/diy/ - Do It Yourself

Anyone else can't stand tutorials made by Indians? I won't bother following it the moment I hear an Indian accent.

>>2611914
it's either that or a 1:43:00 MIT video on how to derive capacitance from maxwell's equations

Anyone able to recommend a basic-decent multimeter for someone looking to carry out basic repairs on household electronics?

Previously someone recommended UNI-T brand but I can see Amazon has one also.

>>2611934
>recommend a basic-decent multimeter

there should be a law that someone so uninformed that they cant choose their own DMM should not own a DMM, or engage in DMM-related activities.
for the safety of themselves and anyone unfortunate enough to be around them.

>>2611939
Aye and there should be a law that companies can't import poorly made shit from Alibaba for profit but that's the world we live in eh.

>>2611934
Uni-T is enough of a brand that I'd trust it with mains electricity. The same applies for some Aneng models (e.g. AN8008/8009 and their lineage), though their wanky modern designs like the Q1 are objectively worse than having a range-switch. I wouldn't really go for an "amazon recommended" DMM.

Personally I'd read through the EEVblog forum multimeter spreadsheet to get an idea of what's out there.

Clamp meters are probably not quite what I'd recommend for someone starting out, you can't just slap them across a normal power cable since you need to seperate out the phase and neutral wires (see that funky plug thing shown in the top image). Also only some clamp meters do DC (and do so with limited accuracy). You're probably better starting off with a conventional meter with its inline current measurement, and a wall plug power meter.

>>2611945
>poorly made shit from Alibaba
Gives people who can't afford anything else passable products, and provides competition to the actual brands to actually be worth what you pay for them. I see nothing wrong with flooding the market with cheap shit, so long as it's not difficult for the average consumer to understand what's good and what's shit. "Amazon recommended" products do blur that line a bit too much for my liking.

Henlo. Like a month ago I posted my dead monitor power supply and I finally got around to testing it. Recap: it wouldn't turn back on after unplugging it unless you waited for a while. It lost power for a long time and now doesn't turn on even if you leave it plugged in all day. I suspected capacitors because of the transient behaviour but they all looked fine visually.

Testing the power outputs of the board with the AC setting on my multimeter I got stable 11V for the +5V outputs, and 45V for the +16V outputs that gradually falls. Testing them on the DC mode I got almost exactly half of the value (5.6V and 21.7V) so I'm not sure if I'm being retarded here. Which mode am I supposed to use?

I've isolated the problems to pic related (which now that I look at it is faintly tinted a darker green but it doesn't come out in the pictures). Blue is all connected to ground, yellow is the low voltage rail, red is the high voltage, pink is reading as zero. Orange is unsure because it reads 0v but doesn't have continuity to ground, so it could be blue or pink.

Is the IC dead? The datasheet says VCC should be 3-26V, but I got 45V on AC and 0V on DC. But power seems to be reaching the output connections so I dunno.

>>2611957
Should also note that the other IC on the board is similar (getting some input voltage, but output is zero).

The fact the pink voltage gradually continuously drops as I measure it still makes me believe its a capacitor problem. I can't determine whether C930 or C925 are dead. It might even be another cap on the same voltage rail.

>>2611656
Sure they do but I won't be using a ribbon cable.
I forgot to mention a crucial piece of information. The connector will be plugged/unplugged by apes and women so it needs to be kinda sturdy.
I have decided on a DC-37 which is kinda cheap from Farnell.

>>2611966
>apes and women

these 50-pin phone connectors have proven rugged enough for some 50 years.

>>2611966
Look up automotive connectors.

>>2611966
>apes and women

these 50-pin phone connectors have proven rugged enough for some 50 years.
whereas those DB-xx connectors can get bent pins, and pins that can submerge into the shell.

>>2611953
>Gives people who can't afford anything else passable products, and provides competition to the actual brands to actually be worth what you pay for them. I see nothing wrong with flooding the market with cheap shit, so long as it's not difficult for the average consumer to understand what's good and what's shit. "Amazon recommended" products do blur that line a bit too much for my liking.
the problem is a ton of them are heavily underbuilt for their rated use, or misses a ten cents component that would made them safe, and tend to self destroy.
I all in favor of having cheap things, but a ton of shit from ali is just ewaste
>>2611934
my usual recommendation for tools that you don't think you are going to use it often, is to go for the second cheapest tier of products, the first tier is usually crap

>>2611989
>I’m second illustration, how is current flowing through the load,

what makes you think current is flowing

> if the neutral isn’t attached to anything? The circuit isn’t complete. How is there voltage measured?

the voltage is measured the same way it always is

>Does the neutral become part of the “hot” line as it goes through the load, so we have voltage but no current?

you're starting to crack the code

>If this is true shouldn’t the voltage be reduced because of the resistance of the load?

oops, back off the track here.

>How would we even test the voltage on that neutral, since one side isn’t connected.

just give up and get a job at starbucks.

>>2611989
Hot is still live without neutral because it's AC. The danger in this scenario is that a connected load will "bridge" the hot and neutral leaving a live wire floating and no return path/breaker function which is a fire and electrocution hazard.

>>2612001
Sorry I deleted my post because I saw that it had voltage since they were bridging it with the multimeter.

I guess I still don’t understand something. Doesn’t AC need both hot and neutral connected to work? Will hot _> load > ground still work? If so why bother with neutral to begin with

>>2612015
>Doesn’t AC need both hot and neutral connected to work?
Yes. Now imagine your body acting as a neutral path.

>>2611993
Damn you guys replied fast. I deleted that post within 30 seconds.

>If this is true shouldn’t the voltage be reduced because of the resistance of the load?

>oops, back off the track here

Why doesn’t voltage drop after the load? The functions as a resistor right?

>>2612017
I get that, but don’t I also need to be connected to the neutral power source as well? So the AC can push/pull. If it were DC it could pull electrons through me from the ground. But with AC I thought it was a push pull with the power source at both ends.

>>2612018
>Why doesn’t voltage drop after the load? The functions as a resistor right?
No current means no voltage across the "resistor".

>>2612015
>why bother with neutral to begin with
Safety. Neutral conductors are bonded to ground at the circuit breaker panel. Each circuit connected to the panel has its own circuit breaker which requires neutral to function. The ground conductor is meant to contain arcs/faults from any hot conductor and discharge through the grounding rod at the service panel while the neutral will quickly trip the breaker due to leaking current between conductors.

>>2611877
try /qtddtot/ or sparky general first
>sparky general
there is no such thing tho, so i ask here, how hard is electrican as trade to get into when it comes to knowledge?

>>2612027
You need to pass an exam to get a license. The best thing to do is get a contractor's license and make other people do all the work, but you need to know what the fuck you're doing.

Hello frembs. I want to get into tinkering with electronics, and feel having a collection of parts and materials would be best, like a big grab bag of resistors, capacitors, some lil chips like a 555 timer etc. Could people here who AREN'T retards like me tell me if this plan makes sense, and if so, if there are any 'grab bags' in particular i should be looking to get. I'm sharing a pic from my favourite childhood tv show as a thank you.

>>2612058
>Could people here who AREN'T retards
he doesn't know

>>2611953
>EEVblog forum multimeter spreadsheet
thanks for the link, i'll check that. my only fear sometimes with spreadsheets is you can get so bogged down in research that you never end up purchasing something
>Gives people who can't afford anything else passable products
Absolutely, I'm not arguing against that. But fort instance with bicycle parts I've purchased second rate tools that break almost immediately because they are made from low grade metal. The old saying, you get what you pay for is often not too far from the truth on the low end.
>>2611988
Good shout.

>>2611977
I did and they're usually
>expensive
>some weird combination of pin thicknesses for high power and signal transfer
>extremely annoying to unplug
>all of the above
>>2611979
Do these go by any other name than RJ21/Telco? I guess I'm not looking in the right place.

>>2612019
Since the neutral is bonded to ground at the service panel, it is connected to the earth, which is a different concept than a DC circuit ground. When something is earthed, a path can form between the hot and anything else that is connected to the earth. The soil, cement and rebar that make up the building you are in can conduct the current from the hot, through your body and back to ground. If you had a ground isolated system, you could potentially touch the hot conductor where you are standing and not get shocked because there is not path through the ground anymore.

>>2612058
If you are serious about this, invest in some decent breadboards, a temperature controlled soldering iron, a bench top power supply with current limiting abilities and a decent dmm. You will eventually want an oscilloscope, I recommend a digital one that includes a function generator. Analog meters are great for up to 60mhz on the high end, but they take some skill to use effectively and may intimidate you away from using them. The key thing you need to do is to isolate all the errors that could occur from your equipment away from your experimentation process. You don't want to lose hours due to a bad component or faulty test equipment.

>>2612157
A note on this:

While buying all this equipment seems like it will make the hobby too expensive, the parts themselves are ultimately dirt cheap. I learned the hard way that having poor equipment and counterfeit parts will take many hours of your life away. You may learn a lot of troubleshooting this way which is good, but it feels a lot better to have things work in the beginning and build your confidence that way.

about smd transformers and inductors

how do you get the specifications of a transformer that is used on any circuit board if the part has no identification number ?
how can i figure out how much windings the thing has and or what wire diameter they used to wind the thing, preferably without having to desolder it and take it apart ???
is there some kind of tester device like these for diodes ?

>>2612173
ring tester
https://danyk.cz/avr_ring2_en.html

>>2612157
>decent breadboards
there is so many crap on the market
can you recommend one or two who match the _decent_ criteria ?

>>2612177
tyvm !11

>>2612147
>by any other name than RJ21/Telco? I

50-pin centronics.
there's also a 36-pin version (see red box in pic)

>>2612194
Thanks for the help.
Also found them under the "Parallel SCSI" or "D-Sub ribbon" names.
They seem to be about 5x the price of the regular DB connectors though. The fact that I need like 12 pairs of them would add up significantly.

I'm looking for some "horizontal" female 24pins ATX socket.
or some ersatz able to host the male socket.
Please, does Anyone know the correct reference?

totally different question. I have four different 3phase cables hooked by grid company on the outside of the wall 6 feet away from desk. From times to times, I feel weird in nerves and muscles, sometimes dizzy, sometimes feeling slight shaking + spasmodic expansion on the elbow or leg
Does someone have knowledge or lectures about Inductives leakages currents and the body as a receptor ?

>>2612222
are you stepping in a spicy puddle?

>>2612222
did you get the vaxx scene?

i'm trying to make a +/- 15V flyback SMPS from a 12V input. i managed to get a regulated +15V output working in LTspice. but when i try adding a center tap and another transformer for the negative half, the regulation on both sides won't go beyond +/- 12V. what am i doing wrong?

yay. our guy.

>>2612357
Is that Chuck or Sneed?

>>2612343
Try two seperate primary coils, and two seperate coupling arguments. Can’t see any reason why it would help though, it’s a shot in the dark. Either your converter is running at maximum duty cycle and can’t make a higher voltage, or your feedback circuit is triggering wrong. Should be pretty easy to figure out which case it is.

Also what method do you use to simulate a flyback (airgapped) transformer as compared to a regular one? FYI a coupling constant of exactly 1 can cause strange results.

>>2612361
Not sure but NC flag is something he could use on himself if you know what i mean wink wink

>>2612357
wtf i love autogynophilia

>>2612021
Note that the ground (bonding) conductor is a lighter gauge wire than the neutral/hot.

The GFCI is supposed to trip quickly.

>>2612343
>>2612375

fixed it. i didn't think to look at the current through Rsense. adding another secondary transformer coil was pushing it into current limiting, which caused my secondary voltages to cap out at +/- 12V. i changed Rsense to 0.25 ohms and i now get a regulated +/- 15V at the output.

>Also what method do you use to simulate a flyback (airgapped) transformer as compared to a regular one? FYI a coupling constant of exactly 1 can cause strange results.

i have no idea, i'm just a hobbyist. i've read about the importance of airgapping for flybacks but i'm not sure how to simulate it -- maybe ltspice's leakage inductance parameter?

anyways, when i actually build this circuit it's going to be african american engineered to hell because the calculations, especially the magnetics and compensation network calculations, are beyond me. i have some chinky ee25 transformers, a spool of 24 awg magnet wire, and one of those $20 combination LCR/transistor tester meters. hoping to get something that barely works.

>>2612498
Good luck then. Personally I'd mess about with simpler topologies until you're more comfortable with SMPS design (e.g. using based TL494 plus a gate driver IC). Don't really get the fixation with flyback converters.

>>2612500
yes, that's a good idea. i do have some tl494s on hand. i've also simulated a buck converter (using the uc38xx/lt12xx IC) in ltspice that seems to work well. i'll probably build the buck converter first, since the feedback loop is simpler.

>>2612058
Get yourself an Arduino starter kit for like 30 bucks

>>2612235
no I got no vaxx hoax, no covid joke. But feel dismantled in the hips like Marconi.

Is this old audio ribbon present into an audio equalizer swappable with such stripe ?
Or should I stick with same model as the distance between wires may have some reality against magnetic fields and noise transmission between channels ?
I'm sick of those old ribbons, they are very sensible to hot air.

>>2612019
you don't NEED to complete a closed path to complete a circuit, look up capacitors.
you can get a shock from an ac line while completely insulated as the current changes direction and charges/discharges you at so many cycles per second.

I have this buck converter circuit built up on a breadboard. I'm supplying it from a bench power supply. With no load connected, it regulates the output to just over 21V, as it should. But when I connect a load (a 80Ω power resistor, which should draw around 260mA), the output drops to around 18V and I have to raise the input voltage to 30V (my PSU max) to raise the input back up. and even then I can reach only 20.9V
Why is this happening? I'm definitely not reaching the current limit set by the .1Ω resistor, so why is the voltage dropping?

>>2612520
>you don't NEED to complete a closed path to complete a circuit, look up capacitors.

Maxwell had to come up with displacement current to do that very thing, and his famous system of equations did not work without it.

The most amazing thing about his work is that he was basically just a math guy who was able to derive the connection between electricity, magnetism, and light, purely from a pile of relationships that the other guys derived from empirical data.

what's up /ohm/ kind of a noob question here. I'm building a kind of game controller that looks like what I have in the picture. There's a big box with a green button that houses an arduino/breadboard, and a bunch of little boxes each with a red button that do not contain brains, just wire. My initial plan was to just wire all of the little boxes directly to the big box with 22AWG, but I realized that's a terrible idea because if someone tugs at their small box it might unplug from the breadboard, and it will be a pain to carry everything around if they're plugged in at all times. I was hoping to mount female I/O on each of the boxes and then plug in the appropriate cables. Is there a standard way to do this? I appreciate the help

>>2612588
Use RCA jacks and patch cables.

>>2612520
>you can get a shock from an ac line while completely insulated as the current changes direction and charges/discharges you at so many cycles per second.
comically wrong

>>2612594
What do I do about the ground wire?

>>2612605
Nvm I got it thanks for the suggestion

Anyone ever tried using a polyfuse to balance lithium cells?

>>2612519
It should be fine.

>>2612582
He did have faraday to ape who had correctly posited everything besides the aether.

>>2612620
Polyfuses increase resistance with each cycle, leading to more cycles—a runaway.

It’s a protection device, under normal circumstances it should never be used.

>>2612552
>that pinout-based circuit diagram
>all the google image results for "MC34063 schematic" are the same
what the fuck, 34063 users have brainrot
also shouldn't the node between the big electrolytic caps be GND not COMP?

>>2612766
>all the google image results for "MC34063 schematic" are the same
I copied it from the "Application Information" section of the datasheet.
>also shouldn't the node between the big electrolytic caps be GND not COMP?
Yeah. I accidentally used the wrong ground symbol.

guys, i used to buy soviet electronics thru ebay. now russians are banned. what are some big electronic parts websites in russia like mouser or RS in the west?

>>2612830
Buy it from other former soviet counties retard.

>>2612552
did you simulate with ESR of inductor quantified?

>>2612154
How is the ground in AC different from the ground in DC? They both provide a source of electrons right?

>>2612830
>>2612834
Most of them are made in Ukraine anyway.

What's this style of potentiometer called (white square in the middle)? From an old Casio keyboard from the early 80s, measures 7mm x 7mm x 1mm, 430 kOhms.

>>2612914
Open wiper, single turn, SIP, ceramic, cermet, trimming potentiometer.

Looks to be reasonably high current, replace with a ½ watt to be sure.

>>2612881
It's a different thing with the same name. DC circuit ground is just the negative prong of the battery or power supply that is common to every circuit node. Ground in the context of an electrical service is physically connected to the GROUND as in the dirt, soil, and foundation. Since the power pole is also grounded, anything connected to the ground can conduct current back to the source. The neutral in mains AC is connected to the ground in the panel so current can flow through your body to get to the source, rather than through the neutral wire. This is why you can get shocked by just touching a hot wire in a mains scenario.

>>2612881
>>2612931
You might ask yourself why we bother grounding things. There are a Few reaons:

1. To provide a path for lighting strikes and EMP's in the utility company system. Without grounding, any large EMP's would completely destroy all of your electronic devices, and in the case of lighting possibly burn your house down.

2. To keep things at the same voltage potential relative to the mains power. If every conductive surface at your house is grounded, sneaky voltages cannot appear across them and shock you. By 'bonding' everything in the system together you don'y need to worry about plugging something in wrong or touching to different metal surfaces in your house. This partly became an issue when grounding was first adoipted. Once the ground is a conductor, anything that isn't at the same potential as the ground becomes a serious shock hazard. This is why almost all countries require conductive surfaces to be bonded together.

3. Fault clearing. Now that all metal chassis and surfaces are bonded together, any lose wires or broken connections are likely to touch a metal surface that is grounded and instantly tripping your breaker. You don't want a potentially lethal voltage sitting against the chassis of a broken appliance and your breaker to remain closed. By bonding all conductive surfaces together you've created a low impedence ground fault path that will quickly trip a breaker and prevent arcing and shock hazard.

If I send an update to a flip-flop, does the state change become visible during the current clock when I send the update, or do I have to wait for the next tick to read the updated state?

>>2612950
Depends on the topology. Master-slave needs first a positive then a negative edge I believe (or vice-versa), while a classic edge-triggered flip-flop will trigger as soon as the first edge comes in and propagates through. See:
https://en.wikipedia.org/wiki/Flip-flop_(electronics)

>>2612950
I’ll assume you’re talking about the legendary 7474.
The D input is available on the leading edge of the clock.
The R and S inputs (reset and set) are available immediately.
There are other types of flip-flops or “latches” without clocks and different behaviours.

>>2612967
>>2612969
I'm trying to do FPGA programming as a total retard who knows nothing about electrical engineering. Basically, if I "assign" to a standard register/flip-flop, should I expect the variable assignment to propagate within the same clock?

>>2612979
>expect the variable assignment to propagate within the same clock?

almost certainly it'll propagate one stage per clock cycle.

>>2612979
>>2612981

the appropriate datasheet will have a timing diagram similar to this.

>>2612981
>>2612982
So if it propagates instaneously, does that mean you can include flip-flops in combinatorial logic?

>>2612990
> can include in combinatorial logic?
A flip flop is made of combinatorial logic.
picrel==7474

>>2612979
If you clock a flip-flop and sample the output of that flip-flop at the same time, you should expect to sample the previous value of the flip-flop, not the new one being clocked in. Invert a D-FF's output and put it into its data pin and you've got a T-FF, not an undefined system. Also look at how a shift register works, all those D-FFs are clocked at the same time but the data propagates only one stage per clock.

>>2612991
That ignores the edge-triggering circuitry that makes it synchronous though.

Is there any problem with using an OPA551 in this configuration? I need a lot of output voltage swing and little bit of current.

>>2612873
The datasheet gives a DCR of 0.059 ohms. I doubt that makes a difference. There's gonna be more resistance in the jumper wires.

I also noticed I made another mistake with the schematic, the timing capacitor is 330 pF, not nF.

>>2613013
>this configuration?

never seen anyone put their load in the feedback loop.
likely wont work as feedback currents are usually super tiny.

So, do they KiCad fags have to select a footprint every time when adding a symbol to the schematics? That's insane. Would you have to literally remember what to search for and look through dozens of look-alike footprints to pick the right one? What happened to the idea of one part one symbol one value one footprint?

>>2613062
It's called a constant current amplifier. That exact schematic has been used in guitar amps for decades. I just want to know if I can swap the chip for a OPA551 and increase the voltage supply to get more output.

It's an unfamiliar chip but it sounds like a miracle. They call it a high voltage high current opamp.

>>2613002
> ignores “magic circuitry”
Here’s another version using nand-complete logic that you can wire up yourself and prove to yourself that it does work, as-is. The magic is in those “cross over” connections.

>>2613150
>The magic

You actually seem to understand the topic. Why use gaytarded words like "magic".

>>2613152
He did voodoo.

I need to solder on some cylinder SMD caps onto this board im recapping. (not the tiny #### ones). What temp should I have my iron at to keep the solder on my tip prior to applying to the pad/'leg'?

I dont know the wattage of the iron, it was a chinesium brand.

Also what is a good temp for just heating up the pad to apply solder to it to avoid overheating and lifting the pad?

>>2613206
Keep it on full blast. Add solder to each pad and alternate the iron between them while you lift the cap with your other hand. Don't pull, just keep some tension on it.

>>2613013
Don't see why not.

>>2613115
For almost all symbols, you select the footprint in as you pick the symbol. Sometimes this is completely automatic (e.g. picking a 1N4148WS which is automatically a SOD323), other times there's a few options (e.g. picking a 74HC14 where you can pick between DIP14 and SOIC14 and maybe a few others). The drop-down menu usually does a passable job of only having relevant footprints, as filtered by the keywords defined in the symbol. But if you're picking resistors, then yeah you'll have a shit-ton to browse through. Thankfully the common SMDs are right at the top, and the DIN0207s are pretty close too. I advise just copy-pasting passives as that preserves footprints, using different symbols like R and R_small to denote different footprints.

It is possible to add default footprint assignments to existing symbols, and naturally you can add footprint keywords and defaults to custom symbols.

If you're not sure what size to go for, then just leave the footprint un-assigned while you work out your schematic, and add it later before exporting the netlist to a board layout.

>>2613150
My bad, I guess I don't understand master-slave circuits well.

>>2613206
More about force minimisation than temperature minimisation. If you can melt one leg and pull the leg off the board without putting much stress on the other leg then that's the easy way to go. Even less stress if you pull the cylinder off and cut the plastic carrier so you can remove the legs independently.
High temperatures definitely increase risk of delamination, but the forces with which you can pull can certainly just delaminate a cold pad. Trying to melt both at once is a bit of a tricky thing to do without two irons or desoldering tweezers, unless you buy chipquik.

>>2612990
>So if it propagates instaneously

it doesnt propagate instantaneously.
the drawing i gave in >>2612982 shows that D input appears at Q output on the falling edge of the *next* clock pulse.
so it could take as long as 1 clock cycle for it to propagate.
(doesnt have to be falling edge, could be rising, depends on particular FF)
the only inputs that propagate instantaneously are things like the RESET pin.

I want to use this Sony Walkman as a preamp. Which terminals on the tape head correspond to those for the 1/4" jack?

>>2613261
Look up the datasheet for TA8132AF.

>>2613261
>>2613267
The signal goes through the volume pot.

>>2613268
Ah, ok. That's helpful, thanks.

>>2613271
np

>>2611877
I need help. I have to get replacement capacitors for C90, C91, C51 and C56 @ 560uf, 200V. I calculated that I needed at least 3.0 rms Amp @ 120 hz for 690W for each capacitor. ESR meter told me that they have ~50 mOhm, which doesn't make sense but OK.

I looked every where, can't find capacitor at specs with 3.0 A ripple current max, or at the strange ESR value.

Am I doing something wrong?

>>2613274
Forgot to circle it, my bad.

>>2613240
ok thanks

>>2613213
I was running it at 250F, its 60 w and that was a fine temp to hold solder on the tip and solder on the smds in seconds without smoke either way.

Testing Google's AI

>>2613274
Have you tried looking for less specific ratings? Like 200-400VDC, 3-10A, 50-10mΩ, etc.

>>2613240
>Don't see why not.
The thing I'm worried about is the dissipation. It's thermally limited but a 8 pin DIP isn't gonna handle much. I need like 3 mA constant current output. The problem is they're saying low voltage swing on the output wastes power in the internal transistors.
So I guess I'm right at the edge of the SOA.. Red is at 1kHz, green is 6kHz (impedance rises with frequency) and blue is 6kHz if I use a 12k resistor instead of 4.7k.
This chip might not be such a miracle at all. Quiescent power consumption is already at .36 watts just by having a +/-24V supply. I might need to lower the supply voltage or boost the current with transistors.

I'm about to attempt to wire the chinkiest, most minimal usb c host to a regulator and then to the esp8266 I want to power. The host has 3 wires: +, -, and c. Apparently, I'm supposed to pull c down to ground to signal the presence of a device with a 5.1k resistor, which I don't have on hand.
>What will happen if I connect c directly to ground?
>Will the previous question be rendered moot because the host will supply power regardless because it has no concept of negotiation?

>>2613425
There is basically no reason to use that 5.1k resistor unless you're bit-banging a USB protocol. The PSU will output 5V regardless, though you'll have to look into whether you need communication to get high current, and I'm 99% sure you'll need negotiation to get a higher voltage. If you need a higher voltage, buy a dedicated IC or try to bit-bang it with V-USB or whatever.

Did rectangular LEDs go out of style or something? Why are they so hard to find right now?

>>2613467
Is it hip to be rectangular? Everybody uses light pipes and diffusers now for panel indicators.

>>2613425
>>What will happen if I connect c directly to ground?
5V 500mA

>>2613480
yeah well everything I still own uses rectangular 2 x 5 x 7 LEDs and if I want to swap them I don't want to buy from china but I guess I'll have to

>>2613482
https://www.mouser.com/ProductDetail/Kingbright/WP113SRSGWT?qs=58z0TXQGVSQ0SKlq5iYnJQ%3D%3D

>>2613483
non-stocked, minimum order 20,000
they only carry 3 rectangular LEDs total, I already checked

>>2613486
Sorry, Anon. I saw everything except the quantity in stock because of tardation.

>>2613486
https://www.ebay.com/itm/265919218411

>>2613488
that's a good one. thanks
I saw plenty like that on ebay but that specific one has a mixed bag option which is cool

>need to get electro education to get loicense
>decide to apply for sparky trade school
>have to bring application in person between 3 to 5 pm
>arrive at 3:15
>already full
it's over

>>2613467
Just sand a round one down.
I just use some wet sandpaper.
You don’t want to inhale epoxy dust.
Then you want some like 800 grit for the top to give you a soft diffuse look.
You already have the raw materials, just DIY it.

>>2611957
>it wouldn't turn back on after unplugging it unless you waited for a while. It lost power for a long time and now doesn't turn on even if you leave it plugged in all day. I suspected capacitors because of the transient behaviour but they all looked fine visually.
Sounds like some kind of resettable or thermal fuse. It usually can take up to an hour for some fuses to reset. What caused the fuse to trigger is most likely a short somewhere.
>Testing the power outputs of the board with the AC setting on my multimeter I got stable 11V for the +5V outputs, and 45V for the +16V outputs
Looks like you have a very noisy output, might be an issue with the snubbing circuit, although it can also be caused by a big short somewhere and your buck converter IC on pic rel cannot keep up. You should definitely use the DC mode for DC voltages, although having such a high AC signal could mean something wrong with the power supply. You need an oscilloscope to diagnose this further, which could tell you if it's just temporary noise you pick up on the multimeter or an actual short.

>I've isolated the problems to pic related (which now that I look at it is faintly tinted a darker green but it doesn't come out in the pictures). Blue is all connected to ground, yellow is the low voltage rail, red is the high voltage, pink is reading as zero. Orange is unsure because it reads 0v but doesn't have continuity to ground, so it could be blue or pink.
>Is the IC dead? The datasheet says VCC should be 3-26V, but I got 45V on AC and 0V on DC. But power seems to be reaching the output connections so I dunno.
It is possible, if the IC has a built in switching transistor, and you have a short somewhere on the board, the switching transistor inside the IC is most likely toast. I recommend first to find the short and then to replace the IC. If the monitor works for a while and then stops working, it might be a short on a part of the circuit that does not run all the time if that helps.

>>2611957
Should you need more help debugging this, let me know we can try to go through it together.

AAAAAA KiCad is driving me fucking nuts.
I accidentally moved the NC symbol and now I can't select it! I keep clicking on it but the pin is selected instead. Zooming in doesn't help. What am I doing wrong? Is there another way to select it or delete it?

>>2613736
Whew. Right click and cut worked. What a piece of shitty software.

>>2613738
>doesn't know how to use software
>caveman mode.exe
You can at any time right click on packed nodes and get a list of each individual components.

Do you know what sucks the most about KiCad? Selecting an operation immediately performs that operation. For example if I hit Q it immediately places NC at the current position insted of switching to NC mode and waiting for a click. Same with wiring. Hit W and it starts wiring immediately like a motherfucker instead of letting me reposition the mouse and click to start wiring. No other EDA or simulator works like that.

>>2613741
Fair enough but isn't that a bug? If NC is tucked inside the pin area like in the picture, KiCad gets confused and cannot select it and selects the pin instead.

>>2613741
>right click on packed nodes
what's a "packed node"?

>>2613738
>What a piece of shitty software.

suck it, young padawan.
us oldfags had PCB software that beeped the buzzer on the motherboard
- every time you hit an undefined keystroke
- every time you typed a letter when it wanted a number, or vice versa
- every time you did anything outside what it was expecting.
- no popup warnings
- no error codes
- just soul-destroying beeping: 200 times an hour, with no way to lower the volume.

>>2613747
I think it means non-binary.

>>2613749
Imagine in a couple of years any kid will be able to just voice control your EDA and other tools and design a space ship in a few minutes.

>>2613751

>>2613751
Yeah, those couple of years will coincide with everything being powered by fusion.

You know most people can read a book faster than they can listen to it, eh? The same thing happens in reverse.

>>2613742
i like it for that reason. less clicking. for all the bloated features in altium, ive never had to use any. using it professionally or recreationally i can do the same shit in kicad faster and with less annoying restrictions imposed for no reason. dunno why companies waste so much money using full altium licenses and expensive pcb fabs when kicad and jlc achieves the same quality for a fraction of the cost

Why use KiCAD if you can pirate latest version of Altium in two clicks?

>>2613759
but kicad has some annoying bullshit like this PWR_FLAG. wtf. i guess one could just accept all the quirks and get used to them.

>>2613761
Piracy is theft. Can you just walk into Walgreens or Walmart and stuff your bags and just casually walk out and expect the security to look the other way without trying to stop you? Same thing with software.

>>2613766
>You wouldn't download a car.
lmao

>>2613766
Are you OK anon? I don't give a flying fuck about that and you should not too.

>>2613744
It's not a bug, it usually happens when you mismatch the grid dimensions. If you place a component on one grid size then move to a larger one, kicad will select only the component that matches the new grid size in the case of two different components very close to each other. To select something on the old grid size, you can either switch the grid size again or can right click and go through all the options. It sucks but there must be a reason why they did it that way.

>>2613761
cos i may do commercial shit

>>2613775
I thought about it since I got bit by it already when I tried wiring two pins together and kept missing, they just wouldn't connect. Now I make sure the grid is set to the recommended 50mil and never touch it.

>>2613696
I already do this when I need a flat top 5mm LED. No way in hell I'm doing that for 5 different sides.

>>2613816
you can always go multiples of 50 mil, i don't remember what the default grid for symbols are, but if they are 100 mils, you can do 100, 50 or 25 just fine, but yeah, keep it at 50 mils unless you need to go smaller.

>>2613828
>multiples of 50 mil
of 25 mils

>>2613381
Da

560uf
200V
10mm lead spacing
22mm dia
radial can
through hole

Can't find ripple rms max @ 120 hz greater than 3 amp.

Can't find DF of 0.01 or ESR of 0.070-0.020. All DFs I can find are 0.1-0.2

Original Capacitors x4

(3) JWCO; 560uf; 200V; vent; KM+ 105°C, PET 1801

(1) JWCO; 560uf; 200V; vent; -40 - 105 C_1, PET 1705

I looked up JWCO KM series datasheet but they don't list 560uf at 200V, as if they don't make them or something, also the ratings on the sheet are not what I need, ESR doesn't match, and max rms ripple less than 1 amp.

>>2613837
>>2613274

I did ripple current calculations with 690W (that is the max).

I am more worried about the ESR now. I can't find one low enough. Mouser filters are worse than digikey.

I need to recap the board, but I need ripple current rms max and matching esr values.

>>2613828
the default is 50mil if i read correctly.

>>2613706
You know those results kinda look like what you’d get if you plug a 110V appliance into 220V. Maybe there’s a switch in the wrong position.

>>2613766
I get this.

>>2613837
Maybe those requirements are just too stringent? See how maximum ripple currents and ESRs are calculated by other people.
I can’t imagine just copying the original values would result in a faulty system after all.

>attached ground probe to neutral
>measure capacitor negative lead: -170 V
>measure VCC: -157V
>attach ground probe to capacitor negative lead
>measure VCC: 6V

What the fuck is going on?

Hello from ChatGPT

here once was a tool named KiCad,
Whose features were quite sad.
It was free and open source,
But it was slow and a source
Of much frustration and pain.

Then came Altium, the chad,
With its features so glad.
It was not free, but it was fast,
And it had a great ui,
And it made KiCad users sad.

So if you're looking for a tool,
That will make you not drool,
Then choose Altium, the chad,
And forget about KiCad, the lad.

The Chad Altium

It was a dark and stormy night.
The wind howled and the rain beat down.
Inside, a young engineer sat at his desk,
Working on a new design.

He was using Altium,
The most powerful PCB design software in the world.
With Altium, he could create anything he could imagine.

He was designing a new type of circuit board,
One that would revolutionize the world.
He was working late into the night,
Driven by passion and ambition.

Suddenly, there was a knock at the door.
The young engineer looked up,
And saw a man standing there.

The man was dressed in a long black coat,
And he had a strange look in his eyes.

"Who are you?" the young engineer asked.

"I am KiCad," the man said.
"I am the god of open source PCB design software."

"What do you want?" the young engineer asked.

"I want to challenge you," KiCad said.
"A duel to the death.
The winner gets to design the future."

The young engineer laughed.
"You're on," he said.

And so, the two of them fought.
They fought for days,
And neither one of them gave an inch.

Finally, on the third day,
The young engineer struck a fatal blow.
KiCad fell to the ground,
And the young engineer was victorious.

He had won the duel,
And he had won the future.
He was now the master of PCB design,
And he would use his power to change the world.

>>2613900
>kookaburra hands typed this post

>>2613891
Idk man, probably some artefact of a bridge rectifier or something otherwise not being ground referenced (e.g. a floating output). Draw a simple schematic.

>>2613838
>I am more worried about the ESR now.

too lazy to re-read all your BS, but from what i remember you're trying to match a measured ESR to a rated ESR.
if so, you're forgetting that the rated value is a guaranteed maximum.
any actual caps you buy will be lower.

>>2613766
>and stuff your bags and just casually walk out and expect the security to look the other way
in SF yes

Are these zeners just used for clamping the input to this pin?
So the voltage can't on the pin won't go higher than 11.7V + the forward voltage of D58 and won't go lower than -6.8V - the forward voltage of D57?

>>2612881
To add to the other answers, the ground of a DC circuit is just a reference point for that circuit and it could potentially be several hundred volts above Earth.
I've seen plenty of interns connect an oscilloscope probe to a floating ground a blow the fuse on the scope.

>>2614001
reverse polarity protection / transient voltage suppression

How do I make a stable power supply?

>>2614014
rubber feet

>>2613930
Pic related is a diagram of the start up circuit.

The power ground (PGND) is at -168V from neutral. As confirmed with a multimeter.

The VCC ic is -154 volts from neutral. As confirmed with a multimeter.

Thus the voltage difference between VCC and PGND should be ~13 V, but the multimeter reads ~6V. Way below the startup voltage and even below the shutdown voltage, the IC isn't turning on at all.

Does anyone have a proper Field Oriented Control schematic? I'd like to simulate it and figure out how it works cause all those youtube videos are not helping much.

>>2614035
>multimeter reads ~6V

you need to cut the trace going from emmiter to chip.
cutting traces is a lot of fun, believe me, when you have a sharp x-acto knife.
coz you can be a pretend surgeon for 1 minute.
if the voltage stays around 6V then you have a bad zener or bad transistor.
if it jumps up to 13V then likely the chip is sucking in too much current.
place various resistors between emmitter and ground to see how much current you can pull before the voltage sags too low for chip operation.

>>2614035
I have no clue but I am just curious, what's the point of resistors and capacitors in series and in parallel? Is this for filtering? Why can't you use one single filter capacitor? Or maybe two, + to N and PGND to N?
What happens if you remove the transistor and the diode, and only keep the zener and the limiting resistor. And what's the purpose of another diode in series with R141k? I've done something similar and it worked well. I just had a single zener and the limiting resistor of about 1k. I didn't need more than 10mA on the output.

I'm using these cheap chink 48V li-ion batteries for a project, what are my chances of a workshop fire?

Zero documentation on the BMS or the cells used, maybe I should rip one open to see what's inside.

>>2614088
Is that 13S2P? 18650s shouldn't get more than 1A current per cell.

>>2614091
13S3P.
For some reason these battery packs are readily available everywhere, often with stickers that state they have '99999 mAh' capacity.

>>2614068
> point of resistors and caps in series and parallel
That would require a novella to explain.
Often you see multiple values of capacitors in parallel because smaller caps are better at removing high-frequency crap and bigger electrolytics are for filtering out the 50Hz mains frequency (turning pulsating DC from the rectifier into smooth DC on the output. At low loads, anyway).

How come the teensy is so expensive?
What's the best one for embedded?

>>2614051
Thank you. It is now reading 13V after removing transistor. A bad transistor wouldn't had effected the voltage imo, at least that which is seen by the zener, I should of clarified that the zener and the VCC read 6V before I removed the bjt, so it has to be the IC taking too much power. The ic is either broken or maybe it is being shorted on one of it's logic pins that output reference voltages or charge a capacitor. If that wasn't the case the IC would be cycling IMO, since start up current is very low and operating current exceeds 1ma, which is the current required to bring the VCC down to 6V when it is not getting voltage from the auxiliary.

Afternoon. I am in the market for a cheap but reasonably decent soldering iron, and have absolutely no idea of what to buy because I'm probably functionally retarded. An Anon (>>2614264) suggested I ask here, so I'm giving it a try. The soldering iron would be mainly used for stuff like soldering and de-soldering battery leads to electronics, and replacing/removing the tabbed watch batteries on older video game cartridges. I am pretty clueless when it comes to soldering, but it's a useful skill I want to learn so I can maintain and repair my collection of older electronics. Help is greatly appreciated, because a I said earlier, I'm functionally retarded with this sort of thing.

>>2614035
Why is that transistor circuit coming from single diode rectified mains, instead of from the bridge rectifier?

>>2614088
Better than a lipo.

>>2614283
Don't take any brand suggestions from here, do your homework on what you can find locally for cheap and make sure whatever you pick you can get cheap tips for it.
I got memed into a pinecil and after the first tip cratered too much I searched to buy more and it was like 6$ a pop before shipping and import taxes.

I'm trying to fix my projector, it's a BenQ MX761 and I just installed a new projector lamp. It keeps overheating after a few minutes after turning on; I tried cleaning out the fans, resetting all settings, putting it in high altitude and economy mode but it still overheats. Any ideas?

>>2614316
Tech support is on >>>/g/

>>2613886
The power plugs are completely different for 110/220, unless you're talking about those switches behind things like computer power supplies that let's you select 110/220V. I personally have never seen that on a monitor and most modern supplies are so called off line universal power supplies that can run on anything between 90-260V 50/60Hz without having to flip any switches. Could be worth looking into though.

>>2614326

It's an electronic problem, sure I might need to replace the ballast but I was wondering if there are any certain electronic components that usually go bad for when projectors overheat; input capacitors etc. or if there were any way to lower the output of the lamp wattage?

>>2614316
Did you replace it with the correct bulb? The bulb might be dissipating more power than the original one.

>>2614330
>if there are any certain electronic components that usually go bad for when projectors overheat
Any short will heat up the thing, but if the product is designed properly, there will be a fuse that kicks in to prevent it from being damaged. The most likely culprit though are shorts on things like power ICs/Transistors (Switching transistors, voltage regulators, switching buck/boost/flyback ICs, etc.). Did you remove any heatsinks during the replacement of the bulb? Make sure to clean them and use good quality thermal paste.

>>2614316
thermal sensor / thermocouple

>>2614332
Yes, I checked the numbers, checked the housing and it all looked the same - even checked the resistance for good measure. It seems a bit harded to lower the output wattage manually since it's not a regular bulb.

>>2614336
or a dead DLP chip

found a bell 220 AC => 12 AC PSU
>why not building a diodes bridge and make it DC
craft my shit from memory, solder prototype board
>oh shit it's not a svastiska layout for diodes
>unsolder & reverse 2 diodes, solder shit again
>fire at will, run the board
>1000µF 16V Sanyo capacitor treebles and pops out
>First dead capa, RIP capa.
>read AC out and it reads ~18V.. ok shit
>unsoldershit, swap tip for long conical one, pump the shit out, >resolder some unknown chink 470µF 35V
>fire again.
Got on multimeter 24V DC on the outside of the capacitor.
Is it normal to have voltage rise that much ?
need to learn about voltage regulator now I think.

>>2614380
Put a load on the output and measure again.

>>2614380
>First dead capa, RIP capa.
Watch the polarity, electrolytics have polarity. You're also operating above the voltage rating of the capacitor, go for a 25V or 50V rated cap.
>read AC out and it reads ~18V.. ok shit
Maybe your transformer is not very accurate in the amount of turns, or maybe you're not running it on 220V but 240 or 260? Confirm that your mains outlet is 220.
>Got on multimeter 24V DC on the outside of the capacitor.
This makes sense if the output of the transformer is 18V RMS. Remember 18V RMS = 25.5V peak, which when rectified (assuming 2x silicon diode @ 1.4V drop) gives ~24V DC, under no load of course. If you connect a load, the capacitor will discharge and you will start getting slight ripples, increase capacitance to reduce ripple.

>>2614283
You will have better luck searching youtube videos of "decent soldering iron for beginners" or "What to get as my first soldering iron". Don't be dismayed by the terms "beginner" or "first". That is just youtube reviewer/tech bro talk for "not at the level of my expensive 20000$ instrument that I show off every video for clicks"

Soldering irons are like linux distros, they are only as good as their package manager.

You might get one you like, but come time to get different soldering tips, branching out with solder tweezers, or just replacing generic parts you might run into apple levels of BS and all sorts of model incompatibly issues with the various assecories and add ons.

>>2614068
>>2614305
https://www.homemade-circuits.com/simple-voltage-regulator-circuits-using-transistor-and-zener-diode/

>>2614424
Thanks Anon, I'll get to looking at review videos, much as I hate having to rely on youtube.

hey I am new to /ohm/, been learning some basic components stuff and I finally got my hands to do some soldering the last few days using some old boards and recycled old components.

I am now trying to practice and learn some schematics reading and circuit design. I have this Heads and tails schematic and I am failing to see where do the pins 1 and 2 from the 4027BE IC go? Been trying to find and read some datasheets about that IC, understanding not even half of it. Can anyone orient me on what to do with those pins? Just leave them not connected to anything? Pins 14 and 15 go to the LEDs. But why are they called Q1 and ~Q1, Q2 and ~Q2?

And one more thing, about copper track brakes, what's the correct approach to do them? In that same 4027BE IC for example, do I need to do breaks between all the pinout horizontally?

>>2614425
there are like ten different schematics there and none of them is even remote close to what i asked about

>>2614486
The 4027 is a dual JK flip-flop. The creator of this circuit only needed one. So he tied the inputs to 0V and left the outputs unused. Take note that he’s also retarded for switching LEDs with common-collector transistors. Transistors aren’t the best thing does a noob to dive right into, but it’s good to internalise the optimal “transistor as a switch” circuit anyhow. The pull-up resistors on the FF are also pretty stupid.

As for the flip-flop’s operation in this circuit, consider the three operations a JK can do. Set, reset, and toggle. When both J and K inputs are held high, then it operates in toggle mode, so each clock pulse will toggle the outputs. Q is the main output (high when set, low when reset), ~Q is just the inverse of Q.

Anyone >>2614041 ?
Yes, I googled it.

>>2614545
I found a TI appnote on it handy, though the Wikipedia page is pretty decent. It’s just a matter of having the right analogue front end, a fast enough 3-phase inverter stage, and the matrix operations going through your microcontroller. I’ll be able to say more once I’m home from work in an hour or so. I’m also working on my BLDC driver, using a DRV830x and an RP2040.

>>2614283
those $16 "90 watt" ones are decent enough, and after about a year you might wanna upgrade to something with temperature control. It's just a hot metal rod either way, it rly doesn't matter which you get as long as it heats up

>>2614425
I understand the theory, but why not pic related?

>>2614570
Non-temp-controlled irons, especially high-power ones, lack the thermal regulation I'd want to recommend. Considering you can get temp-controlled irons for $20-$30, I see no reason to go for a non-controlled iron for a noob. You can even get the thermal performance of cartridge tips using irons like the one EEVblog's Dave didn't understand recently, or just get a cheap T12.

>>2614408
>If you connect a load, the capacitor will discharge and you will start getting slight ripples, increase capacitance to reduce ripple.
Understood.
I continued further with voltage dividers to have 12V and 5V outcomes. I also put capacitors on those 2 lines. Not sure about how engineers chose them. It's fun to learn & discover though.

>>2614610
>temp-controlled irons for $20-$30
bullshit, where? i hope you mean actual temperature feedback and not just some shitty little knob

>>2614283
Soldering is simple, don’t over-think it. A 27 watt is a standard wattage for small electronics. They are self regulating (it won’t run-away like the demon core because of physics) and almost nothing to break.

Try and ignore the fads, larpers and hipsters that will try and steer you to toys that require things like firmware updates.

>>2614626
Pinecil hits that price range

>>2614635
And then you spend another 20$ in tips because the tips are liquid shit that get pitted in an hour of work.
Please shill your consoomer trash on reddit, not here.

>>2614648
>Tell me about your cheap temperature controlled soldering irons
>NOOOO NOT THAT ONE THAT ONE DOESN'T COUNT

>>2614648
I have never had this problem and I exclusively buy chinkshit. Tin your iron.

>>2614649
>what the fuck is money
It stops being that cheap if you go trough 4-5 tips per month. Go search how much the tips cost.

>>2614635
Lol, my pinecil soldering iron running a linux distro just got hacked and it recorded everything and sent it to china

>>2614650
I also exclusively buy chinkshit because I can't afford to blow 20$ on pinecil tips plus shipping and import taxes. Haven't used my pinecil in over 2 years because of this.
Now the1$ chink specials for the cheapest of irons from the local hardware store will pit just as easily as the original™ Pinecil™ garbage but in the 2-3$ range I can find some decent tips that last a long while. The one I'm using right now has lasted me for over a year and there's no visible degradation on it.
So, no, it's not about me not tinning my iron, I always do that, it's just that some tips are absolute garbage and the pinecil tips are exactly that but 10x more expensive because mUh PiNeCiL.

>>2614650
Ironically tinning your iron is part of the problem young one.
Get some picrel.

>>2614650
Ironically, tinning your iron is part of the problem young one. Get picrel.

>>2614653
>RTOS
>linux

>>2614283
EEVlog made a recent test of something in that range.
https://www.youtube.com/watch?v=Ia0C_X0ycFE

>>2614657
I find Broquetas great on daily use. Made in Barcelona.
My other rollers are thicker and it's finally not that good.
0.3, 0.5 are imho good. 1.0 is in the end good for feeding prototypes boards or high voltage connections.

the GEB roller on pic is thick as fuck. I suspect it's a "plumber tin" an electrician forget behind in Parents' house. Plumber tin is really something to avoid at all cost. I made tries with it and it's total garbage, even adding flux.
Hot air soldering with Chipquik is fun to experiment as well.

>>2614626
Some of the ones with a shitty knob do have real temperature feedback, but it's a bit of a crapshoot as to whether you get one with an open-loop phase-fired dimmer. Otherwise there are ones that plug into a control brick with the knob on that, or the irons with a digital readout and 2-3 buttons and 936 tips, pic related.

>>2614648
>>2614655
>buy chinky T12 station
>comes with knockoff T12 tips
>never tin them
>always scrape solder off them once i'm done working via the copper wool
>leave for weeks without use
>often crank up to 420C for XT60s
>repeat
>tips last literal years
skill issue

>>2614545
>>2614552
Yeah so I might have imagined this slideshow as an appnote:
>https://www.ti.com/lit/ml/slyp711/slyp711.pdf
But there still are good appnotes, like these:
>https://www.ti.com/lit/an/spraby9/spraby9.pdf
>https://ww1.microchip.com/downloads/en/Appnotes/doc32126.pdf
Seems they go more into detail about the mathematics and code than the hardware design. The sizing of components follows the same guidelines as any other power electronics project, figure out the size of transistor you need, pick something good and not too expensive, then figure out how fast you need to drive the gates. I'd recommend building a scaled-down but sufficiently fast model (maybe with BJT totem-poles instead of half-bridges) to test with, as any simulator isn't going to be detailed enough to program in such an algorithm. Testing algorithms with python or matlab or what have you is also something to consider, though to get all the transistor losses and such coming out of that would probably be too much work. A rough estimate of the computational hardware requirements can be done by looking at the maximum commutation speed (number of pulses in a coil at maximum rotation speed), multiplying it by some factor (e.g. 50) to get the desired PWM speed, and considering how many operations it will take your microcontroller to perform the vector transforms. In my case even a fairly fast motor speed is quite doable on an RP2040. The ADCs are another question though, I'm having to go somewhat exotic with them, and one for each of the three current channels. Hall sensors too.

Key takeaway from youtube is to not trust youtube, half the vids there are of people talking about sinusoidal control, not field-oriented control. Sinusoidal is just about changing the duty-cycle on the fly to be sinusoidal instead of the more common trapezoidal/square/modified-square, usually just using a lookup table. Take a look at hardware used by SimpleFOC and VESC for examples.

>>2614688
>>2614552
Thank you anon. I've seen some of TI notes before but it seems like everyone just refuse to share their FOC implementation as a schematic and no youtubers actually done this project before, they just explain the theory behind it.
Unfortunately I'm not that good in electronics, I'm just a hobbyist who wants to make the controller PCB and use in my toys. I can't make the schematics from ground up for such complicated problem, even though there are block schemes of the process available. It seems I'll have to wait with that or buy a working off the shelf solution. Kinda unlucky.

Do any of you use cad for your hobby?
In what way?

Found this with a bunch of telefunken radio tubes. Any idea what it is? It sort of looks like a mount for a bulb that would snap into place there are also holes on the bottom where a positive and negative charge might go in.

>>2614801
I design prosthetic camel toes in Fusion 360.

>I'm happy, just received my brand new
fuck you Bezos, bringing a new cycle of boredom for swap and repack.
If only there was other sellers for such book in Europe...

>>2614657
>>2614672
The point there is that savbit has Cu in the alloy that prevents the degradation of the tip. Molten metals can literally dissolve other metals, and it’s one of the reasons lead/tin “stick” so well to copper.

>>2614845
Doesn't help if the tips are shit.
The solder I use is 1% Cu and still eats away any low quality tips. Better tips don't care, they don't get damaged even by lead solder.

What do you do when the tip rejects the solder?

>>2614860
Depends on the condition of the tip. If it's crusty, use a wet sponge and brass wool to clean it up. Tip tinner/cleaner works too, but isn't really needed if you use quality shit and good habits. If the tip is degraded then buy a new tip.

>>2614860
It means your flux has turned to charcoal.
Clean it.
I use a wet sponge most of the time.
Some believe it causes microfractures in the iron plating, but that’s bullshit both by experience and chemistry/physics. I do also have copper mesh stuff when i have to scrub.
All tips should be iron plated, you can re-electroplate yours with ferric chloride (which you might already have from pcbs making)
All tips should be made of copper, which is easy to file down to give it more life. You can re-electroplate it after that if you’re that kind of guy.
It’s not any different than a woodworker taking care of his chisels.
Buying a new tip every time you get a pit in it is insane.

I used to have issues with soldering and soldering accessories. Then I bought a 888D. No issues since then. Best tool I've ever bought.

are there any rules of thumb for how much power a transformer can handle? i need a SMPS that can drop 24V input to 4V output @ 2A, so 8 watts. i'd like to use a flyback topology. i have some ee10 and ee25 transformers on hand which use pc40 magnetic material, but im not sure what size (or larger) i need to support 8 watts.

>>2614088
Once they catch fire you cannot put the fire out, or move it easily. I keep a co2 extinguisher and store them in a metal,box. Hopefully if they go off i will try and get the metal box outdoors quickly.

i want to be able to self monitor my alarm system via basic iot. i have a honeywell ademco vista 15p and it has the ability to dial and send basic messages to pagers (cant believe this is still a thing, and the system is only 5 years old too) via a phone line about the status, like if its alarming or if there is a fault. does anyone know of a gizmo that can read the tip/ring output from the phone line terminals on the board, decipher it and turn it into a sms, and text it to a phone number over the internet?
i would greatly prefer an off the shelf solution, im sure it could be done with an arduino or something but im not a programmer and would rather not learn on a seemingly complicated project like this. also obviously not one with a home server that will render it useless when the company folds in a year.

I want to make the worlds shittiest diy oscilloscope. I have a chink knockoff arduino nano lying around. So basically all i have to do is
>Scale input to 0-3 V
>Feed into Analog pin
>Send numbers to computer

One thing i am worried about is a spike of voltage fucking up the board. Will slapping on a zener diode protect from overvoltage or do i need to make something more complex?

>>2612520
Then why don't birds get shocked from sitting on power lines?

>>2615173
In the case of an arduino, you can set the ADC full range to be significantly less than your power supply range (e.g. a 1V or 2V reference) and rely on diodes to clip any spikes. Just make sure the diodes can react fast enough, you may be best off using a few 1N4148s in series for the speed but their I/V curve isn't as vertical as you may want. Personally I'd have that protection (and a series resistor before the diodes) between the ADC input and the gain stage, and use additional protection for the higher voltage analogue input stage. Remember you're not just protecting your arduino, but your entire computer.

>>2615184
thanks for your insight anon. Sounds pretty close to my plan
>Remember you're not just protecting your arduino, but your entire computer.
I did not consider this. Thanks for the warning.

>>2615189
There do exist analogue optical isolators for before the ADCs if you're paranoid, but they're really not that linear. It would be easier to put optocouplers between an external ADC and the MCU, or to use an optically isolating USB cable.

>>2614871
>All tips should be iron plated
They're nickel plated, retard-bro. Nickel is the only thing that sticks well to copper in electroplating.
If you want to do iron plating you need a few nickel layers first. The nickel does the job just fine by itself though so I can assure you that no manufacturer will waste money to iron plate tips.

>>2615110
>text to a phone
>over the internet
anon i....

>>2613706
Thanks for the help. The thing that's fucking with me the most is how the initial problem transient and related to a loss of power event, which to me suggests a broken cap that fully discharged. Also as you said the output is kinda noisy.

I don't have a scope so I'm pretty limited in by testing. It's gonna be impossible to find a short somewhere, but if there was I have a hunch it will be in one of the chips.

The chip in my pic is an AP1510, which does have switching components. I might just order a replacement off ebay and see what happens. I've never done surface mount components so at least this should be educational.

>>2613886
>You know those results kinda look like what you’d get if you plug a 110V appliance into 220V.
Nah its one of those universal supplies >>2614329. I think I was just retarded and reading DC as AC.

First serious board design using strip/vero-board. Think it turned out decently, though it's heavily dependant on being able to run jumpers beneath a DIP14 socket. Will probably need to change some parts, as on one part I've got a board-mounted connector for the laser but I've just got wires dangling out for the 12V PSU and CNC input. I'd kinda rather put the board in a case and have external connectors for it, though I'm not sure where to fix that box yet. Ideally on one side of my gantry.

>>2615231
It's some simple digital logic to combine the PWM signals from the pilot oscillator and from the CNC mainboard, but only output while the (currently upside-down) laser has a 12V PSU connected. That took 3 gates, while the 4th is used as the oscillator to give a minimal duty-cycle. If I plug 12V into it without the 3.3V CNC machine power coming in, nothing bad happens either, thanks to this clever MOSFET circuit.

Coulda used an MCU, but I think this method is cooler.

>>2615218
yes? you what?

>>2615245
mobile network for sms isnt the internet. its a closed system

>>2615251
ok have it your way
"does anyone know of a gizmo that can read the tip/ring output from the phone line terminals on the board, decipher it and email it?"

>>2615216
No, my smooth brained non-googling troll.
Fe is the main plating material, and has been since the 40s or 50’s when they figured all this out.

Picrel is from Hakko, the first search result. Hakko is a soldering iron and tip manufacturer FYI. I posted the answer from their site since you can’t use internet search.

Don’t forget to tell Hakko they’re “doing it wrong” after you figure out e-mail.

>>2612058
Here are some common components I think you should get:
> 100 of each round resistor value (100, 1k, 10k, 100k, 1M) in 1% tolerance (5% is ok if you can't afford 1%)
> same with ceramic caps (100p, 1n, 10n, 100n, 1u)
> buy some opamps in DIP8, double circuit. I recommend the lm358, the tl072/tl082 or the MCP6002. Or all three.
> If you want to do medium power discrete circuits, you should get some transitors. I recommend the pn2222a (NPN) and the pn2907a (PNP) for bipolar transistors. You sould get a hundred of each.
> get some good quality breadboards, a big one and a few small ones
> get some protoboards.

You don't have to buy all at once. Whenever you buy stuff for a project, get extra quatity for the common stuff. Also, get some good equipment; get a dmm, a power supply, an oscilloscope and a function generator (in that order). A microcontroller board can substitute for the scope and the function generator for a while, but it won't replace them.

>>2614622
You shouldn't use voltage dividers for power applications. Fristly, the dissipate a lot of power continuously, even when no load is using current. Secondly, the rail you're creating will drop a lot once there is any significant load on it. You need to use a linear regulator, either get one (lm317, lm78xx) or build one (search discrete linear regulator). The simplest solution to choose a cap, is to stick to your regulator's datasheet recommendation (that's what most engineers do most of the time).

>>2615284
>The simplest solution to choose a cap, is to stick to your regulator's datasheet recommendation

this part of the post is nonsense.
caps are chosen based on the desired ripple, which is determined by load current.
the manufacturer has no clue what current your load is gonna use, so they cant make a recommendation.
rule of thumb is 2000uF per amp of load current for 50/60Hz stuff.

>>2615343
After a rectifier bridge, yes. After a regulator, just use the recommended value, it's not for ripple rejection, it's for bulk decoupling.

This AI is full of shit. Sad.

>>2615432
AI is going to make people even more retarded than they already are. Maybe that's the AI apocalypse we were warned about. It's Dunning-Kruger on steroids.

>>2615433
I bet the masses are not going to use it anyway. Except for dumb students who will get burned on shit like this and will give up.
And those who already have some knowledge will just add it to the list of tools to use, with known caveats etc.

>>2615443
>I bet the masses are not going to use it anyway.
The biggest developers of AI are trying to make it into GoogleV2.0, so almost everyone would be using that over "dumb" search engines. The Bing AI thing already looks things up online for you, and that's how most models will work in the future. Probably.

Bought a Wavetek 288 function generator lads, man this things a beast.

>>2614283

Just get a used Hakko 936, the thing can take all sorts of abuse and has served me well since I first got into soldering back in 2007.

>>2611877
>want to make an MCU controlled buck converter so that i don't have to waste money on breakout boards
>open YouTube to learn more about the subject (hobbyist, so no professional background)
>long ass math equations about inductor current etc
>"transfer function"
>"phase margin"
>"Laplace transform"
>"process plant diagram"
Damn what the fuck i just want to power my nrf24 radio

>>2615224
>I might just order a replacement off ebay
Don't do that just yet. It looks like you have a short on your supply, if you don't fix it a replacement chip won't do anything. Furthermore, from your circuit diagram, it looks like there is a temp sensor built into the chip, so it's unlikely it got burned unless there was some kind of voltage spike that fried it. I recommend a bit more testing, look at capacitors, they might actually be shorted. Focus on electrolytic ones or anything that looks fried. You shoudl also be able to probe stuff around the chip and see if you're geting the correct inputs and outputs. If you don't have the correct inputs, then it's normal for the outputs to be wrong as well.

>>2615483
got the wavetek 182a

>>2615496
Most of that is required knowledge if you want to design your own switching regulator.
>Damn what the fuck i just want to power my nrf24 radio
Why do you want to build your own regulator then? You can find so many ready made buck converters for cheap. Also why the MCU control? Do you need to change output voltages often?

>>2615533
>Also why the MCU control?

if your TV remote, toothbrush, and rectal thermometer dont have a computer in them, how can you even show your face in public?

Digital logic is cool and good but I have a much harder time with digital signal processing or discrete random signals. What can I do to be better? Pls no more oppenheim

>>2615443
It’s harder to weed out all the subtle lies because the algorithm works off the markov chain probability for the next word based on the training set, so everything has the appearance of a confident source of information with perfect diction and grammar. Imagine an all-reddit/all-twitter set.

It’s already a problem that AI training sets are already feeding off themselves. It’s like taking a xerox copy of a xerox copy… eventually it will be either all black or all white.

>>2615533
0 points for not using the Black regulator.
Roman black developed it a while back as a minimalist switching regulator that’s easy to fix, is cheap, and works well.

test

>>2615110
https://forum.arduino.cc/t/solved-serial-comm-with-security-panel/246301/3

Remember LED autim from 2021?
It is time to autism again.
Mainly because OSRAM have released CRI 90 195 lm/w QD LEDs and I will order some, but also because I found Sunlike knock-off LEDs from Aliexpress where seller used Seoul Semiconductor shilling material but Seoul Semiconductor doesn't make SunLike LEDs in 2835. And I have around 250 OSRAM JELMS2 4014 LEDs left. Oh, and I got a new work bench light which came with awful T8 LED tubes that are useless, and inefficient as fuck as CRI of them is absolte dogshit (80 CRI OSRAM JELMS2 outperfroms it in terms of CRI) and they flicker. So I want to manufacture my own lights, again, with LEDs I want.

This time I'd do the different autism. First of all, I've finally read a thermal design manual from OSRAM which basically tells that LEDs dont convert all energy into heat, but rather you can calculate it yourself given formula, bla-bla... In short,
>120 lm/W -> 60% heat
>200 lm/W -> 50% heat
Which allows me to run LEDs hotter at higher current this time, especially since QD-LED can't be jewified (lm/w is more-less constant in its case). And China full spectrum LED (supposedly Honglitronic job, but seller didnt say anything) has piss poor efficiency anyway.
Next innovation is power supply. It is still a capacitive dropper, but this time with 2 x 470uF 100V capacitors which do form a low pass filter which gets rid of nasty 100Hz flicker. It also comes with benefit of having a power factor of 0.25, to make my power company happy.

So, question is, did Zheng chink me with high-CRI LEDs? Did anyone order noname "sunlike" LEDs?

In case you're wondering what did happen to my previous light - nothing, still works.

>>2615644
Use an inductor instead of the 56Ω resistor.

>>2615656
I could, but consider following. Only 165.79mW are lost on the resistor, which means overall efficiency is about 96-98%. Also in the end I decided to use 47Ω.
In order to cut off 100Hz, your filter cut off freq. has to be like a decade lower, so 10Hz. So inductor has to be quite big.

>>2615666
>overall efficiency is about 96-98%
Such a low drop across your resistors/capacitors implies two negative things:
>a change from 220Vrms to 225Vrms or 230Vrms will significantly change brightness and possibly overdrive your LEDs
Definitely simulate that.
>the NTC of the LEDs may not be properly be compensated for, leaving you vulnerable to thermal runaway
Not sure about this one being an issue, but you can try modelling your LEDs with 10% lower forward voltages to see if their power consumption increases or decreases.

Also 10Ω means 22A peak worst case, in which case you'd dump ~10J into it. A rudimentary calculation suggests that 10J will explode such a small resistor. Start your sim with a 90° phase shift to your AC input and see what happens.

Does a 1kw electric oil radiator actually use 1kwh, or does it use less?

My gut is telling me it cycles on and off and uses less than that depending on the ambient temperature and how hot the safety sensor gets. But I'm not an electrician and so I doubt myself, and I thought I'd see what you guys thought.

>Why do you want to know this, anon? What's the reason?
I'm a computer science student, and essentially wondering if I should just max out my GPU doing useful tasks instead. If it's putting out more or less the same amount of heat, it may be cost efficient.

>>2615496
>MCU
Fuck, can I escape that marvel shit anywhere?

>>2615533
>Why do you want to build your own regulator then?
Because it's cheaper to buy two mosfets and an inductor than to buy an inductor and a specialised buck IC. Plus i want to do it once just for experience

I've seen a few circuits where a voltage to frequency converter is used to convert an analog value to a frequency for transmitting through an optocouple and then a frequency to voltage converter is used on the other side to change it back.

I have a board now that I know takes in an analog voltage between 0 and 10 and uses it to control a speed of a motor.
I followed the signal to a small 6 pin chip that is also connected to an optoisolator so I assumed it was the same thing but it appears to be giving out a 2KHz wave and the duty cycle is varying depending on the voltage.
I can't for the life of me find what the IC is (pic related, it's an SOT23-6, Marking says A0B2 although the last two are different across the few boards I have so I imagine they're a batch number and only the A0 is relevant).

What do you even call a chip that does this?

There's a dead short in this hilti battery charger I own, guessing to start at the psu section and what are some other components that fail in these things? Although I'm decent at soldering can't say the same for my troubleshooting skills, in this case would injecting a signal into the charger with my function generator and tracing with my o scope be appropriate?

>>2615817
Probably less, I'm pretty sure they have a thermostat in them as opposed to a phase-fired dimmer. But how often they stay off is really dependant on how quickly head is convected away. For a large or leaky room with some airflow, the thing may never turn itself off. Using GPUs for heating is by no means a new idea, IIRC bunkerbro was mining crypto to heat his bunker.

>>2615835
Use ASICs everywhere. Watch out for ASICs that are actually MCUs in disguise, like the FT232. They glow in the dark.

>>2615836
>it's cheaper to buy two mosfets and an inductor than to buy an inductor and a specialised buck IC
Buy a fuckton of TL494s and learn to use them. It's good practice since they're very versatile (albeit outdated) chips, somehow they're still the closest thing to a jellybean SMPS IC. You can use the integrated transistors for low power buck or boost style circuits, while pairing them with the right FET(s) and gate drive circuitry can easily get a 500W converter. They have dual feedback loops built into them with ports for external references, so it's trivial to make a CC/CV supply or a supply with temperature or other analogue regulation.

>>2615855
It may just be a single comparator if there's a triangle wave coming from elsewhere. Otherwise, I imagine it's an integrated PWM oscillator, similar to a 555 but more specialised. Maybe even a repurposed SMPS PWM controller. Chances are the PWM isn't even being turned back to an analogue value, but rather being fed directly to the motor controlling MOSFET.

>>2615860
Test the power transistors and diodes with your DMM. If that fails, mock up a quick milliohm meter using a 9V battery, an LM317, and 4 alligator clips, to try and trace the short. When using seperate clips instead of kelvin clips, your voltage sensing clips need to be closer together than the current emitting clips.

>>2615862
>Test the power transistors and diodes with your DMM. If that fails, mock up a quick milliohm meter using a 9V battery, an LM317, and 4 alligator clips, to try and trace the short. When using seperate clips instead of kelvin clips, your voltage sensing clips need to be closer together than the current emitting clips.

I'll def do that, thanks but also you should know there's a big RF Choke where it looks like the trace literally blew apart like something catastrophic happened, how would I test the coil without an impedance meter or something that reads inductance?

>>2615862
>Chances are the PWM isn't even being turned back to an analogue value, but rather being fed directly to the motor controlling MOSFET

I'm sure it is, I haven't gotten to that side of the board yet.
I was just mentioning the voltage to frequency and frequency to voltage circuits I'd seen, usually for temperature measurement on high power IGBTs.

This board is just for controlling a big fan so yeah, probably the signal doesn't get changed back to analog

>>2615864
If there's a blown trace then you may well find that applying your function generator to the inductor doesn't end up putting a voltage across anything else. I'd give it a shot, send it some different frequency sine waves with a known series resistor and measure the voltage across the resistor and inductor to determine the -3dB frequency, or the 45° phase point. Ideally you have the ground clip between the components and measure just the inductor voltage with one channel and just the resistor voltage (i.e. the current) with the other channel, but that requires that your fgen is isolated from your scope ground. Or that your scope has a subtraction mode. It's even easier to find the inductance if you put a cap in parallel with the inductor. Or maybe you can read off an inductance value if you squint.

I'm guessing the choke is fine, a blown trace suggests that was the weakest link and the fault happened really fucking fast. Dead inductors usually happen from insulation overheating, which is a lot slower. Not like the wire inside it couldn't explode, but I'm guessing the resistivity of the magnet wire and the trace were of different magnitudes, and only the higher one cooked off.
Where in the circuit is the choke? If it's in series with the power transistor, it could be that the transistor dying short-circuit caused the overcurrent event. If it's on the input then it could be anything.

>>2615865
>usually for temperature measurement on high power IGBTs
Oh that's where you were seeing them. I've never seen such a circuit before, but that's a situation where it makes sense. Wonder if it's easy to run 1-wire through an opto? Some cheap digital temperature sensors use 1-wire. SPI would be a lot easier in this kind of situation.

>>2615876
IGBTs usually have an NTC thermistor built onto the substrate so you generally just use a voltage divider to get an analog voltage back.
Pic related is from a board that drives three IGBTs

>>2615836
LM2596 ICs are dirt cheap
https://www.aliexpress.com/item/1005001582744801.html
XL6009s are also fairly cheap if you want to dip into boost converters too.
https://www.aliexpress.com/item/32220046693.html
There's also the option of using TL494s as another anon mentioned above.
I'd say using either of these is preferable to dedicating an arduino to the task.

>i want to do it once just for experience
That's fine but you shouldn't trust whatever you build blindly. Ideally you should have a scope and test it under load to check for excessive noise and/or voltage spikes before attempting to power anything with it.
I still recommend you get a ready built buck board to power that radio so you don't risk damaging it.

>>2615855
>>2615862
So my mystery chip is wired like this.
Strangely, there's an unpopulated 5 pin header connected to it which makes me think it's an ICSP header for programming.

Not sure about the transistor / zener on the right, constant current source or something?

>>2615892
I think the transistor pin 3 changes voltage when supply voltage gets above 4V or so. I'd guess it to be an MCU judging by the header, probably an OTP jobbie like a padauk.

>>2615907
I'd say it's a microcontroller myself.
Can't be many ones in SOT-23 that would have an ADC and a capture / compare module

>>2615814
>Definitely simulate that.
There is a slight change in brightness, but I made sure that at 265V (highest you can get from socket) current through LEDs isn't insane.
>Not sure about this one being an issue, but you can try modelling your LEDs with 10% lower forward voltages to see if their power consumption increases or decreases.
Nothing would really happen, because capacitive dropper kinda acts like constant current source. You can even add or remove couple LEDs from schematic and it would work more-less the same.
>Also 10Ω means 22A peak worst case, in which case you'd dump ~10J into it. A rudimentary calculation suggests that 10J will explode such a small resistor. Start your sim with a 90° phase shift to your AC input and see what happens.
If there were nothing on the input - yes.
But there is a capacitive dropper, which will limit curent to around 100mA. In worst case scenario when capacitors are charged to -310V and you connect it at the peak of sinewave, capacitors dont store enough energy to blow anything. But still it is a good idea to put a discharge resistor in parallel.

>>2615912
Droppers are the worst for LEDs. I hate droppers.

>>2615914
I dont care. All I care is that I get 40-80W of light, power grid gets 15VA of load, and I pay just for 4W.

>>2615642
well if that isn't exactly what I'm looking for with most of the work already done for me

>>2615817
>wondering if I should just max out my GPU doing useful tasks instead.

no, silly boy
the replacement cost of a new computer is $1200
the replacement cost of a new radiator is $60

>>2615817
>wondering if I should just max out my GPU doing useful tasks instead.

no, silly boy
the replacement cost of a new computer is $1200
the replacement cost of a new radiator is $60

>I'm a computer science student

oh, so for you, that's 2260 and 74 in octal
which is a ratio of 10100 to one in binary

>>2615817
Just run both, if the GPU output is enough the radiator won't run, if it's not enough then the radiator turns on so you stay warm, ultimately all energy used will go to heat from either device.
Heat pumps are the exception.

>>2615945
but the radiator will heat up the GPU and then you have a thermal centipede situation on your hands.

>>2615942
>>the replacement cost of a new radiator is $60
sus

>>2615961
>>$60
>sus

$70 canadian is like $13.50 USD

>>2615967
it has to be at least $13.75 by now.

>>2615836
look up biricha digital on YouTube. he has a free software for stm32 that does the calculations and code generation for you. his channel even has a start to finish video on generating the code, flashing it, and measuring the output with a scope.

>>2615836
>>2615974
https://www.youtube.com/watch?v=a64UIK1dFXA

fwiw i have an stm32g474 nucleo (i think) that's sitting in my parts box because i also want to make an MCU controlled buck converter. the math is definitely beyond me, but the software tool seems to do basically everything for you.

and at the same time i've also found myself learning how to build analog buck controllers. there's integrated solutions like >>2615887 says, which are simple to work with because the data sheet gives you basically everything, included a suggested layout. and there's more general PWM ICs like >>2615862 says, which require more knowledge and planning for design and layout, but can be extremely cheap. i bought some chink brand uc38xx ICs on LCSC for like $0.25 each.

Can i run this with a 3 phase inverter that has an output of 3x230V?
Im confused whether this is a 200 or 150 V motor. Can any germans here enlighten me what Erreg. is?

>>2615912
>But there is a capacitive dropper, which will limit curent to around 100mA
Only for 50Hz. For a really high frequency spike, like what happens when you flick the lamp on in the middle of an AC cycle, it goes straight through the capacitor. That's what the series resistor is for in the first place, otherwise you'd ditch it and just use a capacitor. As I say, simulate it with a 90° offset to the AC source.

Also add a discharge resistor, touching the prongs of a recently unplugged cap dropper can give you a bit of a buzz.

>>2616073
If I had to guess? No. It's a 150V motor.

>Conductive tape, the tape only conducts through the Z axis, not along its length
Huh, neat

>>2616099
conductive threads interwoven with non-conductive threads, backed by a plastic film and adhesive?

does it make sense to have a power plane on the top layer to connect all the VCC pins together similar to the ground plane on the bottom layer? or is it better to just connect VCC pins together with tracks?

>>2616345
*traces

>>2616345
It’s bad for PSRR, since traces are going to capacitively couple to both rails. That said, if you already have a low-noise power supply, or don’t need high PSRR due to it being a digital circuit, it can be a neat shortcut to getting low impedance throughout your board. But I’d look into why people use star grounding to decide whether a full retard ground plane is the way to go.

>>2616408
>be me, buy pic rel for some copper plate soldering
>little fiddly but mostly does the job
>pick it up again a week later to do another bit.
>tip doesnt wet, solder rolling off, tip looks pitted, pain to get enough heat into the copper without a wetted tip
>maybe should have wetted the tip before i put it away? w/e, i'll just buy another one
>new one arrives, works fine first time, wetted the tip before i put it away, pull it out to finish the job 24 hrs later and tip isn;t wettable
no amount of wire and fux cleaning is bringing it back to life. wtf am i doing wrong? or does this cheap tip just destroy itself after one use? It's not temp controller so i can get glowing red. wondering if thats the issue.
on a side note, anyone got recommendations for a good 100w+ soldering iron? my fx888d isn;t up to the job.

>>2616408
didn't mean to (you), sorry

>>2616591
>recommendations for a good 100w+ soldering iron? my fx888d isn;t up to the job
What exactly are you soldering?

>>2616593
copper sheet, 0.6mm onto 0.3 nickel strip thats spot welded onto a battery, so needs to have enough mass to quickly dump it into the copper plate but without heating up and damaging the cell. picrel is the end of the tip after one use.
don't mind paying a few quid for a quality one but don't want to go overboard as it's probably going to be only used a handful of times again, if any.

Do you guys know of any diodes that could allow 12-24 volts forward and block 10kv backwards?

>>2616591
>a good 100w+ soldering iron?

any 100W iron is gonna destroy tips.
even good tips like Weller or Hakko.
that much heat will ''carbonize'' the surface.
so, live with it.
anyway, you can always file away all the black crud, get back to bare metal and tin it.
that'll give a few minutes of usage before you have to do it all over.

>>2616598
that's what i was thinking, although no amount of cleaning will make it wettable again, no wire brush, wet sponge, flux or even sandpaper.
guess i'll just need to squash the solder on, or try another iron that has temp control.

>>2616595
https://www.amazon.com/HAKKO-soldering-processing-HAKKO-MATCHLESS-571/dp/B002MRR6R4

You should season the tip before you use it, and always tin before you power it down.

>>2616602
seen that one, but it's £191 here. probably just going to buy 5 of the £10 quick cheap ones and throw them out after each session, lol.

I have a board that has an expensive (€15 odd) and difficult to get quad op amp on it.
It has a 8MHz bandwidth and 20V slew rate.
The four circuits in it are used as in pic related.
First one seems to just be a comparator with schottky diodes to clamp the inputs.

Second and third seem to just be precision rectifiers followed by low pass filters (480 and 4.8KHz cut off)

I imagine it's the fourth circuit that needs the higher bandwidth or slew rate?
It's basically driving a linear optocoupler so the feedback loop is through the LED and photodiode.
Would the high bandwidth be needed for this and if so how do I calculate the required bandwidth?
Hopefully it's overkill and I can replace it with something easier to get.
The current in the LED to the current in the photodiode is 0.007

>>>/vg/429688447
Artificial Academy 2 General /aa2g/ #1279
Who Framed /aa2g/? Edition

Welcome, this general is for the discussion of ILLUSION's Artificial Academy 2.

COPY ERROR MESSAGES WITH CTRL+C, PASTE THEM WITH CTRL+V INTO GOOGLE TRANSLATE. JUST CLICK THE WINDOW AND PRESS CTRL + C, IT WORKS.

>Downloads:
/aa2g/ Pre-Installed Game, AA2Mini: https://tsukiyo.me/AAA/AA2MiniPPX.xml
AAUnlimited updates: https://github.com/aa2g/AA2Unlimited/releases
Anon's Modded Pre-Install: https://pastebin.com/42JS3q6E

>Information:
AA2Mini Install Guide:
https://docs.google.com/document/d/e/2PACX-1vS8Ap6CrmSNXRsKG9jsIMqHYuHM3Cfs5qE5nX6iIgfzLlcWnmiwzmOrp27ytEMX03lFNRR7U5UXJalA/pub
General FAQ:
https://web.archive.org/web/20200216045726/https://pastebin.com/bhrA6iGx
AAU Guide and Resources (Modules, Tans, Props, Poses, and More):
https://docs.google.com/spreadsheets/d/17qb1X0oOdMKU4OIDp8AfFdLtl5y_4jeOOQfPQ2F-PKQ/edit#gid=0

>Character Cards [Database], now with a list of every NonOC in the megas:
https://docs.google.com/spreadsheets/d/1niC6g-Xd2a2yaY98NBFdAXnURi4ly2-lKty69rkQbJ0/edit#gid=2085826690
https://db.bepis.moe/aa2/

>Mods & More:
Mods for AAU/AA2Mini (ppx format, the mediafire has everything):
https://www.mediafire.com/folder/vwrmdohus4vhh/Mods
/aa2g/ Modding Reference Guide (Slot lists for Hair/Clothes/Faces, List Guides, and More):
https://docs.google.com/spreadsheets/d/1gwmoVpKuSuF0PtEPLEB17eK_dexPaKU106ShZEpBLhg/edit#gid=1751233129
Booru: https://aau.booru.org

>HELP! I have a Nvidia card and my game crashes on startup!
Try the dgVoodoo option in the new win10fix settings.
Alternative: Update your AAU and see if it happens again. If so, disable win10fix, enable wined3d and software vertex processing.
>HELP! Required Windows 11 update broke things!
winkey+R -> ms-settings:developers -> Terminal=Windows Console Host

Previous Thread:
>>>/vg/428858839 are you a pussy ass nigger no then come play

>>2616610
god, i really hate these fucking bots. gookmoot needs to do something about it.

>>2616597
> block 10kv with a diode
HV diodes exist, e.g. microwave ovens. Multiple diodes in series increase the voltage at the expense of higher voltage drops when forward biased.

Maybe you should consider an optoisolator.

>>2616606
>it's the fourth circuit that needs the higher bandwidth or slew rate?

seriously doubt it.
that 150pF in the feedback loop is meant to cut off any high frequencies.
starts attenuating at around 100Khz.
at 8Mhz it's attenuating 750x.

>>2615862
> Outdated TL494s
I’ve got several SMPSs that still use a μΑ723, including a modern PC power supply. Maybe anon should start with that civilization-creating building block and move on from there.

Afternoon, /ohm/. I was the clueless Anon that posted a few days back about wanting a soldering iron. Just wanted to say thanks for all the replies and advice you Anons gave me. Ended up getting a cheap iron off of Walmart with plans on doing proper research and buying a proper one in the future. https://www.walmart.com/ip/SREMTCH-10-In-1-Portable-60W-Soldering-Iron-Kit-Heat-Pencil-Repair-Soldering-Tool-Blue/516044715 if anyone is interested in the model or at laughing at me for buying a piece of chinkshit. Again, thanks Anons. You guys are great and helped me a lot.

I have a question. When I read about LiFePo4 batteries it always says the charging current is around 1C. What does it really mean? I've looked at the meaning of C-Rate of a battery, but there's something wrong. Typical Tesla LiFePo battery capacity is around 26.000 Ah, so if I calculate the charging current I get
26.000Ah / 1C = 26000 Amps, which is obviously bullshit since 26000A x 400V = 10.400kW.

So what am I missing?

>>2616652
>1C
Capacity of the battery (or cell) in mAh divided by 1000.

26000mAh / 1000 = 26A

>>2616659
No, it's not 26000mAh lol, it is 26.000Ah like I said.

>>2616667
Is that a European comma or a standard decimal?
26,000Ah at 1C is 26,000A current.

>>2616668
Anon it's a fucking Tesla battery pack, how can you think it's a decimal point? Do you even know how little 26Ah is? Typical car battery is 70Ah lol.

Why would you answer then if you don't understand yourself.

>>2616670
Again, 26,000Ah at 1C = 26,000A current. Any other questions?

>>2616671
Yeah, it's bullshit like I said. Waiting for someone else to help.

>>2616672
You ever heard of fractions?

>>2616652
>So what am I missing?
braincells, please drink the contents under your sink and remove yourself from the gene pool.

>>2616591
Tin with zinc chloride plumbing flux

Bump >>2616652

Have anyone ordered PCB with lewd silkscreen? Do chinks even snow allow it?

>>2616671
> 26000A
What AWG wire would you need to run for that kind of ampacity?

>>2616703
>AWG
No- you'd need bus bars, a welder's helmet and a rubber suit.

Can someone explain me one thing? How does DJI video transmitter have better radio penetration, than video transmitters from other manufacturers, while using the same RF power and the same frequencies? How is this possible?

>>2616599
100W irons will creep up above like 400C without a load on them. If you get something temperature regulated that won't happen.

>>2616729
Different modulation?

>>2616820
Maybe they use zmodem protocol. lmao

>>2616652
the cells aren't all in parallel.

>The P85 pack contains 7,104 lithium-ion battery cells in 16 modules wired in series (14 in the flat section and two stacked on the front).[88] Each module contains 6 groups of 74 cells[89] wired in parallel; the 6 groups are then wired in series within the module.[89][90][91][92]

74P x 3.4AH/cell = ~250AH
6S x 4.2V/cell = ~25V
6S74P = 25V, 250AH
16 modules in series = 96S74P effective configuration = 400V, 250AH = 100KW

V3 supercharger can do up to 325KW at 400V output, or 812A peak.

>>2616848
Thank you anon, now it makes sence for me.

How do they choose what voltage and current to make a motor for? In this case they have chosen 400V 250A, but they could go 200V 500A by rewiring batteries. How would you choose that if you'd have to put an electric motor with batteries on a buggy vehicle or a motorcycle for example?

>>2616859
i'm not an engineer, just a hobbyist, and not even a skilled hobbyist, but...

P=I^2*R
P=V*I

for a given power output, you want as high a voltage as possible to limit I^2*R losses. in general there are practical ceilings on voltage due to things like insulation requirements or semiconductor ratings.

learning to route.
what do you think? i poured the entire 3.3V copper plane and then stitched 3 islands together with a couple of cross under traces. is that acceptable? do you notice any obvious issues overall?

>>2616870
>any obvious issues

where is your anime gurl?

>>2616683
> tin with acid flux
Good idea, also you could tin the tip using one of the lead-free solder formulations, with silver and/or copper in the alloy—they probably don’t dissolve the Cu tip as much.
The flux I have also has solder paste already in it, too.

>>2616870
do you have a schematic? assuming it's an MCU but it would be helpful to see what else is on the board.

>>2616862
That’s right; also the lower the voltage, the thinner the wire you can use (and less losses) for the same power.

Most common PVC insulation for north american wire is rated for around 300V, so 400V is is quite an achievement.

You can see this same trend in cordless power tools, like the 40V battery powered lawnmower.
Theoretically, a 3.7V lawnmower would be equally powerful if all cells were wired in parallel, but you’d need 10 times the amperage, and, hence, really thick wire to carry it.

>>2616896
it is work in progress, it is an MCU and a sensor, some of it are just random connections, i just wanted to test KiCad functionality.

>>2616687
No but I’m considering minting some cute and funny PCBs. Ask the ham radio thread too, they post lewd images on SSTV.

>>2616870
Is that sideways(?) connector on the right side of the board J3? Looks like plugging something into it would cause the connector to scrape by U1.

>>2616906
Add plugs or footprints around the op-amp for building negative feedback loops with resistors and capacitors.

>>2616906
J3 is a screw terminal, I was thinking something like this but KiCad doesn't have a 3d model for it, and I am not even sure if I picked the right footprint. So I am not sure how it should be oriented. I am going to search for a 3d model for it or maybe also together with a footprint to be sure it fits properly.

>>2616870
dont use thin traces without reason.

>>2616894
I think lead-free dissolves copper much faster, as lead doesn't form an alloy with copper. Dunno.
Lead-free is junk anyway, all alloys are like Linux distros and communism, sure SAC305 would work, wait no it is Sn99.99NiGe+ trust me flows just as good as leaded solder. Better leave it for plumbers to solder pipes with potable water.
I've got a roll for cheap of Ag0.3 Cu0.7 or something like this with soldering iron handpiece (in combo), it is usable but eh, doesnt feel as nice as 60/40.
As for flux, it is a good idea to drop couple solder blobs in it.
Other thing that fucks up tips is wet sponge, because you should use distilled water, but nobody does and you get limescale on tip which is annoying. It is much better to use dry toilet paper to clean tips, or brass wool.
>>2616078
I did specify initial conditions of caps to be charged and specified AC to be out of phase, and inrush was 6A or so but it is so short it can't really heat anything up and it would get absorbed by those 470 uF caps I think. But then idk, electrolytics caps suck for high frequency.
>>2616926
Shitting in the air is not a rocket science. Nobody would even try finding you, unless you SSTV over police or idk ATC band.

Anyway, ordered all components, PCBs. I still wonder if LEDs I bought from aliexpress are legit deal (as they put small chinese dick on korean patents) or just chink scam.

I have a question /ohm/. What piece of shi- PCB CAD software do you use?
https://strawpoll.com/polls/NPgxEom24Z2

>>2616870
That 10 pin connector
• Don’t put V+ and ground next to each other.
• Rework the pinout so traces go directly to your chip without looping around things like gnd and +3.3v
>>2616962
> thin traces == no
True. Make sure you get ALL your free copper.

>>2616971
> get absorbed by 470 uF
Put TVS diode or MOV to protect

> distilled water
Good point

>>2616928
Those are the MetzConnect Type055 / RT015xx

>>2616971
>Lead-free is junk anyway
Try the tin-bismuth solder alloys, they're brilliant. Probably not as structural as lead-tin, but it flows just as nicely and is less stress on sensitive components.
>6A
That's probably fine, assuming that fuse is slow-blow. What the hell are those capacitor footprints anyhow? Just SMT pads for soldering standard radial caps to?

>>2615876
>but that requires that your fgen is isolated from your scope ground

where would I place the scope ground on the circuit?

>>2617022

>>2617015
> bismuth
Also good for belly aches

>>2617048
truly the best "heavy metal"

>>2615231
ouch oof ah the feature creep

>>2616906
you'll want decoupling on your op amp and BME280. like someone else said, add feedback traces/pads for your op amp. you can leave them NC/not connected if you don't want them.

i'm just a really shitty amateur but the placement of your op amp might not be optimal, since you're running its output relatively close to """high speed""" digital signals on U3/BME280.

your LDO needs input and output capacitance but it appears to only have one of them. read the data sheet for your specific model. it may or may not support ceramic capacitors. if it doesn't support ceramic capacitors, you can work around this by adding a series resistor to the capacitor, to mimic the ESR of a tantalum capacitor.

Herro Anon.
I'm trying to repair a headphone cable, but can't get the inbuilt button controls to work.

What does this white wire do?

>>2617103
>What

you have 4 parts on the plug, which means you have 4 solder points.
find the fourth point and solder it there.

retards at the shop broke my dashcam
what would it take to build one myself? i think most of these are too expensive for what they offer. besides i do not want to install some chink shit app just to get the video from them.

>>2617109
Webcam + SBC + MotionEyeOS

>>2617015
>That's probably fine, assuming that fuse is slow-blow.
It lasts much less than a half cycle.
> What the hell are those capacitor footprints anyhow?
Proprietary footprint. Basically SMD pads big enough to solder THT. Originally I wanted to mount only capacitors this way, but then I checked price of 1/4W SMD resistors and THT resistors were cheaper, so I decided to use this for resistors too.
>Try the tin-bismuth solder alloys, they're brilliant. Probably not as structural as lead-tin, but it flows just as nicely and is less stress on sensitive components.
Idk. I never used one. But consider following. Bismuth is 4x more expensive than lead. And im not restricted by RoHS so lol.

Which of these capacitors are on the audio signal path on this preamp?

>>2617002
>10 pin connector
Well in this case I probably could since it is just a test circuit but typically I don't have much control over the pinout since it has to match the cable connector that plugs into it. For example, here is a stand J-TAG connector. It differs a bit from the one I am using, but VDD is still right next to GND and there is no way to change that unless you make a custom cable. I agree, the pinout is terrible for routing. There are always some parts that are awkward to route no matter how you change the pinout (like 555)

>dont use thin traces without reason.
>Make sure you get ALL your free copper.
Yeah I agree, I was just following some tutorials (Eric Bogatin's class series, really like them) and set the defaults for different netclasses: 6mil for signal traces and 20mil for power traces and 25/13mil for vias. However I found that 20mil could be awkward to connect to smallish pins and also 20mil sounds like an overkill for low power low current applications.

Also I hate mm for PCB routing, like wtf is 0.7899999999 mm. mils are way more convenient since they are always nice round numbers. and common values are easy to remember and convert to mm (like 100mil etc).
However I really wish that the pitch was specified in mils, but it is traditionally in mm, and I memorized common pitches a while ago, like 0.5mm 0.65mm 1.27mm etc. and switching to mils for pitch sizes would confuse me. Damn, there is no good solution other than switching between mils and mm all the time.

>>2617141
> Also I hate mm for PCB routing, like wtf is 0.7899999999 mm.
And how do you get this?

>>2616979
pen and paper

I bought these "Rechargeable Alkaline Battery"s from Aliexpress, anyone know what charging parameters I should use to charge them? Maybe just charge them to 1.5V like it says on the label, or should I try 1.6V or 1.65V which is the approximate voltage of new alkaline batteries? I have a CC CV power supply, so was just gonna set it to 1.5V 100mA.

THanks

>>2617187
Maybe I'm wrong, but you should do two in series at 3V. NiMH can be charged at up to 1C, but keep it low (0.2C) for longer life.

>>2617187
>I have a CC CV power supply

have had very bad experience with charging these.
if you do it the CC CV way there's a good chance they'll start to leak very soon.
i also used a dedicated eneloop charger (which is all digital and shit) and they also ended up leaky.
just not as soon.
suggest you read more stuff.
i dont read, so i just chucked mine in a box, and throw them out 1-by-1 as they get leaky.
https://www.panasonicbatteryproducts.com/eneloop_rechargeable_batteries/eneloop-rechargeable-batteries-aa-4-pack-with-individual-charger/

>>2617206
Yes I've charged them a couple of times and they seem pretty bad, didn't hold much charge, and of the four I just tested the voltage on 2 were just over 1V, one was 150mV, and one was MINUS 36mV. So not a very good sign. I don't think I'll bother with these again.

>>2617143
one example if it was some kind of a default, like a clearance, set as a round number or a fraction in mm, 0.2mm and you convert to mils and get 7.87402mil

>>2617253
You should work in mm from the very beginning and have no fractions.

>>2617141
have you never heard of a "micron?"

>>2617298

Le Président de la France?

I have some fluorescent tube ballasts here, the top shows wiring for one 36W tube, the bottom has wiring diagrams for two 36W tubes or four 18W tubes. Anyone know if I could make the top ballast run two 18W tubes using the wiring shown in the bottom ballast?

>>2617338
No. that's Maricon

>>2617298
A ram manufacturer?

>>2617433
The difference is that the middle two filaments won't get hot. It may be that the bottom ballast gives a higher striking voltage, allowing it to trigger a tube with only one hot filament per tube, while the top one wouldn't.

Looking at the "U-OUT" values, the top outputs 280V while the bottom outputs 430V, which suggests to me that you need that kind of higher voltage to run two tubes in series even after they're triggered. So I'd suggest no.