit is enough. Also could be that the panel itself is damaged, and check in the assembly/connector part that all the wiring is correct (just compare working with non working).

>>2693504
>Is this something i can fix in the panel?
If it's just a broken tab and it's accessible, then yes.

>>2693505

Its not on the array, its propped up against the jump i think or something in that pic, i had 5 of the renogy 205 watts but only put the 4 good ones up, pic was just for attention.
It prolly is shaded there, i was just done putting the stump vallast array together.

Ill open it up and see what we got, i didnt see any imperfections looking at the face, but im a scavenger not an electric pro.

Yeeeah i guess it looks like that got hot.
The silicone was annoying to chip out.

>>2693458
I’ve always thought that someone who needs a logic analyzer, is also someone who has the capabilities to build one.

I’ve built things on a small scale using discrete logic to look for things, or example, to detect and hold when two lines go high when they should never do that, or set up a one-shot to detect pulses that ate too long/short and use oscilloscope to calibrate it.

If you’re reverse engineering a chinese 5G cell phone tower from square one, and the parts have no markings, then you’re probably gonna need something fancier.

>>2693466
> inductors?
I think inductors are just hard to fabricate on an IC so they avoid those designs.

This detail in picrel is common as mud in a lot of amplifier designs. You often see the coil formed around the resistor. I’m guessing it has something to do with the fact that this thing can probably drive higher currents than a typical op-amp.

>put Altium Designer in my resume
>doing interview with 3 engineers and one HM
>one of them asks me how I got Altium Designer experience if I had to pay thousands for the license
>Me: "Uhh, I just got it online"
>"So... you pirated it?"
>Me: "Uhh, yes."
>"...We don't support that kind of stuff at our company. We have a very tight-knit relationship with Altium and want to foster it."
>rest of interview goes horribly

are you actually fucking kidding me
i'm going to kill myself

>>2693578
Should have just told them you had a student license.

>>2693578
You just got cucked by proprietary software. Should have listed KiCad instead. That's also probably not the reason why you didn't pass. Normies can smell autism from a mile away. A right kind of person would pass this interview with a firm hand shake alone. In an interview first few minutes determine whether you'll pass or not, subsequent questions are there just as an excuse for interviewers to waste time so that they don't have to go back to work.

>LG Gram
>Good laptop
>2 years later, FN key dies
>Linux sometimes think there is no ACPI
>Touchpad erratic

Re-soldered touchpad with leaded solder, cuz i saw crack there in solder joints...
Re-flowed EC/KBC/SuperIO or whatever you call it, because if you press on keys TYGH very hard it started working again, the FN key and rest of shit, and there is Microchip EC controller in that place, and from what I learned is that this black box is a bit more important than CPU since it tells powerlines when to start and shit, and also processes keyboard.
seems to work.

Welp. I guess I didn't buy hot air for nothing, saved my bacon.
But what flux should I use? Home-made rosin flux kinda sucks for BGA.

>>2693615
liquid rosin flux

>>2693618
That's basically what I have except I made it a thicker paste.
Problem is that it gets dark and I don't see stuff.
If you're making DIY projects with big SMDs or THTs, that's fine. But with BGA and super micro tiny components its a pain. Other thing thats kinda annoying is that alcohol boils off and can lift shit.

So I take back whatever I said about normal flux couple years ago, there is a reason why ppl use it. Yes, it is a worse flux in terms of removing oxides, but if board is clean, visibility might be more important

>>2693623
Are you just reflowing the chip or are you replacing it? If you're balling/reballing, then use RMA-218 flux paste.

>>2693626
>Are you just reflowing the chip
I did just reflow it, because I've no skills, nor stencil for reballing
>or are you replacing it
Im pretty sure all Microchip/SMSC EC controllers have firmware, and I don't have it.
>If you're balling/reballing, then use RMA-218 flux paste.
Is it safe to order shit from aliexpress?

>>2693628
>Is it safe to order shit from aliexpress?
Depends on the seller. It's chink ebay (caveat emptor).

>>2693637
I mean, will I get electronics flux, or zinc chloride shit that will short everything

>>2693646
I really don't know, Anon. Are there other suppliers that deliver to your area?

>>2693432
anyone ever try to run a normal lab scope (siglent sds1104x-e in my case) off a pure sine inverter? looking to do some field work and dont want to drop another $500+ on a good quality portable scope (micsig.)

>>2693649
Same aliexpress but locally with "u need it today huh?" tax.

>>2693534
I wouldn't have 100% capabilities but even if I had I'd lack the time to do so. Also what you're proposing is more akin to a custom test device than a full fledged LA. There's a plethora of cases that are not as extreme as the example you brought in which you have to decode protocols, possibly custom ones, at speeds higher than say 50-60Mhz and some custom built device that took a couple of weekends to make simply isn't going to cut it.
Not saying one needs a fancy LA worth thousand of bucks ofc

>>2693667
Where are you? Where I live there’s Element 14 and RS components with warehouses in the country, though their selection isn’t nearly as good as the American and UK warehouses, and they’re more expensive than Digi-Key.

>>2693749
I think I have RS, but they are cunts.
X is not shipped to my country because lithium battery, Y isnt shipped because fuck you, Z is not shipped because RoHS and I don't have RoHS in my country.
And there is a high chance they will just say "no" because it says "dangerous good" on the site.

Meanwhile Zheng will just scam. I suffer.

>>2693778
Well, I have amazon also, and if it doesn't contain undocumented lithium batteries shit usually arrives.
Why can't euros into trading like at all? Zheng is best at selling, but euros are the worst, followed by Koreans.

>>2693781
>>2693778
Btw, any other chink companies like LCSC?

>>2693500
Oh. I just did a quick skim through the wikipedia article and it looked like it could only display text characters. But maybe that was a feature of the MDA board and not the screen itself.

>a good supplier for PIC32 chips in europe
Buy enough that the shipping is free. I'm sure there's a bunch of stuff for various projects you want to buy. IIRC Arrow had better shipping thresholds than Digi-Key, but that was like 3 years ago. Check Mouser too.

>>2693578
Shoulda sperged out about how non-free software is immoral, would have been slightly preferable to coming off as amoral yourself.

>>2693785
Chinks mainly just sell to themselves, LCSC is the exception and not the rule.

>>2693646
just buy real solder, chinktrash really isn't much cheaper and will only waste your time

>>2693578
>playing the corpokike game
you get what you deserve

>>2693840
Question was about the flux.
Solder I can buy locally. Or aliexpress and gamble. Usually you look at the price. And soon I'd probably be able to buy it only from aliexpress, because solder I buy is european. And you know, they will ban it everywhere, would be only available for aerospace or whatever. Would be expensive as fuck, and chinks would keep on making good things, because they dont care.
Thing with flux is that I've no fucking idea how much proper flux should worth, nor who even makes it, as I've been using rosin as a flux whole my life, and now I just realized it sucks for stuff u barely see cuz it gets brown.

Thing is that zinc chloride is much cheaper than pine rosin, so im a little bit afraid of buying flux from china, since zinc chloride is corrosive and conductive, which is good for original purpose, soldering copper pipes and galvanized sheet, but sucks for electronics.

>>2693580
This

Or cover as kicad.

>>2693846
If you buy Aliexpress solder, buy Mechanic branded stuff. I bought some lead-free bismuth solder from them on LCSC a few years ago and loved it, but LCSC no longer sell tools and consumables. Since Mechanic is a Chinese brand, you should be able to find legit Mechanic stuff on Ali. Importantly, you can find decent spec sheets for their fluxes and solders.

If you do get zinc chloride, you'll know. It's water soluble, while most fluxes are only alcohol soluble. Even if you do get it on your PCB, with any luck you can just wash it off with water in an ultrasonic cleaner. Then wash in alcohol a few times to displace the water. But that might not be an option if it's stuck under a BGA.

>>2693846
>buying tree sap

I've got a component that simply WILL NOT take solder no matter what I do to it. I think the dumbfuck manufacturer chrome-plated the pins or something. Maybe they fucking teflon-coated them for all I know.

I just got a wire-wrapping tool; how good is a connection from just doing wire-wrapping? If I solder over the wire-wrap, would that guarantee conductivity?

It's a low power signal connection, 3.3V, 10-20 milliamps.

>>2693869
the pressure of the wire makes a cold weld with the pin
it's a pretty damn solid connection

>>2693881
Thanks, here's hoping. Do I need to do anything to keep the wire tight while wrapping?

>>2693881
>>2693886
I should maybe mention that the component isn't designed with a standard-sized square pin. It's close enough in size to a .025 pin that I can still wrap it, but that's it.

>>2693887
just use square wire

>>2693886
>to keep the wire tight while wrapping?

if you're doing it by hand, or on a smooth surface, you have to hold the wire back to create stretch.
if you have the correct tool, and corners on the pins, the tool will provide the stretch you need to make air-tight connections.
i've undone 20-yo WW connections where the outside of the wire has gone all black, but the inside is all shiny and chrome.

>>2693869
Should work well, assuming the coating is conductive.

I'm coming from a traditional tech and programming background, how hard is it to get into control engineering or learn? Can I do it on the job? Is it worth it bros? Seems stable and has a lot of opportunity in my area

is the atlas ESR70 the only "cheap" (under $200) handheld ESR meter that does >=100 KHz AND has some level of input protection? the de-5000 looked great until i saw it has 0 input protection and uses an impossible to get IC to do literally everything.

>>2693924
Make one yourself?

>>2693924
>pass known impedence sine source over cap to ground
>measure output voltage
wa la

>>2693961
>wa la
I hate to be that guy, but it's "voila".

>>2693961
yeah i've seen the function generator/oscilloscope solution but i dont want to dick around with it every time i want to measure ESR, especially in-circuit.

>>2693972
>I hate to be that guy, but it's "voila".

I wish I didn't have to be "that guy", but IT'S A FUCKING MEME NEWFAG

>>2693985
>IT'S A FUCKING MEME NEWFAG
Not a very popular meme.
>I was just pretending to be retarded.
lmao

>>2693859
Tree sap isn't rosin, it contains too much turpentines or whatever.
>>2693853
Well, LCSC is more-less reputable supplier, but Aliexpress is full of scumbags, so nothing stops them from selling fake mechanic.
>If you do get zinc chloride, you'll know. It's water soluble,
Pure zinc chloride is, but let's say plumbing flux, some are petroleum based and kinda behave like typical electronics flux if you try to clean them off with water.
Maybe PH measuring stick? Rosin flux should be neutral. And zinc chloride should be acidic

>>2693972
its wallah brohter.

>>2694026
Wallah ackbar

Am I delusional? Is it possible to take the displayport output and the charging port of a laptop and combine them into one USB-C power-in and dp-alt mode out?

>>2694063
>Am I delusional?
Maybe.
>Is it possible
Most likely yes.

>>2693889
>>2693892
>>2693893
Thanks folks. Also picrelated for >>2693889 :-)

>>2693972
Viola Voila, is that you?

>>2693985
So is the traditional response.

In this car wiring diagram, what does the dashed line between pin 1 and 2 of connector C234 in the Heater control assembly represent?

>>2694071
>freres and bros

>>2694084
It means you should ignore it.

>>2694084
I’m just guessing it means the shields are connected.
I would also say, that is one wire with two conductors and sharing a single shield.

>>2694118
I find it odd that the blower control switch would have an insulated lead, but it could be to avoid popping on the receiver when switching speeds. Who knows.

>>2694016
> rosin flux is neutral
It’s acidic at high temperatures. That’s why we use it.
Zinc chloride is more acidic at higher temperatures but still acidic at room temperature.

I think you can probably use zinc chloride on electronics soldering just fine as long as you clean it off. Even in plumbing applications you’re supposed to clean off any any extra ZnCl2 and there is a special procedure to flush the whole system in critical systems. It’s a PITA. This is less of a problem in plumbing because the acid “wears out” when it erodes oxides and the underlying metal, and in plumbing the copper is thick-ass compared to a copper trace.
Note that if you leave rosin flux on around a hot resistor that will keep the acidic nature of the flux activated and you’ll see similar effects to zinc chloride.
Leaving zinc chloride on the board would be akin to that shit that leaks out of alkaline cells. Ive had to fix dozens of things that were basically destroyed by that, so be careful and never use alkaline cells in your devices. I’ve never had ni-cd or ni-mh cells leak, they’re a bit more costly but there’re 1000x better in lifetime capacity and I’ve never had one leak anything.

>>2693924
For in-circuit measurement, some guy made his own circuit that specifically uses really low voltages so it won't interfere with anything, you might be able to find it.

>>2694122
Zinc chloride under a BGA or QFN sounds like hell to clean, even with an ultrasonic cleaner. I'd only use it with through-hole soldering. Maybe it's good for large things like XT60s, since it won't burn away like rosin does.

>>2694122
>It’s acidic at high temperatures. That’s why we use it.
Yep, and neutral-ish at room temperature.
>I think you can probably use zinc chloride on electronics soldering just fine as long as you clean it off.
I did that in a past and it worked fine, but problem is that you have to wash it off somehow.
>Note that if you leave rosin flux on around a hot resistor that will keep the acidic nature of the flux activated and you’ll see similar effects to zinc chloride.
Eh, I think it takes much more temperature to make it active, like 170C.
Idk.

>>2694145
> not for bga
Yeah, anything with connections under the chip you’re never gonna want to use zinc chloride unless you have lab-like facilities.

>>2694149
> resistors don’t get that hot
They sure do. In fact, that’s their function. There are tons of things out there that save 1¢ to under-spec the resistor wattage. I’ve seen many desolder themselves and fall out.

My furnace pcb is one, the FR2 board was dark brown around the resistor, which was shocking because they used the full lead length to raise it up a ½" off the pcb. Instead of getting proper resistor wattage. This is not what you want to see in your furnace. It turns out that it was fine, and within tolerance, but it’s probably lose to the solder melt point since the solder was turning dark gray.

>>2694197
>They sure do. In fact, that’s their function. There are tons of things out there that save 1¢ to under-spec the resistor wattage. I’ve seen many desolder themselves and fall out.
Wow, so its not just me who used 1/2W resistor for 2W lol
>My furnace pcb is one, the FR2 board was dark brown around the resistor
It can get brown at lower temps however.
>but it’s probably lose to the solder melt point since the solder was turning dark gray.
Solder gets black due to oxidation, which happens faster at higher temps.
Idk, still dont think rosin flux would corrode it.

want to buy an ffc breakout board. can I just get a 60 pin and use it for anything under 60 pins? or do ffc tend not to self align in a larger connector upon insertion?

>>2694283
>can I just get a 60 pin and use it for anything under 60 pins?
Sorta.
As long as pitch matches, you can do it

>>2694283
Use a 60p ribbon with a 60p connector and ground out the unused pins or leave them floating.

>>2694063
For a start, you could buy a usb hub that fits your use case, and open it up. Or search for open schematics of usb hubs.

No transformer I ever wounded ever output anything on the secondary
How do I git gud?

>>2694794
>ever output anything

that's impossible.
- your measuring tool is malfunctioning
- you're probing the wrong wires, like maybe one probe on primary and another on secondary.
- your meter is on DC volts
- you forgot to remove the enamel insulation on the tips of the wires

recommendation: use your body's natural ability to measure voltages.
for high voltages, grab wires between two hands.
for lower voltages (less than 100V), use your tongue.

>>2694801

also
- you're feeding the x-former (that's how the cool kids spell it) a DC voltage

>>2694802
>>2694801
I think I just need more turns
Why is it so tiresome to wind transformers? Fucking 0.1mm wire doesn't help either

>>2694794
What's the input frequency and voltage? Waveform too I guess.

The higher frequency you go, the less inductance you need for it to act as an effective transformer. I'd advise calculating or measuring the inductance of the input side, with the inductance you can calculate the reactance at your frequency, and hence if the no-load current through the primary will be too high for your source. You want the voltage drop across the winding resistance to be significantly lower than the voltage drop across the inductive reactance.

pic rel simplified version of what i would like to do. The voltage divider is obvious, the more switches i put on simultaneously, the dimmer everything gets. Question is to i "just" put a voltage on every lane or is there something smarter i can do? I'm thinking how PSUs have multiple lanes as well.

>>2694871

add 2 silicon diodes in series with battery
they'll drop voltage by 1.3V
there'll be a very slight dimming as more lites are lit
to avoid that, use 6 diodes, 2 in series with each lite

I studied that a capacitor works by creating total induction of an electric charge from a conductor to another conductor (like two parallel planes, a sphere inside a hollow sphere, etc)
In practice, in real capacitors like the pic I posted what shape do the conductors inside the capacitor have? I don't understand how they really are on the inside and can't find any source that explains it that isn't a picture of two parallel planes

>>2694922
>in real capacitors like the pic I posted what shape do the conductors inside the capacitor have?
You can always take one apart and see for yourself. In case of that black tubular style capacitor you'll find a scroll of 2 metallic strips separated by some insulator. Everything will also be moist because of electrolyte. Ceramic one is even simpler consisting of a sandwich of metal/ceramic layers. It's all just parallel plates, the real magic is in materials.

>>2694922
Either multiple plates stacked on top of each other or two long strips rolled into a cylinder. At least these are the ones I know of.

>>2694871
Meant to ask do I "just" put a voltage regulator on every lane. Dunno why I dropped so many words.

>>2694879
I don't really get how this is supposed to work, but I'll try later.

In which kind of applications do signal relays find use? Can those be driven directly from micro digital outputs or do they need some driver IC?

>>2694944
>applications

there are 100's of applications. most popular two are:
- current multiplier: a tiny mercury thermostat can use a relay turn on/off a 2000W heater.
- galvanic isolation: select one of two sound inputs without sharing a ground. prevents ground loops in concerts.

>need driver IC

depends on size.
very small ones can be driven from a micro-controller. around 20mA.
bigger ones need a transistor. up to 1A.
even bigger will need a power transistor, or an intermediate relay. up to 20A.

>>2694936
>I'll try later.

you need to test it with an actual load.
maybe a 100-ohm resistor.
if you just use your meter, it'll look like it's not working.

>>2694936
I’d just use a single voltage regulator. Ensure whatever regulator you pick can still output 2.5V even with the minimum battery voltage at its input. If it’s a lithium ion, then a cutoff voltage of 3.2 is sensible, meaning a dropout voltage of no more than 0.7V. Maybe an AMS1117 can do that, but an LM317 definitely can’t. The HT7xxx series are good for really low dropout voltages.

>>2694944
Swapping around range resistors in an auto-ranging ADC, same for various R,L,C components for an LCR meter.

Anyone know how these Anderson Powerpole 45A connectors work? My instinct was to have the wire's insulation ggo into the big V-shamped crimp and the stripped conductor go into the little one, but videos I've seen on youtube have the stripped conductor in the big V. I can't tell what the videos are using the little one for.

>>2695050
The stripped end of the wire goes in the big "V" and the little tabs crimp the insulation.

>>2695050
Keep your crimp hand strong.

>>2695052
What? The wire comes in from the right in my picture with the some amount stripped on the left

>>2695054

>>2695057
These aren't 45A connectors. They don't have the little tabs

>>2694963
>actual load
I used some LEDs from my starter kit. I don't really get how it works, but 1 to 2 LEDs are perfect, third one is noticeably dimmer, but good enough desu.

>>2695030
So LDOs can keep output voltage steady over multiple loads as well? I thought i'd end with something like pic rel with 1.66V at each diode opposed to 3.3V.

>>2695089
try this and mind the voltage drop across the LEDs and Schottky.

>>2695050
I think the wire is meant to be thick enough not to need tug resistance, and the planting housing around the connector provides the bending resistance.

>>2695089
Yea they can keep the output steady. Loads in parallel don’t split voltage.
The loads are incandescent lamps, right?

>>2695093
AMS1117s already have an integrated anti-parallel diode.
Also there’s a 2.5V one last I remember so no need for an adjustable one.

>>2695097
>incandescent lamps
LEDs and maybe more, but always the same load as in between switch and ground.
>Loads in parallel don’t split voltage.
Ahh, then excuse my confusion, please. From what I had left in my mind it was them getting split. I'll probably get best over the basics again first. Thanks for the pointer on AMS1117.

>>2695089
>1 to 2 LEDs are perfect, third one is noticeably dimmer

this is normal as each LED has diff characteristics. even if same model number.
when you parallel them, the aggressive ones suck in more current than the shy ones.

anyway, if your load is actually LEDs, then your entire approach is wrong.
you dont control for the voltage, you control for the current using series resistors.
try using this site to calculate your resistors: https://www.amplifiedparts.com/tech-articles/led-parallel-series-calculator

>>2695103
>LEDs
Those need a constant current, not a constant voltage. At the very least a constant voltage with a current limiting resistor dropping at least ~15% of the LED's forward voltage.

Now that I can access my datasheet database I can see that the AMS1117's minimum dropout voltage is 1.1V typical, 1.3V maximum. That's measured at 800mA, but I doubt you'd get much below 1V even at 20mA. I'd probably want to choose a different chip, especially if you're SMD averse.

The HT7325 has an excellent dropout voltage of 0.11V at 2.5V output, and comes in a TO-92 package.

>>2695165
I was assuming he was putting them in parallel, and the lack of brightness with 3 was a result of his low number of secondary turns.

>>2695168
>I was assuming he was putting them in parallel
Meant to write series. But if the transformer winding and LEDs on switches are the same project, then I was wrong about them in series, and wrong about the dropout mattering. Well dropout might still matter if the transformer voltage fluctuates a lot.

Anyhow, if they are actually well matched LEDs in parallel, and adding more causes the voltage to sag, then there's really fuck-all power capability being delivered by the transformer.

Instead of LEDs, I'd test it using a length of nichrome wire, or a resistor subsitution wheel (fun project if you have a bunch of rotary switches lying about) or wires in a bucket of salt water.

I have a 3v micro controller esp32 and i need to connect 5V to one of its GPIO pins.
But because of reasons i cannot do a voltage divider, so what if i do
5V --- 100k resistor --- GPIO
Would that work?

>>2695203
The digital inputs of an ESP32 should be 5V tolerant.

>>2695205
they most definitely aren't

>>2695206
That's not what Espressif says:
https://voodoo.business/2021/05/19/are-the-esp32-and-esp8266-5v-tolerant-yes-they-officially-are/

>>2695207
the people who made esp32 say it's not
but at the same time they say it should probably be fine if you slap like a 10k in series just don't do this in like any devices that will be operating life support in hospitals etc

>>2695208
All that American ESP janny says there is that this module, that has an ESP32 on, doesn't do any level shifting. And that the ESP32 datasheet does not say that they are 5V tolerant (why that is is explained in the link)

The link literally quotes the CEO and founder of Espressif, Teo Swee Ann (can't post the direct link to his post because 4chan thinks it's spam).
They are _most definitely_ 5V tolerant.

>>2695203
Yeah the internal protection diodes will eat that up. It's not advised, but it won't damage anything.

just don't forget that if you plug 5v into the adc pin you will only be able to measure half of it because the adc will just spit max value for anything over rail voltage

>>2695203
just use a fucking transistor you sperg

You guys worked with Li-ion batteries before?

I was planning on building a sort of "modular" battery, due to airline limitations. Basically it's comes in 5S "blocks" that snap in series to form a 15S row, which then snaps with another row to make a 2P. Total of 15S2P.
I checked with the airline, they're fine with that so long as each block can be seperated and is below 100Whr...

In terms of Thermal control and voltage balance.
Each block has it's own BMS, which then connects to a "Mother" BMS keeping each block similarly charged with a thermal cutoff. If any block gets too hot the whole thing shuts off.
To prevent that from happening, cells are mounted directly to a heatsink in the electronic box through Cho-Therm (or other electrically insulated thermally conductive material)

It sounds good to me but I'm still putting together a shopping list of materials and judging feasability/safety...
I considered LiFePO4 batteries but they're significantly less energy dense and I can't find any reliable suppliers, OTOH I found a supllier for Samsung 21700 Li-ion with decently high current ratings.

I am a complete noob. Is it possible to convert a DSP circuit like this into a C computer program? Can it be done pretty easily?

>>2695249
https://www.native-instruments.com/en/products/komplete/guitar/guitar-rig-7-pro/

https://www.ikmultimedia.com/products/amplitube5/

>>2695249
If you know basic algebra and can analyze circuits then it should be doable. You need to find a mathematical function that gives you output for given input. The tricky thing here is modeling transistors as nonlinear devices so you need to get familiar with ebers-moll model. From schematic you can see that you can split circuit into 2 parts: nonlinear transistors and output linear filter. Output part is easy, just find a transfer function and use bilinear transform to convert to z-domain, which is just a simple digital structure with sample delay elements. Transistor part is harder if you want to fully model nonlinear effects. You could probably split circuit in nonlinear amplifier and linear feedback path where PINCH input determines DC operating point. It will take some math, but you can start by playing with matlab/octave/scilab to plot responses, spice to study original circuit or pen and paper to assemble your model.

>>2695244
>Each block has it's own BMS,
That's called a string controller.
> which then connects to a "Mother" BMS keeping each block similarly charged with a thermal cutoff.
Sounds alright
> judging feasibility/safety.
There's really no way to do this at home. You'll need to do a UN38.3 test which costs $$$ and iirc is destructive so you'll need two. What's the purpose of this home build?
> t work in battery manufacturing

>>2695244
That pic is mostly related to e-bike/e-scooter battery packs where vibrations cause all sort of shit.
>I checked with the airline
They are not cool with anything DIY. Your best bet is modifying factory power banks, or at least getting powerbank case from aliexpress for 100W PD, since those use some series connection stuff.
It has to look like a factory device. TSA workers are dumb, and everything that looks DIY, you'd have to throw away or worse, you'd get v&. I fucking hate TSA can't wait to get enough money to buy my own plane so I can just put giant ass dick on all dumb restrictions they put. I will fly with water and batteries, dumb assholes. Fucking hate TSA

https://www.airwheel.net/home/Fit/2075
More u can figure out here.

>>2695249
You could build it, and study its net effects with an oscilloscope (and FFT) and try to build a model with different elements to achieve the results.

I’m guessing this is a guitar effect, so it might have harmonic distortion, clipping and various filter effects.

Never fear, even tube amplfier effects can be exactly matched with digital filters.

>>2695244
> airline, they're fine with that
That is unexpectedly enlightened of them
> each block can be separated
This is a good idea in vacuo.
A lot of battery failures start with a single cell, and end up causing a cascade failure.
Cells are welded because it’s the cheapest possible option, and easily roboticized. Same with packs not being modular… it’s just cheaper, and if you have an electric car, you get to pay tens of thousands for a while new pack once the bms can no longer mask out the bad cells well enough.

My dad is looking for an inexpensive, stand alone program for documenting the electrical wiring in his house. Apparently everything at the low end of the market is pretty much just line drawing software with some symbols while everything at the high end costs five figures or more and have annual fees. The house was built in 1901 and has had several additions. Before he bought the house, it went through at least two phases of being wired up with whatever was common at the time (he's guessing 1920s and 1950s). Dad has done his own electrical work in the decades he has owned the house, all of which was properly permitted and inspected (he's not an electrician but he's nerdy enough to read up on code and stick to it). Some of the old stuff was replaced but apparently the county doesn't require everything to be brought up to code like some places do.
I think what he's looking for is a way to document all of the circuits in the house, what kind of wiring is on each, all the branches, outlets, and loads so he can do "what-if" type scenarios for adding to the system or redoing parts of it. I don't know what he's willing to pay but I'd guess under $500 and for it to be a one time purchase.

Input offset voltage seems to be defined as "the differential DC voltage required between the inputs to make the output zero". Seen this many places. But isnt this true also for transition from low to high? So that the offset voltage is a "dead zone" both ways so to speak. For example if the positive input is at fix 5v reference and say the compararator has a Vos of 5 mV then the negative input would need to go above 5.005V for it to output low, but wouldnt it also need to go down below 4.995V for it to go high again?

>>2695442
It isn't a dead-zone, it's an offset. If your input offset voltage is 5mV, then the transition point will be when the inverting input is 5mV below the non-inverting input, regardless of which direction the output is travelling. It isn't hysteresis.

>>2695414
>program for documenting the electrical wiring in his house.

dumb idea.
it'll take 10x longer to learn the software than to do it by hand.
coz all EDA software is retarded and 20 years behind the times.

>>2695630
You know, it would be good to have an EDA software like KiCAD, where you design the actual electrical layout in the eeschema-like portion, and arrange it into the right shape on the pcb-like portion that also contains the room layout. Could even have seperate schematics for pneumatics, hvac, water, etc.

Might be able to make a KiCAD plugin for it.

>>2695249
>I am a complete noob. Is it possible to convert a DSP circuit like this into a C computer program? Can it be done pretty easily?
If you can do it, you are easily in the top 5% of programmers

What can produce such distortions on a grid 3phase cable ?
With some RF scanner, I survey it happening frequently at morning between 10 and noon.

>>2695658
What would the method be? I think I'd just analytically describe the circuit (>>2695272) and try to express that as a Z transform or two and a taylor polynomial. A lookup table for soft clipping would kinda work, but that pinch and wool would probably change the curve. Hence using a polynomial where you can change the values. Maybe a lookup table would also be required.

>>2695662
Idk, do smart meters send information back along the wires themselves? The fact that you're getting not only 100Hz and 50Hz but 25Hz sub-harmonics suggests it's not a conventional mains appliance at all. I don't think phase-fired control or half-cycle skipping could give these results, unless it's somehow skipping (half)cycles consistently across two mains cycles.

The 25Hz derived harmonics seem to have about the same strength as the 100Hz derived harmonics (which will be from things with bridge rectifiers feeding caps or maybe non-ohmic loads like mains LED strings) so it's hardly small. Looks like all the noise is relative to a single phase anyhow, so I'd guess it to be from something on a single-phase outlet and on the same phase as other appliances (especially big SMPSs).

All I can say is to try and measure this at different positions about the building, maybe specifically close to certain phases. With any luck it will be stronger near the source and weaker near things with X suppression caps.

Hi guys
A few years ago I assembled a guitar amplifier kit
I noticed that even with the standby switch disengaged (i.e. supply voltage to the cathodes is cut), it produces a lot of noise (high white noise floor + salt and pepper pops). It's a class A/B amplifier with 2x6L6 power tubes.
I haven't managed to find the schematics yet; going through those will be a priority. Until then, does anyone have any ideas what might be causing this? AFAIK the only thing that should get any power in standby are the cathode heating elements.

>>2695637
That's a breddy gud idea, Anon. It's the beginnings of a universal CAD application.

>>2695731
Reflow the tube sockets and any other joints that look bad, especially mechanical joints like the input jack(s) and volume/tone pots. Seems like a bad ground connection.

>>2695739
and wash the pots with contact cleaner

>>2695739
Good shout, thanks. I was a total beginner at the time, there's a definite possibility of bad solder joints.

>>2695742
Wash, as in remove from the circuit and submerge them, or just clean the contacts before resoldering?

>>2695743
You just spray the pots with contact cleaner and twist them back and forth several times to clean the wipers. I always spray them twice for good measure, and it helps to lubricate them afterward (or use cleaner with lubricant) so you don't have to clean them again for a long time.

>>2695743
Also, you don't need to remove them from the board.

>>2695746
>>2695747
Alright, thanks! I'll check the solder connections first, I haven't got any contact cleaner at the moment but I do have soldering equipment and it is a likely culprit. Dirty pots would just scratch when you fiddle with them, wouldn't they?

>>2695748
>Dirty pots would just scratch when you fiddle with them, wouldn't they?
Yes, but if you're repairing an old amp it's something you should do anyway.

>>2695750
Yeah, thanks for the heads-up. I'll do it as a maintenance task.
"Old" is a relative term, I built it maybe 10 years ago.

>>2695748
>I'll check the solder connections first
Use lots of flux, and none of that plumbing stuff either. lmao

>>2695751
No problem, Anon. 10 years is pretty young and the pots shouldn't be that dirty unless they were sitting in a box in your garage or attic. You'll know after you reflow the board whether you need contact cleaner or not.

I think i'm going to buy this.
I currenty have a 20 watt mastercraft iron, I like it because the tip lasts forever if you sand it a bit, but I have a dead gba battery (ez flash v 3 in 1 for dslite, can't fit a holder), thinking about buying a cheap battery and soldering the wire directly to the cell, if it explodes I will have a few backups cells to try again. I don't really care about longevity. I think a higher power iron would help prevent it from exploding.
The space requirements are very tight, original cell was CR 1220 I was looking deeply for a tiny lipo cell, but I can't find any except on that's on alibaba only. There was one I found on aliexpress but it was 2x larger (wont fit, even with sanding).
https://www.alibaba.com/product-detail/High-quality-lipo-0-9mm-Ultrathin_1600911870788.html
I also thought about using a apple airpod battery (it barely fits with 2mm), but it seems like soldering a connection will be tricky, and I bet it has a chip that prevents directly accessing the battery, and I don't see anything online about taking apart the cell so I bet the chip is inside the cell itself.

>>2696067
>I think i'm going to buy this.
forgot both the image and link and mentioning what it was...
https://www.aliexpress.com/item/1005006113401276.html

wot power supply should I use with my pinecil v2? picrel seems ok but they sell a GaN wall wort for $15 less

I blew up my liitokala charger/analyzer, and I want to build smth very similar, but with PC output.
I'm thinking, how would I achieve something like that?
First thing that comes to mind is to use a uC and couple op-amps to control charge current, discharge current. uC is just pulling up resistors of whatever bridge divider needed to get whatever voltage necessary for CC/CV.

But its scary to make your own power supply for lithium batteries...

>>2696098
Actually, I can probably use some sort of clock signal (or even VCO), and counter, which would be cheaper than uC, but no computer comm, which is secondary goal

Are there any services for pcb/schematic designs for small orders? Not printing, but actual design. I'm trying to tweak some existing files for a gamepad mod. Basically I just want to swap these three pin tact switches for two pins, as well as invert the directions of the inputs. So swap up and down, left and right. Unfortunately, I don't really have the time to learn an entirely new discipline, so I'm looking to outsourcing. Pic related is what I'm trying to tweak.

>>2696147
>swap up and down, left and right
Cut the traces and bodge them.

>>2696081
you probably already bought it, but I am the one buying the cheapo quicko iron for $10 (technically $15 CAD), it's probably a house fire, but the irons I have seen for $5 have the old tiny hakko tips, and long handle to tip distance.
But is there any reason to get a pinecil unless it's like $25?
I found a seguire s99 which is basically identical except it has no DC jack, and it uses JBC 245 tips (and there are other smaller JBC variants if you don't need the heat). But since I am CAD maybe you can find a pinecil for cheaper than the s99.
I'm not buying it though because it's more expensive and I also don't have a quick charger, but I do have a junker laptop power adapter.

>>2696067
>>2696068
I'd want to see if there's a halfway decent feedback loop in there, or just some shitty PWM. Maybe look for a teardown.

>>2696147
If that file is in a format that can be opened by an eeCAD program you have access to (e.g. KiCAD), then learning to modify it would take literally 20 minutes.

>>2696181
>But is there any reason to get a pinecil unless it's like $25?
It is, ignoring the PSU and shipping costs.
>S99
Hmm, a 150W C245 iron with hot-swappable tips. Looks really good to be honest. Could benefit from a ridge yo push your fingers up against though.

Still no portable 200W C245 iron, especially one with an XT60 directly on the back. I might buy a C245 controller (as seen in the Aixun T3A) and mount it on a milwaukee battery clip.

>>2696114
>>2696098
Looks fine for charger. but what about capacity measurement.
Need to think about the logic that would charge, discharge, reset the counter, etc.
>just fucking buy liitokala again
No. It has microcontroller, which is botnet

>>2696401
well, maybe some capacitors need to be added for stability of op-amps, idk.
But next part is basically making a VCO that is fed to the 4026 or 4033.

Would it be cheaper to use uC? yes.

>>2696404
Actually, why the fuck I have to invent a lithium battery charger, if TP4056 is cheap and everywhere. I can probably even put couple of them in parallel if needed.

>>2696407
Or use a TP4056 as the source signal for an op-amp that controls a high-power linear pass transistor. Or even to feed a switching regulator like a TL494.

I bought a fan system, which have a remote control potentiometer on a wire. The wire is finally not long enough to route it from the attics to the control room. I could just put a longer wire, with potential problems of length and loss. I also may bypass this cable, instead of the potentiometer, I'd be interested to put a microcontroller in it. With an ESP8266, I'm able to remote control its speed via wlan.
The power area needs to be open and put a 3.3V rail for the ESP. I think of using a buck converter, like https://www.aliexpress.com/item/4000836594835.html
But buying blindly can lead to dangerous hazard.. do you have some good and tiny buck converter to recommend ?

>>2696479
Loss isn't a big issue when the potentiometer probably has 4 orders of magnitude more resistance than the wire.

Consider just using a linear regulator if the power consumption is low enough (it probably will be).

>>2693432
Posted on ham but I'm asking here too. Do you guys have any idea of a jammer to fuck with the upstairs tv? I've got a deaf boomer for a neighbor. I doubt he's getting his TV through fiber optics, let's imply there's coax. As for the tv I have no ideea, it could be a modern LCD type.

>>2696533
probably cable tv if he's a boomer. look for the cable junction box inside. find which cable is his, unplug it, then plug it into a mains.

>>2696555
xDD

>>2696479
What's your input voltage?

>>2696418
>>2696407
It werks!
Current: 200mA
20 pulses.
Mom look no computer charger

>>2696693
And DPDT switch will change between discharge/charge mode.

>>2696288
>I'd want to see if there's a halfway decent feedback loop in there, or just some shitty PWM. Maybe look for a teardown.
I've thought about it, I think I am going to go with a Sequre S60P c210 with the IS tip (curved conical), seems like a really useful size, tip being a cm or so shorter, and the tip type.
I am going to power it with 30watts (what I have). It's probably going to be annoying for fat lead free joints (not a power upgrade from my mastercraft 25watt), but it just takes time and shoving a bunch of leaded solder and wicking it out.
I could also buy a used 65 watt usbc laptop charger for $20 cad (I could get a gan adapter for about the same, but I don't have an extra usbc cable).

>>2694871
That's not really a voltage divider. All lights with a closed switch will have the same voltage, no matter how many switches are closed.
What exactly are you trying to do? Am I taking it you're trying to power light bulbs with a lipo?

>>2695050
Seems to me they're just standoffs so that the contact doesn't move around.

>>2696068
For portable, I'd rather get a gas-powered one.
Got picrel, works great.

>>2696068
Get the Miniware TS101. It's the creme de la creme of chinky mini soldering irons. https://www.aliexpress.com/item/1005004804485790.html
But can probably find it local.

>>2696873
>$55
why would you ever buy that instead of a pinecil?

>>2696694
Okay. Is it OK to use a microcontroller as a dumb counter? Its just cheaper this way, esp with shit like idk, 8051 or Padauk whatever shit.
>>2696888
Why would you buy a pinecil, if you can get T12 op-amp based driver for way less without shitty firmwares. Its a fucking iron. Its job to get hot and maintain temp, now show hentai on the screen or whatever.
Botnet-free too.
>but muh USB PD
get activator card and use it with T12 based handle thingy. Zero fucks given

>>2696909
>Is it OK to use a microcontroller as a dumb counter
Absolutely, it allows for expanding the features later if you need it too.
>8051 or Padauk whatever shit
The hot new cheap MCU is the CH32V003.

>if you can get T12 op-amp based driver
Digital temperature control is something I'd trust a lot more than an op-amp based solution. Op-amp-based assumes the temperature/voltage curve is linear, and often doesn't do cold-junction compensation. You can't even be sure it's closed-loop in the first place, don't want to open-loop PWM a cartridge tip. With an MCU in there, you get lookup tables, sleep timers, thermal runaway detection, etc. for free. and can easily change the configuration yourself.
Anyhow, the comparison was with the TS101 which is basically identical but double the price.
>now show hentai on the screen or whatever
Yeah ok I'll get around to it, but I can't post that directly on a blue board.
>Botnet-free
The firmware is free as in freedom.

>activator card
A what?

>>2696915
>You can't even be sure it's closed-loop in the first place, don't want to open-loop PWM a cartridge
With open loop it would be glowing red hot at higher current.
>A what?
I meant negotiation IC, USB PD trigger.
>The hot new cheap MCU is the CH32V003.
I thought about this, but I'd have to buy the programmer for it. STC8 micros are prob most interesting for me as they are programmed via UART.
Idk. With micro I don't need all those opamp circuitry and I can integrate shit in software, not to mention that I can also calculate power easily, given I find micro with 2 ADCs.

>>2696693
Wait why did you make it MORE complicated? Take note that if you plan on leaving cells connected for more than an hour or two, you'll want to stop feeding them a constant 4.2V as that will degrade them. A low-current stage for when they're below 3V would be good too.

>>2696927
>USB PD trigger
Oh yeah. To be honest I'd rather use either the Pinecil or TS101 with the barrel jack, you can get 24V that way and hence more power. But that's coming from someone who doesn't have any USB C PSUs or power banks, USB C is definitely a good way of getting reasonably high-power for mobile soldering.

The problem with not using USB C is there's no negotiation for power level; without USB C you can't limit the iron to only draw 45W if you're on a 45W capable PSU. Which would be handy on a lot of barrel-jack PSUs, maybe the solution here is to just set the maximum power in software, plus use of brownout detection.

>I'd have to buy the programmer for it
That's what I was apprehensive about at first, but they make this neat starter kit as seen here:
>>2690812
>>2690906
The kit coming with a dev-board, a bunch of raw MCUs, and a programmer is something I'd like to see more of.
Personally I think having a cheap hardware programmer with good software support (e.g. STlink V2, USBasp, UART dongle) is perfectly fine. PICkits are cutting it close, you can get clones of the 2 pretty cheaply but not all PICs can be programmed with the earlier ones.

UART is nice, but does it need to use the DTR line to reset itself for programming like STM32s do? I've never really liked that feature, because not all UART implementations use those flow control pins. From what I see they only have RTS/CTS, assuming they use any flow control at all.
There's also live debug support to consider.

>>2696941
>Wait why did you make it MORE complicated?
Top part is just a power supply, bottom one is counter, impulse each 10 mAh, since I originally planned to use CD4026/4033 to count pulses, but for some reason those ICs are kinda rare and expensive, especially SMD.
> Take note that if you plan on leaving cells connected for more than an hour or two, you'll want to stop feeding them a constant 4.2V as that will degrade them.
Which is why in the end I thought about TP4056, as it does pretty much everything you expect from lithium charger.
>A low-current stage for when they're below 3V would be good too.
TP4056 does this, but I think it is a gimmick, because cells discharged below 2.5V apparently get lithium plated, and well, with high current you will know that something have happened faster
Possibly with audio-visual-olfactory alarm :D
>Personally I think having a cheap hardware programmer with good software support (e.g. STlink V2, USBasp, UART dongle) is perfectly fine.
WCH programmer is kinda like STLink. Actually, it probably can program STM32s too.
>UART is nice, but does it need to use the DTR line to reset itself for programming like STM32s do?
I don't think so. It just has tx, rx and power with ground.

But yeah, if I'm gonna use a microcontroller, I wonder what would be more precise, integrating in software, or in hardware with op-amps, where microcontroller works as a glorified counter IC.

>>2693432
Maybe you guys could help.
Not gunna go deep into detail on the project and why im doing this specific lighting configuration.
Im installing some can lights for ambient lighting and replacing a ceiling fan with a top down spotlight. It's 4/14 wire to the ceiling fan box run to 2 switches for standard fan/light setup. Im running the cans on b/w/gr and the spotlight on r/w/gr.
If I run a dimmer for the cans, will the effect the center spotlight ? Part of me says yes, the other says no. Id hate to buy an expensive dimmer just to find out it won't work as intended.

>>2697167
If you wire the dimmer between the switch and the light by breaking the red traveler wire (hot) then it should work (on the light only).

>>2697167
>>2697170
I'm retarded, nevermind. Yes, it will dim both the can lights and fan.

>>2697170
>>2697171
I haven't verified yet, but there may be 2 independent feeds to these switches with the 4/14 running to the fan box. If that's the case, the found and neutral wires will be attached for both circuits with independent hots. How will the neutral wires react to a dimmer on one circuit? I've got no clue how dimmers actually work and haven't had the chance to research yet ( :( wage slaving)

Long IC is looong

>>2697167
I don't get it, post a wiring digram.

>>2697202
Hope this helps. Hard to do those from a phone. we're assuming its a 2 line feed to the switches as diagramed. I wont be able to verify until I get home.
Im not sure if a 1 line feed in this application is applicable or not, but an answer for both scenarios would be cool.
I did learn that dimmers work with a triac and doesn't change V, I, or R, but limits the amount of time V and I are present through its cycle at the load. If im correct, the hot wire (black) is the only line to be effected by the triac, thus a dimmer on one switch should not effect the load on the other, but im still not certain.

>>2697226
If you install the dimmer on the right switch then the spotlight will be unaffected (assuming your wiring schematic is correct).

>>2697226
>Hope this helps.

for future reference, this is what your schematic should look like.
it has the advantage that you can draw it in a minute, and others can understand all of its complexity in just 0.2 seconds.

>>2697226
White lines are neutral, the black and red lines are lives and are each only used by one light or the other. So putting a dimmer switch each on the red and black wires should do what you want it to.

I'm trying to fix a lead-acid battery charger, taking it apart it's dead-simple. There's a simple step-down transformer followed by a rectifier, and on the negative side there's a series component. Pic related. After that the battery and panel meter are connected in parallel with this output. So I assume this series component is acting like a current limiting component of some sort, but considering it's reading 320kΩ I doubt it's working properly. I would guess it's a PTC thermistor that limits the current when the battery voltage is significantly less than the transformer voltage, but allows a decent current to keep flowing even when it gets close to the source voltage.

But I can't seem to find the part name. The markings embossed the front of the metal can are:
>323 (in small lettering)
>CD5
>82B (in small lettering)
And on the back:
>A1250 -or- A125D (in large lettering and actually printed in black, on the actual strip that's wrapped in the metal can)
>Control Devices Inc (in small lettering on the side of the metal can)
My closest guess is that it's a 125°C PTC thermistor, but that brand shows up no results, and I can't see anything in the same strange package. As I alluded to it looks kinda like it's a strip of something metallic wrapped in a metal can, and there's a plasticy layer between them.

>>2697247

best guess: self-resetting fuse

>>2697247
>>2697253

this one looks even go want to do look more like.

>>2697253
It's the size of a TO-220, looks like one too.
If it was a resetting fuse, I think if you connected a battery down at 9V to this ~14V PSU it would just trip immediately, and take ages to charge at a very low duty-cycle. Unless the effective impedance of the transformer prevents that. Also just connecting a rectified transformer output directly across a lead-acid battery feels really stupid, but maybe I'm wrong and this is common for lead-acids.

Does look like it might be a bimetallic strip, but it's not reading open-circuit but rather just a few hundred kΩ. Maybe it's a PTC being used as a resetting fuse, the thing says it's a 4A charger. I can easily get a hold of self-resetting thermal circuit breakers and 3.2, 3.7, and 5A PTC "fuses". Guess I'll try them out and see what works.

I may just recommend that the guy this belongs to gets a multi-stage charger instead.

>>2697256
Ah, that is a bimetallic strip, is it? I'll look into those.

>>2697239
I had a crude one drawn out but wasn't sure on how to depict 4 wire properly. I didnt have enough time to research proper symbols so I just scribbled on a ceiling fan diagram I had saved.
Showing the speciation between hots is the most important point I was trying to make.
>>2697242
Thanks for the help, everyone. My biggest concern was how neutrals being tied in together would effect the spotlight with no dimmer. If I thought it through all the way, I would have realized all your neutrals and grounds are tied in at the bus bar anyway and couldn't possibly effect anything.
Im headed back home to go play in the attic again. Might pick up a dimmer on the way.

>>2697258
Ah yep, definitely a bimetallic strip. Looks like it died from arcing too much.
I'll likely just solder in one of these:
https://www.jaycar.co.nz/4a-pcb-mount-thermal-circuit-breaker-spst/p/SF2284
Seems like the trip time should be fine. Maybe even a 3A will work.

Also my 4.5Ah 12V test SLA started popping its rubber vent caps after I left it on the charger for 10 minutes. This guy should really get a multi-stage charger.

>third attempt at a milliohm meter actually works
woo

My circuit has three ranges, 10mA, 100mA, and 1A. Any idea what sort of battery I should run it on? Currently, when it's not connected across something, there will be 20mA of quiescent current (power LED), so even at the 10A range it doesn't draw much power. For a safety feature, I should add a 1A diode to shunt it out to prevent excessive voltage when in-circuit probing, but that will make the quiescent current worse. So worst case scenario it may be experiencing 1A for ~1 minute timespans. I think it will run fine on two 18650s, but I wouldn't have a protection circuit. Should I run it off 4-6 series AAs instead?

>>2693432
I've got a small 370w ac motor that plugs directly into the mains. Any simple way I can control the rpms without having to buy a 200$ frequency converter?

>>2697380
>20mA of quiescent current (power LED)
I always preferred 1-2mA for my LEDs personally

>>2697506
>control the rpms

induction motor: belts, gears, clutches, variable frequency drives
universal motor: speed controller

>>2697518
Yeah it's an induction motor with different pulley sizes. Isn't there some schematic online that I could make from scratch?

>>2697523
arguably you could build yourself a variable frequency inverter

Anyone have any experience working with spring steel? I am planning on building a bunch of pushbuttons that use copper rivets as the contacts. Spring steel will give tension and be the conductive element. Would it also be a good idea to invest in nickel and gold electroplating solutions for a one-off project?

Working on animated Halloween decorations.
I want to amplify these things, using guitar / stereo amps and patch cords I have laying around.

Pic related is one of the circuit boards. Yellow to left is puting out 3W on 8 ohm speaker. IC2 controls it. I could probably pull unamplified power from that IC, but I don't trust my soldering / electronic skills enough to try it.

What's the best / safest way to pull signal off the amplified speaker? I'm not going for audiophile quality, just want something I can amplify over the voices of all the chattering neighborhood kids.

Once I get the circuits I'll add 1/4" mono jacks to run signal to the amps.

>>2697586
Nickel or tin plating is probably fine, maybe even copper. Nickel plating is pretty easy to do yourself, you just electrochemically dissolve nickel metal in vinegar and use that as your plating bath. Might be able to get away with no plating if you use dielectric grease or contact cleaning lubricant inside the switch. It’s also an option to try to passivate the spring-steel with a cold-bluing or an electrochemically grown oxide layer.

Maybe there’s a metallurgy thread?

>>2697647
Voltage divider, maybe with clipping diodes. Though I’d first trace the schematic to see if there’s an easy through-hole lead to solder a wire to.

You gonna blast the kids eardrums with feedback?

I got a dimmer, but I think I fucked up. Got a 3 way switch kit (pic rel). Im just looking for verification prior to getting started. My switch box is exactly wired up as previously diagramed >>2697226. The switches i bought are picrel.
Im assuming I'm going to wire the dimmer: black->black(source)/red->black(can feed)/gr->gr/blue->wire nut
and the other switch:
black->black(source)/red->red(spotlight feed/gr->gr/blue->wire nut
I'll follow up with a better pic of the 3way diagram provided with these switches

>>2697716

I'm not doing a good job removing that in order to get a good solder. Anyone have a method for that? I really don't have room for the long lead wires that are on these as they come from the factory.

>>2697724
One time at Vietnam war camp I used a Dremel on old fingernail clippers to mike tiny wire strippers. It's also how I lost my eye and got a bionic optical implant for WW3.

>>2697727
Shit, I should have just tried the .2mm stripper :)
Looks like it is peeling off the coating.
Appreciated, good luck with your eye!

>>2697729
You can use a lighter too, if you're into that kind of thing.

I ordered a bunch of NE5532 op amps. Need a circuit to bring a 0-5V input out to -10 to +10V using -12V and +12V as the supply to the chip. Tried using a 370K to 100K (4 to 1 basically) resistor divider from output to inverting input and get an 11 to 16V range... Basically the same 5V going in. Help?

>>2697753
*3.7K to 1K, sorry

>>2697754
The circuit I tried, pic rel

>>2697753
>>2697759
put it in differential mode with a flat 2.5V on the +ve input. then set your gain accordingly. been a while since ive messed around with analog shit so i could be wrong

>>2697766
Appreciate the advice! Got an example schematic to toss at me?

>>2697768
you can just do an internet search for differential opamp example

>>2697776
OK thank you!

>>2697768
threw this together, see if it works

>>2697828
remember thats inverting, you can do it non inverting but if i recall theres some tradeoffs. better then to throw the output into an inverting unity gain buffer if you want the signal polarity to match

>>2697828
>>2697840
I'm pretty sure you can just swap the 2.5V source and the input signal to make it non-inverting, with no downside. If you want to get the 2.5V without a voltage reference (e.g. LM336 or TL431) you might want to use a voltage divider, in which it's a lot easier to use the non-inverting input than the inverting input for that.

>>2697828
Nice, I'll definitely try this out!
>>2697847
Excellent, thanks

>>2697847
>2.5V reference via voltage divider
I might just use a crowbar zener regulator with resistor. Either way sounds good, thx..

>>2697847
i think that was one of those "depends on the part" things. for a ref voltage i used a divider with caps, bead, and tvs. seemed super stable when last i did that. could also just use a zener

>>2693432
anyone ever use dental loupes for smd soldering? i'm working with 0402 for the first time and they're pretty small. don't have the desk space for a stereoscopic microscope, unfortunately. these SMEDENT™ loupes are pretty cheap and the focal length is reasonable.

>>2697716
>Got a 3 way switch kit (pic rel)

i'm sensing massive confusion here.
attached diagram shows how a 3-way switch is normally used.
two switches are used to control one light.
you flick switch at the bottom of the stairs to turn on the stair light, then flick the other switch at the top to turn it off.
this bears no resemblance to what you need. so 3-way switches make no sense here.
still, you can use them like a normal switch by ignoring blue wire. just make sure it's isolated.

>I got a dimmer

so where is it?
is it also a 3-way dimmer?
if not, just wire its two wires exactly like you would wire one of the switches in >>2697239.

Existential question wiring a RCA Audio connector on this cable.
Would you put both lines into the core and connect the braid to the outer "ground" ?
or
neglecting the use of the braid, use one wire to the core, one wire to the outer "ground" ?

>>2697893
>221 / 63 / 72 / 1
each time I visit my dentist I watch his device with interest. He uses Petzl brand.

>>2697667
>Voltage divider
Can anyone validate this circuit is correct?
OK. So... existing circuit is 6V, guitar input is looking for ~1V. So, I need about 80% reduction in voltage. Which I should get from pic related. Right?
I'm not going to do anything sophisticated like cutting off the main speaker, or adding a Gain pot, since built in speaker volume will act as the "gain" in this case. The guitar amp will outcompete the 3W speaker for volume.
> You gonna blast the kids eardrums with feedback?
No mic, no feedback. The problem with these mechatronic displays is that the volume is plenty loud for inside, but outside you can't hear it. I've got guitar gear and long patch cords, so plan to use those for some amplification.
Also may build a couple audio amps to old speaker boxes; assume above will work for those as well.

>>2697667
>if there’s an easy through-hole lead to solder a wire to
There's not; whole thing is SMD which I don't really have the equipment to work on or modify. DIP-size stuff is smallest I can deal with. Also trying to avoid anything on board if possible; tying to main speaker is imho easier / safer for the electronics.

>>2697918
>Would you

the white and yellow are for two signals, likely L and R audio.
the braid and copper foil are connected together as a shield and as a ground.
if you have a mono signal, you can join yellow and white, or ignore one of them.

>>2697932
>validate this circuit

the voltage divider and output signal should be referenced to Audio- not to Gnd.
they're probably the same but not necessarily so.
(your battery symbol is upside down, BTW).
as for your voltage divider, it reduces signal by 20% not 80%.
flip them around for 82% reduction.

>>2697941
>flip them around, reference, battery
OK, so this, then. Was wondering about the ground as reference as well. And good catch on backward divider.
I've been drawing that battery symbol backwards my entire life. MechE, I only dabble in electronics.
Thanks for help.

This kid has to post here, right? Is there any other community who recycles lead acid batteries properly?

>>2697945
>OK, so this, then.

yep, all good.

>>2697945
>OK, so this, then

yep, all good.
except you can drop the resistor values by a factor of 10 or even 100, so the input impedance of the amp wont affect the ratio as much.

still, a pot or trimmer is always better than a fixed divider, coz right now you're just guessing what you may need.

>>2697946
its a facebook tier meme retard

>>2697932

>>2697964
Oh ok I don’t go on Facebook. Thanks for letting me know all the freshest Facebook memes since you spend so much time there.

please don't make fun of my autism

I want to create a speaker that plays songs from multiple sd cards like a pseudo radio with different "stations". I would like to have physical switches that, when pressed, will start playing music from that sd card (or folder, if it's possible to do it through software)

Is it possible? If so what's the best way to do it

>>2698048
I'll toss this in, for what it's worth. In my experience if an SD card playing an MP3 loses power it is much more likely to suffer file system damage than a usb stick doing the same function. I'd use usbs.

>>2698052
Yea yea I said SD just to mean removable storage, it's ok if it uses usb
The question basically boils down to:
Is there a way to have a physical switch control the swap of multiple usb drives? Sorta like a hdmi switch

>>2698048
>>2698054
These playback modules are sold for Arduino units, but they don't require Arduino to drive them.
You could stack up a dozen of these, with a dozen SD cards, and fire them up with a pushbutton. They can be configured (IIRC) to just start playing the entire card on input from one of the triggers.
Built in amp, power off 5V wall wart.
Have fun, and post a pic when you are done.

>>2698057
I'll check it out thanks
Trying to understand if it's worth buying a bunch of these over some shenanigans with a raspberry pi since I have a couple hanging around

>>2698048
>>2698057
Could this work? It costs like 3 dollars

>>2697963
OK. So, for an adjustable gain, it would look something like this, correct? This would give a range of 20-43% reduction in voltage. Assume if it's too high, I'll get clipping / distortion. Resistor values are just ballpark...
I've several 10K pots laying around for this sort of thing. I'll either swap resistors until it sounds right on a breadboard, or do pic related.

> you can drop the resistor values by a factor of 10 or even 100
So the input impedance of guitar amp is something like 1M. The multiplier value I picked for resistors was just taken from another circuit. What's the guidelines here... practical difference between using 470K or 47 ohm for Z1, for example?

>>2698057
Do these really contain a full firmware with mp3 decoder? Why even.

>>2697945
If it's a differential output amplifier, then that Audio- being used as a ground rail will be going up and down all over the place. This might be a noise issue, and will definitely be a problem if you want to use the same power source or otherwise have commoned grounds.

I'd just use the 0V rail as a reference, doesn't really matter that you're only sampling half the signal because both channels should be symmetrical anyhow.

>>2698057
Or just connect a single one of these to an mcu/arduino and tell it to swap to a different folder at the push of a button or twist of a rotary encoder.

Fill it with the GTA 5 or Fallout 4 radio tracks.

>>2698069
>20Ω
Do you mean 20kΩ for that top resistor?

>What's the guidelines here
Make the input impedance be significantly greater than the output impedance of the thing it's being fed by. In this case, your voltage divider impedance should be roughly the geometric average of the guitar amp's input impedance and the speaker amp's output impedance. Say that's 8Ω and 1MΩ, the mean is ~2.8k. You're probably fine staying within an order of magnitude on either side of that value (280Ω to 28kΩ). Go too low and you'll load down the speaker amp, go too high and you may start to get a noisier or weaker signal as seen by the guitar amp.

Though that assumes the input and output impedances of your divider are equal. In reality, if you have a 4.7k/1k divider, the input impedance is 5.7k and your output impedance is 1k, so you have a bit more leeway here.

>>2698080
>Do you mean 20kΩ for that top resistor?
Yes; that's a typo.
> geometric average of the guitar amp's input impedance and the speaker amp's output impedance.
OK, got it. Thanks for the overview.

>>2698073
IIRC from the datasheet, they'll play either WAV or MP3 formatted files. Basic chipset's meant for stuff like simple toys or those cards that sing when you open them.
I'm using one right now for a scream robot; it senses people, then turns and screams at them.

>>2698069
>for an adjustable gain, it would look something like this, correct?

i'd get rid of the 5K. it serves no purpose other than to set a minimum level.
if you make it zero, the minimum output level is 0%, which is what every volume pot should start at.
i'd also change the 20K to 0, but i have the courage of a minor god.
for the weak or timid, i'd suggest replacing 20K by 40K.
this gives a 5-to-1 division, so a 5V output will feed 1V to the amp.
(because to put 3W into an 8-ohm speaker requires 4.9Vrms)

>What's the guidelines here

rule of thumb is Z2 should be 10x smaller than input impedance.
this will create a 10% error from the original calculated ratio.
10% is an acceptable error, except if you're making condoms.
if you make it 100x smaller, then error is only 1%, which is good enough for condom manufacturers.

>>2698048
You could just put the "stations" on one card, then let an rPi decide which block of memory to read when an external input pin is toggled
This is a better question for /mcg/

>>2698109
So for those following along: This is what the final working circuit ended up looking like, after breadboarding for about 15 minutes.
10k pot had waaaaay too much gain. 1K seemed like a good range, and as suggested I added a gain pot so I can adjust it for different amps.

why do power supplies typically use type II/III compensation network configurations rather than PID? sorry for such a broad and dumb question, but im trying to learn the bare basics of analog control systems and power supplies.

>>2698376
because pid is software driven

>>2698384
i don't understand. you can absolutely build a PID control system with analog circuitry.

>>2698385
a lot of components for very little performance gain. also for everyone of those components chained together you get mad latency if youre not using purpose built parts.

>>2698376
Those compensation networks are PID controllers. Your pic related has proportional, integral and derivative term. It's the same thing just approached from a different angle. When dealing with power supplies it's easier to talk about compensation of loop response, because those things can be measured directly, rather than some PID gain parameters that don't map conveniently to a power supply problem. PID is a theoretical tool, compensators are practical realizations.

Can I charge a 2.7V supercap by PWMing a 5V input to an appropriate average voltage without damaging it?

I have a Wavetek 2001 Signal Generator, and it doesnt work. It's PSU keeps cooking off two electrolytic caps that are supposed to be 100mF and 50V, which I have replaced with 220mF 50V caps, which are still boiling. I have checked every diode, capacitor and transistor in the unit. The only weird thing I have found is that one of power transistors reads 200 to 300 instead of the usual 6-700 when checked as a diode. Any tips?

>>2698414
>PWM...an appropriate average voltage

this technique works fine with things like light bulbs and motors that have inertia.
but caps dont, so they're feeling every single slap of 5V violence on their delicate 2.7V-rated dielectric faces.

>>2698421
>Any tips?

pro-tip: dont buy instruments that use switching power supplies.
coz power-supply deaths account for about half of all modern unit failures.
and the reason my 1973 Heathkit oscilloscope is still going strong is coz it's powered by a heavy iron transformer.

what i'd do is figure out the voltages, and see if i can bodge an external supply to provide said voltages.
coz switchers cant be fixed except by experts in the arcane.
even if you replace every single component with an exact replacement, it wont work unless you also do the ritual incantation.

>>2698443
>but caps dont
Or do they? Basic property of capacitance is ability to resist changes in voltage. If you slap a capacitor with step input voltage across it will rise with (1-exp(-t/(RC)))*v_in. There should be nothing wrong with charging it with 5V as long as you stop charging before you hit 2.7V. PWMing input voltage would just make it charge slower. PSU will crap out before cap does, because it wouldn't be able to handle amps capacitor would pull and you would have to include series resistance to limit max current. If capacitor can provide 100A when discharging it also means it will happily eat 100A when charging.

>>2698414
Only if you use a resistor or inductor to limit the charge-speed. Otherwise you're both feeding it the full 5V for a short instant, and dumping a lot of current into it. A resistor will waste half your energy as heat, which an inductor won't. While you could just give it 54% duty-cycle and it should get to 2.7V eventually, it would be a lot faster to give it the full 5V initially and use feedback to prevent the voltage and current from going too high. An IC like the TL494 PWM controller already has two connected feedback loops exactly for this kind of situation. Ooops, just invented the buck converter.

>>2698421
Disconnect the PSU and measure its voltage before you pop anything else.

>>2698454
>implying resistor

>>2698528
>implying ideal voltage source with superconducting wires and ideal capacitor without parasitic resistance

>>2698206
So, when I moved this circuit from breadboard to soldered, I accidentally swapped the white and yellow wires (output to amp.) Doing this, the circuit didn't work... basically just made amplified 60 cycle hum out of the guitar amp. Swapped them back, worked again.
I don't understand why that would be the case. Reversing the inputs (green wires) made no difference. Since the entire thing's unreferenced (no grounds anywhere... mounting box is plastic) I don't get why polarity at the guitar amp would matter.
Pic of the actual circuit in non-working form, circuit diagram >>2698206

>>2697828
0-5V to -10v-+10v op amp guy here. I built the circuit I'm replying to, but it responds exactly as shown- I need a linear response so that the 20v swing tracks linearly with the 0-5v swing, so say you fed a sine wave centered on 2.5V it would be reproduced 4X larger on output?

>>2698619
You grounded the amplifier input to the pot housing, and grounded the pot output to the jack's shield. The ground reference would be the amplifier (through the jack and cable shield).

>>2698531
I dare you to find an equivalent series inductance spec for a supercap and calculate the required frequency for it not to explode. I calculated the current you’d end up with from the specified ESR, it was like 200A.

>>2698620
Right... the amp itself is grounded. Ofc. Thanks; will pay better attention to that next time.

>>2698623
>Right... the amp itself is grounded. Ofc. Thanks. I've got a couple more to build yet, so will keep better track of that.

>>2698393
>a lot of components
two op amps and a handful of resistors? how poor are you that you can't afford $1.50 worth of parts?

>>2698652
>$1.50 worth of parts
in my cuntry that is 3.5 trillion bolivars or 12 worm-less banana. only king have 12 banana. who need more than 12 banana?

>>2698620
idkwtf you're saying, looks pretty linear to me
revised to be non-inverting, and to use the more common 5V voltage reference

>>2698658
OK I must have boned the wiring somehow. The wire-up job I did outputs 1 or 19 volts roughly and transitions right at 2.5V input. I'll revise to this schematic and check for errors too, thanks again!

>>2698659
*-9v to +9,x, roughly. No doubt my bad or these ebay chips suck. Regardless will go to this revised layout.

>>2698660
What chip? It's pretty common for an op-amp to not get all the way to its rails.

As for the square wave output, be sure that you don't have positive feedback by accident, pic related.

I have an LED light that had the controller break. The whole thing ran off a 20V power supply. I tried a direct 100 ohm resistor to the led panel and it used 42 mA and measured at 16V on the led array. So the resistor is dropping 4 volts at .042A. Or .168W.
The light is less bright than before. Is there any harm in running it like this besides wearing out the LEDs from getting constant current instead of PWM?

>>2698451
It's from the late 70s, no switching required. Just pretty complicated cause it has to make a bunch of different but precise voltages for the oscillators.
>>2698528
Thats what I did, sadly it seems that in the unit itself some tiny electrolytics have gone bad as well. Resulting in a short between +18V and - 18V.

>>2698741
Arguably it's better than PWM, they should be running more efficiently with a constant current. That's assuming it's just a dumb LED panel with nothing but LEDs and maybe resistors.

Are there any power banks or batteries where you can read the current charge, for outside use?

I'm thinking of constructing a simple weather station, but don't want to run cables to it, trying to evaluate a solar and wind powered station.

Retard here, I have a few peltier modules and I want to control their power output somewhat. I don't need fine control, just having 3 or 4 power levels would be enough.
Since I'm powering everything from a PC power supply I was thinking of connecting between different rails to get different power levels. Like connecting between +12V and +5V to get 7V across the peltier.
The question I have is how resistant are PC power supplies to reverse current flow? I'm looking at several amps of current so I want to know if the PSU can handle that safely.

what's the most likely topology of this power supply? flyback/cascaded flybacks? https://sprucescience.com/products/lab-mate-high-voltage-power-supply-30kv-400microamp/

Did anyone try perverted shit like 7 segment display out of PCB?
Thing is that I wanted to use PCB as a light diffuser for custom indicator lamps, and I wanted to use 7 segment display because they are dirt cheap. But unfortunately there is a problem, 7 segment display is pretty tall, like idk, 6 mm or so, and it would limit viewing angles quite a bit.
So idk what to do, mill out a hole for 7 segment display and mount it flush with PCB, or try making shitty 7 segment display using 0402 LEDs soldered upside down.

After some experiments I figured that red is the best color, since it seems to scatter the least in FR4, but still I'm kinda worried about problems with either contrast/brightness, or bleed-through

>>2698831
Proper reverse mount LEDs actually exist.

>>2698838
No reverse mounted LEDs in Romania.
But yeah, how would 7 segment shit work? Given its diffused by FR4

>>2698831
https://www.amazon.com/Dim-It-Light-Dimming-Sheets/dp/B006R0VWSG/

>>2698842
It won't fix bleed through.
I kinda want smth like this, but 7 segment, and couple other random gimmick lights.

>>2698853
Maybe you could put drops of tinted epoxy on each LED.

>>2698876
and give each a light sanding for extra diffusion.

>>2698878
>>2698876
I'm worried it would be too diffused and light up inactive segments.
After all entire display is 3x2 cm more less

>>2698903
Edge-lit acrylic segments with the LEDs at the periphery?

>>2698716
Oh I'm not terribly worried about hitting exactly 10V off the 12, there is some ability in the PLC it will be feeding to be adjusted for reaction envelope amplitude plus I could pad down the voltages to the amp itself with diodes or adjust resistor values if the signal's too big. Using an NE5532 I think it's called, a low noise non-rail-to-rail audio chip. I figured it'd be alright for this task due to it's stability. I will be feeding this thing off an arduino doing PWM into a resistor-cap arrangement to produce the 5V control signal.

>>2698915
Thing is that I want custom display of sort. On the PCB.
And idk if it would work.
Well, simple words would work fine, but 7 segment is kinda tight and im afraid of bleedthrouhg which iwll turn 1 into 3 for ex, and 3 into 8.

>>2698790
Yeah that 5V rail isn't going to be able to sink current. You'd be much better off with a buck converter, that way you can continuously vary the power going into the peltier and implement a PID or even MPPT algorithm. Peltiers get inefficient if you shove TOO MUCH power through them.

>>2698927
I think the 5532 is a JFET op-amp, try a more common chip like the LM358.

>>2698968
I got like a hundred of the 5532, you don't think they'll be good for this?

I'm trying to use an OLED display with an arduino pro micro. The pro micro has a stable 5V output, but the documentation for the display says that it should be supplied with 2.5V ~ 4.7V. I think I could just use a 5V input and it won't matter but I tried to do it properly and got a voltage regulator (LM3940) to drop the voltage down to 3.3V. But, depending on the current drawn ( = the amount of lit pixels) the output voltage from the voltage regulator fluctuates between around 3.3V and 2.5V. And when the voltage is low, the screen dims, so it's annoying. When I'm using the 5V output directly its doesn't fluctuate. And I know the display works at 3.3V cos I've also used it with another MCU that has a 3.3V output.
What am I doing wrong here.

>>2698974
Read the datasheet. It says +/- 12V minimum output swing when given +/- 15V supply rails. +/- 13V output swing is listed as typical, but you shouldn’t design a circuit around that.

>>2698979
Did you read the regulator’s datasheet? It could be you’re drawing too much current. Also are your input signals still 5V? Or are you running the entire pro micro at 3.3V? Or are you using a level shifter

Hello, could you give me advice as to what to do with my amp. I have my parents Technics SU-X102 with the speakers and I'm trying to get it to work. I know that it worked fine when they stored it.
I connected my laptop, and I can hear the sound coming from inside the amp. I clean cut the cables (bare wire connections) and i tested the speakers with a battery, they make sound so they're probably fine.
I'm guessing the issue is around the audio outlet then, but I don't know anything about this stuff. Could someone give me an idea of what I should look for and what I could try.
Thank you

>>2698990
What are the implications though? Are you saying it won't faithfully reproduce the signal? That it won't attain + and - 10V when fed 24V? What makes this chip unsuitable?

>>2698998
Yeah, in other words the output will typically get only 2V away from the rails, worst case it will be as far as 3V away from the rails.

>>2698968
>buck converter
No idea how to control one with an arduino and I don't want to wait another 3 months for some beefy enough bucks from china.

>implement a PID or even MPPT algorithm
I'm brainlet when it comes to arduino programming so that's out of the question. I'll have to make do with relays, the code should be simple enough to implement even for me.

>>2698991
>give me an idea of what I should look for and what I could try.

- is muting indicator on?
- is your volume set to approx -40dB?
- have you selected the source that you actually connected wires to?
- have you tried various sources?
- is there sound from the headphone jack?
- have you stripped the end of the speaker wires before you plugged them in?
- are you sure your laptop is outputting sound? tested it with headphones?
- where did your cable come from? does it have 3 or 4 connection points on the 3.5mm side?
- if you hold a paper clip and jam into the center of any RCA input, do you hear a hum?

best guess is you're using a non-standard audio cable (from a Raspi or an old cell phone) which is causing you to short the input signal.

Would adding such a filter improve some noise I'm having on unrelated analog inputs of the controller, or is it better to have one for the controller too?

>>2699037
It may work, but be wary with LC filters. If you’re setting your motor to a 100% duty cycle and suddenly stop it, the drop in current will cause a voltage spike. You can calculate the voltage surge from the inductor into the motor driver’s input capacitance, to be sure it isn’t going to exceed any maximum voltage ratings. If it will, drop the L/C ratio or add a TVS diode.
I’d want to put similar filters on the MCU too, in fact I’d try filters on the MCU first. Less of a pulsed load, and more important as far as the ADC is concerned.

In addition, funky ground connections may be causing you trouble. I’d want to take differential ADC measurements to prevent common-mode noise issues from voltage drops on a common ground, or else use a sufficiently robust star grounding.
Good filtration on your analogue power rails in particular will shield your ADC from noise getting to it from the rest of the MCU.

>>2699003
Oh that's fine, I've got 2 volts head space on either side and if I only end up with +and- 9 instead of 10 that's within adjustment tolerance of the input I'm feeding. I think I'll be OK, but thanks for bringing it up!

>>2699096
Might need to drop the gain a bit so you don’t get clipping distortion, assuming the original signal’s voltage range won’t be changing.

>>2699167
There should be some leeway, the arduino's output gain will have software and, if I implement it, a gain pot for the output. So long as the circuit works "mostly" I can tweak the rest to fit around it.

I've an evil plan.
Buy 100 UV-C LEDs. Chinks sell ones that have no visible light produced.
Make a light fixture out of them.
Bring it to uni, turn it on.
Make people wonder why tf they have eyepain.

Should I do this? Should I become the reason why UV-C LEDs would be sold only for people with license?

>>2699011
>No idea how to control one with an arduino
Pic related. In this case you don't need voltage feedback, but I probably would do temperature feedback just by slapping a thermistor or whatever on the cold side and increasing or decreasing the duty-cycle to keep it roughly constant. That way instead of setting it to 25%, 50%, 75% or 100% power, you'd set it to 10°C, 5°C, 0°C, -5°C as an example. That way it wouldn't be influenced by ambient temperature as much. Though if you're doing something where thermal power matters more than absolute temperature you might prefer to stay open-loop, or even add a different feedback loop by calculating effective peltier power level. But open-loop (no feedback) is still perfectly usable.

>>2699185
You'd be an antisocial ass. Unless you were doing it for killing bacteria and viruses, then you're a covid hero (even though we've known since 2020 that the virus doesn't spread from countertops). If you use UVB LEDs then you can say you're just giving indoorsy people enough vitamin D to bolster their immune system (you'll be arrested for spreading malinformation).

I think Big Clive has a story about people putting mercury UVC lamps in a nightclub for their cool glow and giving people sunburnt eyes. UVC is nasty shit, man.

>>2699185
>should I commit a felony

Not sure if this is the right place to ask but I've got a 4 stroke that's fucked and I think it's an electrical problem.

Tested battery with volt meter, reads 20/200 so I'm assuming it's fine.
Repaired corroded battery leads.
Ignition switch sometimes does somthing, sometimes does nothing.
When it does something the starter motor spins up and engages the fly wheel, but doesn't seem to be able to turn the flywheel thorough a full engine cycle.

I can turn the flywheel through a full cycle manually, it's stiff at a certain point but it does turn through.

If I keep the ignition engaged when it can't turn the wheel it just goes *bzzzsssshhhh* which is not good.

/0/ doesn't seem to have a thread for actual mechanics

>>2699216
Pic relate, the offender in question.

>>2699216
>Repaired corroded battery leads.
>Ignition switch sometimes does somthing, sometimes does nothing.
It's all probably caked with grime or oxide. You probably should clean and test every single component with contacts. Does you millimeter have continuity function (beeps when you short leads together)? You can use that or if it doesn't have that just use resistance measurement and make sure any contact resistance is below 1 ohm.

>>2699196
Not first time.
>>2699194
>You'd be an antisocial ass. Unless you were doing it for killing bacteria and viruses,
Oh yea, ofc offcieer, it is to kill bacteria and virus. On eyes too. Ever seen floaters? UVC remove them :D
>If you use UVB LEDs then you can say you're just giving indoorsy people enough vitamin D to bolster their immune system (you'll be arrested for spreading malinformation).
I wonder if it would be enough, because output power is really small, like 42 mW for 4.5W input. (compare that with visible light LED where nearly 40-50% of energy converted into the light)
>UVC is nasty shit, man.
I know, I was smart enough to weld without eye protection (just cover arc with hand), but reflections still burnt the fuck out of the eye and it was a painful night that day.

But yeah, how many LEDs I'd need for this kek. Like hundred of them or so? given 4mW output

>>2699223
I don't even know how to isolation test individual components, or would be a fraught endeavour. The battery I can check, the resistance I can check.
Good ideas

>>2699185
>>2699228
I only nave experience with fluorescent UVC tubes, they do smell of ozone and require a quartz glass fixture to pass through.
Last time I checked legit ones powerful enough were pretty expensive, the chink ones are gonna be fake or very low power at best.

>>2699194
>>2699196
"Weapon or destructive device instructions" you fucking retards

>>2699070
Thanks for the input, I've added filters to both but it didn't make much of a difference, connecting PE as close as possible to the analog ground improved things a bit but I've solved it mostly by firmware increasing the number of averaged samples and adding a few previous measurements to smooth things out.

How do you guys deal w pic related? Thinking of getting an 8 pin box and just separating by decade.
>>2699223
You’re describing classic corroded terminal issue. Check terminals between battery and stater. I bet you’ve got somewhere with bad connection or acid rotted wire.

>>2699315
>separating by decade.

not efficient.
min you need is 12 boxes with 4 sections each.
each box is an E12 value (10, 12, 15, 18, 22, 27, 33, 39, 47, 56, 68, 82)
then each section holds 1 or 2 decades.

>>2698990
The datasheet for the regulator (LM3940) says "Iout (max) (A): 1" and further down "Iq (typ) (mA) 0.11". I'm not sure what Iq means.
The datasheet for the display (CH1115) says "maximum common sink current: 64mA; typical common sink current: 38.4mA".

I never thought about the inputs. It says 1.65V - 3.5V on the datasheet but I've just been using the signal directly, which I understand is 5V. Would just using resistors be fine to decrease the voltage here, because I assume voltage fluctuations won't matter, so long as it's within the range? Although, I don't know

The pro micro has to run at 5V because of the processor.
I'm not using a level shifter.

Actually, looking at the datasheet, theres this image of a "typical use". Do I need capacitors?

I don't have much experience with electronics aside from high school physics and a embedded system unit at uni.

>>2699337
Current status

>>2699253
its for erasing eproms i swear

>>2697941
>the white and yellow are for two signals, likely L and R audio.
Thanks to have confirmed my questionnings. I expected to receive cables with just one core and found this obvious. I put both into the plug and it's making super fine cables.

>>2699338
are you using a breadboard and jumpers or did you solder everything to perfboard? how long are your data and clock wires? what is the clock frequency?

>>2699345

>>2699353
Tried to find a datasheet for that specific module but only found another module with the same display. If there is a sot regulator on the module then you can power it directly from the 5V USB source. In that case the display is self-regulated to 3.3V and the I2C logic will be at 3.3V, so you need a level shifter between the display and arduino i/o.
The only other thing to watch out for is shitty breaboard connections and jumpers the diameter of a cat's whiskers, made from mystery metal.

>spend an hour soldering 0606 and 0603 components
>wire it up
>doesn't work

I have a computer mouse with a bright LED under the mouse wheel. I don't see a good way to disassemble and reassemble it and it's a cheap mouse so I thought about going through the gap between mouse wheel and chassis with a pair of tweezers and stab the LED. Would this pose a risk to the rest of the circuit inside. Don't want the mouse to start producing weird inputs

>>2699405
Welcome to the club, newguy.
>>2699410
>stab the LED
Post pics of your mouse. lmao

>>2699411
I already applied the treatment to the two mice and they seem to work fine
Roccat Burst Core

>>2699412

>>2699414
thx i'll clean it up some day

>>2699418
Get yourself a set of light, medium, and heavy guitar picks for shimming and prying delicate plastic cases apart. I like Jim Dunlop.

>>2699419
thx for the tip. i've been using an ikea membership card which i dissected in a couple of pieces but they're not rigid.

>>2699412
>they seem to work fine

in 2023, if your mouse or keyboard dont have RGB people gonna pity you like they do a poor shoe-less, orphan negro boy.

>>2699315
I have a clear file binder with each sheet split into 6 sections for storing particular values. But it doesn't scale that well. I think I'll go for a 3D printed mass storage system.

>>2699337
>I'm not sure what Iq means
Quiescent current, the amount of wasted current the regulator consumes that doesn't pass through to the load.

>I've just been using the signal directly
Well it's I2C, so presumably you're using the built-in I2C hardware on the pro micro. Did you enable pull-up resistors on the pro micro, or are there pull-up resistors somewhere else, like on the display board itself? If the pull-ups are on the display, then they'll be pulling up to 3.3V, so no problem there at all. If the pullups are on the MCU or otherwise going to 5V, then it's technically a problem, but if the display has protection diodes it should probably handle it anyhow.

>The pro micro has to run at 5V because of the processor.
Well there's nothing stopping you from feeding 3.3V to its 5V pin, but there may be some software settings like clock speed you need to change. I wish more dev-boards came with a switch to change voltage.

>Do I need capacitors
From the datasheet:
>The output capacitor is critical to maintaining regulator stability, and must meet the required conditions for both equivalent series resistance (ESR) and minimum amount of capacitance.
Looks like the regulator is rated as stable for low currents, but not without a capacitor. Chances are it's oscillating and causing the display to dim, you could check that if you had an oscope.

>>2699344
Actually maybe the two channels are for differential mono audio. Shove a balun in either end for even more noise immunity!

data sheet says to use an n-fet to PWM the input of this LED driver. sounds great but 0% duty cycle PWM (i.e. no MCU input) means 100% duty cycle on the driver. is there a way to do have 0% duty cycle PWM = 0% duty cycle on the driver?

>>2699497
BJT inverter, or just invert the MCU PWM output in software and put a pull-up resistor on the FET for startup.

>>2699253
Cuck. Snitches arent welcome

>>2699622
I generally prefer it to /b/ and /pol/ fuckers coming here. One of them asked if he could give himself an erection through electric shocks, and I told him it would be extremely painful. The retard didn’t even get the bane quote, fucking disgusting.

Interesting article about prolonging battery life:
https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries

>>2699695
I think this is kinda incomplete and dishonest a bit.
It is more correct to measure total energy delivered instead of cycles, and I don't think just limiting the voltage will solve the problem.

If you take a look at modern phones, they all charge cells to insanely high voltages, like 4.45V or 4.5V to squeeze that extra capacity, and yet batteries aren't dying any faster.
I think it has to do something with charging algorithm. I think it is no longer CC-CV. Other thing is that at higher temperatures lithium plating, main reason behind capacity loss and self-dischage, doesn't occur.
Or idk. It is really hard to find stuff about this, probably because most technologies are under nda.
>but tesla
Tesla has to keep voltage lower than 4.2 for regenbraking.

>>2699702
>It is more correct to measure total energy delivered instead of cycles
Well I think the point is that there's multiple compounding methods of wear. While the energy taken from the cell will be proportional to the amount of charge moved through it and hence will roughly correspond to things like dendrite formation and such, there's also other methods.

Interesting to hear that they use high voltages in cell phones though.

i kinda like designing strip-board prototypes, takes a little while but somehow it ends up less stressful than designing single-sided boards to etch
only one jumper wire on this one, which is pretty nice

>>2698979
Is that one of those 128x64 oled displays? I run those off nanos and always just use the 5v output from the nano.
No idea why your ps is doing that tho.

>>2699328
I like the idea of using e12 values; it’s easier to visually seperate by decade than e12. Usually when I’m designing I’m thinking “I need some resistor around 50k” then just grab the closest I can find, which means looking by decade vs e12. But if I’m going to sort I could do better.
>>2699450
Huh. Had not considered envelopes. I have those clear hobby cases for most stuff. Just ran across these tubes and may check them out.
I’ve looked at those 3D print versions; my thing is you can get clear hobby storage cases for like $5 at hobby lobby, and print would take a lot longer.

>>2699328 #
I like the idea of using e12 values; it’s easier to visually seperate by decade than e12. Usually when I’m designing I’m thinking “I need some resistor around 50k” then just grab the closest I can find, which means looking by decade vs e12. But if I’m going to sort I could do better.
>>2699450 #
Huh. Had not considered envelopes. I have those clear hobby cases for most stuff. Just ran across these tubes and may check them out.
I’ve looked at those 3D print versions; my thing is you can get clear hobby storage cases for like $5 at hobby lobby, and print would take a lot longer

>>2699735
What sw do you use for this?
Did a strip board for first time last week, manually moved from breadboard. Was a little trickier than I’d thought having not done one before.

>>2699411
>>2699412
it came back live and it's flashing red. sorry i won't do it again o_o

>>2699783
You could print a silkscreen layer on a piece of paper and tape it to the board, then push the THT leads through the paper, or use heat transfer paper to label the strip board.

>>2699787
If you don't want to desolder anything then open it up and use fingernail clippers to butcher the LED. Just try not to rip up the traces.

>>2699783
>What sw
kicad

>>2699725
>Well I think the point is that there's multiple compounding methods of wear.
Probably.
>Interesting to hear that they use high voltages in cell phones though.
I figured it out when relatively new phone did puff up (which was connected pretty much 24/7), but older phone (which still charges to 4.5V) still didn't.
I think there is more to wear than just charge current, cut off voltage, temperature, time being at 100% SoC, etc.

General consensus seems to be.
>charging at too low of temp - bad - lithium plating
>charging at high current - bad - lithium plating
>discharging to zero - bad
>high temperature - bad
>prolonged storage at 100% SoC - bad
>high discharge current - bad
So I think in order to get most out of battery, you should charge it with 0.25C or idk, less, discharge with less than max advertised discharge current. Probably charging to 4.10V instead of 4.5V helps too, and you will lose <10% capacity anyway.
Not gonna pretend being smart, this is what i found.

But you'd be surprised about real world battery damage... Shitty BMSes. Lack of balancing would kill any cell fast.

>>2699735
>>2699783
>stripboards
>breadboards
>etching
ditch all that garbage and solder point-to-point
use copper tape as a ground plane
you can thank me later

>>2699802
But yeah. Also I think 100% SoC is 4.5V in all these research papers, but idk
>High discharge current
Easy to avoid by proper battery pack design
>High charge current
Easy to avoid by charging slowly
>Low SoC
Easy to avoid with BMS
>High SoC
Easy to avoid with BMS and charging to lower voltage.
>Low temp
Useful for storage lol
>High temp
Useful for charging at fast currents, but sucks in everything else.


Idk. Complex subject. Most people can't drive, which is a 2D task, and this is much more D thing to understand

is there anything wrong or visibly retarded with this? i want to drive some RGB LEDs from an 18650 cell and an MCU providing PWM dimming, and i want to do this power efficiently as possible. i doubt i'll run anywhere near 100% output so my main concern is will i have any issues running these boost drivers at lower duty cycle %s?

>>2697178
>current year
>producing 56-pin IC in DIP
Lacking behind 30 years while still basically controlling the industry isn't normal.
But on TI it is.

>>2699809
another question i had, how do i size the balancing resistors? i know that one constraint is my maximum boost controller output voltage, and i think i'm fine (despite them not defining it explicitly.)

>AL3019A = 3.45V x 8 = ~28V maximum output
>Worst case is Max VF on G/B channel; 3.4V x 4 series = 13.6V
>Balancing resistors see 20 mA maximum; V = 20 mA * 100 Ohms = 2V
>13.6V VF + 2V resistor = 15.6V < ~28V maximum boost output

how do i calculate if my 100R balancing resistors are large enough to effectively control for differences in VF among the 4 parallel strings?

>>2699814
>how do i calculate if my 100R balancing resistors are large enough to effectively control for differences in VF among the 4 parallel strings?
Easiest way is to just plot I(V) curve for all LED corners and then pick acceptable current spread. Once you have the spread you can just draw a line through that spread which will give you required resistance.

Another option is to skip resistors and use BJT current mirrors instead. Simply replace resistor with NPN transistor and then use its vbe voltage to copy current to other strings.

one of my air pods died. it is not repairable but i am just curious what do you think could potentially die there? none of the well known revival techniques (forget, reset, clean etc) worked for me. it seems to be actually dead. all it does it emits some faint clicking noise. and it died suddenly after a few years of gentle use. no big loss but i am just curious what could kill them.

>>2699874
Some possibilities. A) battery gave up. B) speaker failed. C) skin oils/sweat infiltrated the thing and corroded/shorted the circuitry. And if it's an Apple product there's always D), they decided you'd enjoyed it long enough and it's built-in timer expired.

>>2699874
>i am just curious what could kill them.
Knowing apple and how they deal with their hardware it's probably designed to fail at some point.

>>2699874
malik retard

>>2699874
>suddenly
mRNA spikes in your earbuds

I made a schematic in kicad for a project I want to try out, but then I realized I didn't know how to simulate with kicad+ngspice. Throwing the whole hog into it was a big failure so I'm working with a simple toy resistor + LED circuit.
Still can't get it to work right. Diagram is pic related.
Spice just refuses to give me current values for the led or the resistor.
Also, I can't tune anything. No error messages, it just does nothing, nothing is added to the tune list no matter what I do.

What am I doing wrong here?

>>2699874
>(((air pods)))

>>2699939
Spice in kicad is a bit of an afterthought and a hacked solution. It assumes you already know your way around plain text-mode spice and understand how it works without any graphical interfaces. Just use ltspice/microcap/qucs/pspice/tinati/simetrix/pen and paper. Mixing simulation and pcb schematics is not a good practice anyways.

>>2699949
Extremely gay. Thanks for the input though, guess I'll look into that. Hopefully there's some non-sucky tool around so I don't have to p&p the whole damn thing.

>>2699949
>Mixing simulation and pcb schematics is not a good practice anyways.
Not him but that's a horrible situation. How am I supposed to validate my schematics?? If you build an analog circuit, you can simulate in LTSpice all you want but if you cannot easily import your schematics from LTSpice to your EDA software, you'd have to redraw it from scratch and make a bunch of mistakes and re-order your PCBs a few times until you get it right. This is the elephant in the room that nobody talks about.

>>2699783
Ki CAD. It's not designed for it, but with a bit of work it's a good way of organising your parts to be compact.

>>2699790
This is a very cool idea.

>>2699804
>not dremelling up a piece of copperclad for manhattan islands

>>2699809
Something tells me that PWMing the control lines isn't going to have a linear response. The boost IC will increase its duty-cycle immediately after being enabled again to make up for the lower feedback voltage. Instead you'll want to either feed an analogue voltage into the control pins via DAC or filtered PWM (not sure if the boost IC can handle that) or try to interfere with the feedback network directly. In this case I think it would have to involve external error amplifiers (differential op-amp circuits) reading the current from the LED strings and comparing them to voltage references, as set by DACs or filtered PWM, and feeding the output from those error amps into the FB pin. The gain would have to be rather low (10-100 or so) otherwise it might get unstable, but you can add low-frequency gain with capacitors (turning the error amp into a PI controller) in order for it to stably approach the desired set-point.

>>2699814
Pretty sure 100Ω balancing resistors should be fine, the differences between LEDs is largely not random, but a change as a function of temperature and operating current, which should be roughly shared between them.

>>2699973
You can export a netlist from KiCAD to LTSpice IIRC.

>>2699991
>You can export a netlist from KiCAD to LTSpice IIRC.
That's probably only good for passive elements. You may still run into problems when importing opamps and even transistors since the pinouts of your parts may not match the schematic in LTSpce. Altium can import LTSpice schematics but I ran into problems like these. Had to re-create a bunch of parts so it wasn't really worth it.

>>2699973
Do you think that they simulate everything in one file and with one tool when they build a 777 jetliner? No. In engineering we have this amazing tool called (black-box) abstraction. You are supposed to break down your system into small sub-components that adhere to your required specifications. You only simulate those small pieces of a more complex system. Simulating whole circuit is just asking for problems. Do you trust your models? Did you verify them? Do models include parameter variations? Did you include layout parasitics? Do you even know how layout affects your performance? What EDA software has is
>means of abstracting circuits (sheets, hierarchical sheets)
>ERC processor to catch obvious mistakes
Spice won't help you with symbols or footprints. What good is a simulation if you picked BEC instead of BCE symbol for your transistor. Some things you just have to verify manually. If you're having trouble with that then approach problem systematically and create a checklist of tasks to catch errors. Redrawing schematics shouldn't be a difficult task. If your schematic looks like a bunch of tangled up spaghetti you should reconsider if you can apply more abstractions to reduce complexity.