/diy/ - Do It Yourself

>rules

fuck you

>>1759670
>reddit spacing
no u

>>1759651
What the fuck you fag? The last thread is not even half way full

>>1759896
some pigfucker posted a new thread without posting a link in the old thread
good night sweet bread

I'm making my first PCB and I have barely any idea what I'm doing.
I need a footprint for this part for Autodesk Eagle
https://www.digikey.com/product-detail/en/eaton-electronics-division/HCM0703-3R3-R/513-1704-1-ND/2295063

I can't figure out how to create the part in Eagle. All of the tutorials I have found are fucking terrible.

>>1761861
What have you tried? From your post it's evident you've done some work already. Do you have a point where you got stuck?

>>1761861
look in the datasheet for a recommended footprint, and look through the SMD inductor library to find something close enough until you find something or get bored

>>1762326
Terrible idea in practice.
>>1761861
Learning how to design your own footprints is a necessary skill if you plan on doing anything PCB related. Footprint libraries are shit. Also, if you happen to make any, feel free to upload them to online libraries :)

>>1761861
I usually use snapeda to find component footprints, but for something as simple as an inductor like that I just measure the dimensions with a caliper and create two pads with the Polygon button on the board editor (or Rectangle I can't remember) and then connect the traces by renaming the wires and the pads so that they connect with each other. If I'm feeling lazy I'll just put the pads and lay the traces over them, it'll show up as overlapping error, but that's fine.

I cant post pictures so let me explain this

i need a leads that can slip over a solid pins sticking out. pins small in diameter probably 12 gauge pins of so.
i need to be able to slip a lead over these pins and conect to my aligator clips for my meter

what would be the name of these female slip over leads and website to buy them ?

thank you
will post picture if it will let me later

my probes keep falling out from my multimeter because the plugs are shit so i want to snip them and solder them to bannana plugs instead (i just got my mind totally blown when i found out you can connect bannana plugs into the multimeter holes)
anyway is there something special in the probes like some super calibrated resistance or some shit like with oscope probes or is it safe to snip?

my motherboard is shit and doesnt allow chassis fan control
could i use a mosfet and a potentiometer to control the fan speed?
the fan draws 0.3A at 12V, but the fet shouldnt get too hot because when it drops a lot of voltage, the current should drop aswell, right?

TO220, IRLZ44N

>>1762575
lrn2Arduino, it's the easiest shit ever. The software even comes with the code, prewritten, for analogue voltage in -> PWM out, which is what you need to accurately control mosfets

>>1762575
>mosfet in linear mode
It will heat up a bit, senpai, but maybe?

I am a bit scared to press the ON button.
I think i read and understood the article on charging lead acids correctly, but if it turns out i'm retarded, i will end up with a face full of boiling battery acid on fire
But i mean, the values should be correct they are even on the battery

>>1762528
>anyway is there something special in the probes like some super calibrated resistance or some shit
Yes but not for what you are doing. I you don't know why they are special, then you don't need the calibration. I mean you literally didn't bother reading whats on the probes or your MM is so cheap it doesn't have the markings. cheap MMs are like 5 bucks. If you are fucking with expensive one, that's retarded.

>>1762596
>i will end up with a face full of boiling battery acid on fire
So do it outside, wear a face shield, and put it in a secondary spill container.
Drape the walls in plastic if you're stuck in an apartment

http://www.ti.com/lit/ds/symlink/ne5532.pdf

Is it even possible to use this one as a conventional non inverting amplifier?
I tried to get a gain of 11 but no matter what voltage I put on the input it always output 4.4V
I just bought whatever was available. I don't even know what to compare the datasheet with because I don't even know what a conventional OP amp is.

Also what are those resistors on the inputs and why do I have to put them there?

Hi there. I need some ideas, guys.

EE student, more into RF and CNC stuff.

(Sorry for autistic sketch, but it illustrates well).

I have to find a way how to get the speed of a vehicle without any intervention (or very minimal) or modification of it. I would use the information for driving the motor of trailer. Why, you may ask? This is assuming that car is not "torquey" at all ant electric motor helps to reduce towing forces.
I myself came to a conclusion that there's no good way to measure speed this way (laser Doppler method is hard to implement both by price and and processing).

Although it is possible to measure tensile force with a load cell at car-trailer junction and use it as guide for speed adjustment - motor would have to increase it's speed in order to make tensile force low. Although it may seem like a good choice, still, I don't have a clear picture how to implement it, nor I have any other viable solutions.

So, maybe you can give me some insights how should I do it?

Any help would be welcome!

>>1762629
Hi, what's the supply voltage and what's the amplitude of a signal

>>1762649
You need to sense the force on the trailer tongue. That is the only reasonable way to do it. How is that hard to implement? If the force is high, increase motor power. If the force is low, maintain motor power. If the force is negative, brake the motor.

>>1762655
I'm quite profane when it comes to motors, although it seems that I came to a good idea.

Just how exactly to place a load cell on a tongue? Plus, I assume it will measure tensile force, yeah?

>>1762670
I'll tell you, but only once you explain how the hell exactly the speed sensing idea is supposed to work.

>>1762653
supply voltage is 5V
the non inverting input is fed by a piezo that only supplies a few millivolts
The output would be 5V maximum, that I know.
I tried it with bigger signals too but the output wouldn't change from 4.4V

Sup, I would like to make a small timer thing. Here is what I want it to do in order of operations.
>1. Get signal from inductive input
>2. If inductive input senses absense of metal it starts a timer
>3. After timer hits 0 it sends an output signal to a solenoid

Is this feasible with a pi zero or a micro arduino?

>>1762678
oh whoops, forgot. Must be able to be powered by a small battery up to a AA

>>1762629
You set the gain with RF and RG resistors. Gain = Vin(RF/RG+1)

>>1762687
So like a normal non inverting op amp then?
What's Rin supposed to do then?

>>1762671
Assuming that motor has enough torque, using laser Doppler technique could be theoretically possible just to get speed information. Yeah, it is a naive approach, but..

>>1762689
Yes, it is. You can configure any way you want. Usually Rin is not needed and its usage is far fetched - from fault current protection for expensive sensing equipment (if for example power supply is shorted) to removing input bias (no matter how great input impedance of op-amp is, it still has some current running through it, so resistor in this place tries to mitigate this effect (so that there's no so-called zero drift)

>>1762678
Absolutely.
>>1762683
>Must be able to be powered by a small battery up to a AA
first result on google says you need a boost converter, and the arduino nano in sleep mode consumes something like 15mA.
Consider using a hardwired power switch.

>>1762697
oh well, sounds like what I may want is a mechanical timer...

I was going to make a simulation flashbang grenade

How do you go from ugly blocky pwm or inverter output to beautiful smooth sine wave?

>>1762710
Filtering, filtering and again - filtering. Plus, never expect a perfect sine wave from inverter - it's energetic properties should matter more than it's waveform

>>1762710
7 component Chebyshev filter.
How the fuck do digital signal generators even do it without being 80% toroid by weight?

>>1762715
High resolution DAC I guess plays a bigger role

Which one is better for use with microcontroller and no driver IC: common anode, or common cathode 7-segment 3 digits display?

>>1762678
You are better off with a rc timer made with electronic components man

>>1762710
MORE FILTERS!!!

>>1762693
thanks for the info
I'll try again tomorrow on a different breadboard and with another NE5532

Just finished soldering the display board to my new LCD screen to find out there is gunk stuck deep in the layers...
Wonder how difficult it would be to surgery it out of there

>>1762670
>>1762649
The speed of the car and the trailer are going to be exactly the same at all times.
You're an EE? Think of it this way: the link between the car and trailer is a very high impedance path. It conducts force with little or no movement.
So unless you lower the impedance of the trailer tongue (by allowing it to move in response to force) you will have no meaningful speed signal to use.

>>1762678
Maybe a metal detection circuit connected to an arduino?
What is the solenoid activating?
A little more detail would help.

>>1762649
can you not stick encoders on the trailer wheels? Or is that counted as part of the vehicle you're not supposed to modify? Because that would easily give you the velocity of the ensemble.
The load cell is a pretty good idea too, but I wonder about corner cases like slopes/parking

>>1762649
Instead of lasers, maybe capacitive or inductive sensors.

>>1762675
did you read the datasheet for the NE5532 to see if you can even run it from a single 5V supply
spoiler: no
fix: use a proper 5V RRIO opamp instead of high-voltage audio shit

>>1762683
what kind of solenoid do you honestly expect is going to run on a single battery? work out your I/Os first then worry about control

>>1762705
toys like this are good 555 projects but you're gonna need more voltage/current than that. only an idiot MBA fuck would make requirements as to the power supply without any clue what they're actually drawing from it

>>1762715
it's called a DAC, arduino faggot

>>1762727
common anode. sink drivers are no weaker than source drivers, oftentimes actually stronger

>>1762785
you're done with the protective layer. you can peel it off. the mattress police won't SWAT you in the middle of the night

>>1762792
>high impedance
>conducts current with little or no voltage
why aren't you killing yourself

>>1762649
why not just put a micro switch on the tongue and let it play a little bit over the ball, so that the motor is powered only whenr the ball is pulling on the front edge of the tongue?
>inb4 blue board
or, if the motor in the trailer is especially strong, use your strain gage output, suitably filtered and amplified, to PWM-drive a pass transistor for the motor, and tune the gain and offset empirically

>>1762819
>>high impedance
>>conducts current with little or no voltage
Current is movement, voltage is force.
There's force and nothing moves (car and trailer don't move relative to each other) - That's high impedance.
use your brain.

>>1762825
>There's force and nothing moves (car and trailer don't move relative to each other) - That's high impedance.
tighter coupling = lower resistance between the two nodes = lower impedance to changes on either side of the member being analyzed
>use your brain
stop trolling on 4chan

>>1761861
>not using based KiCAD

>>1762575
Get a 12V motor speed controller, PWM > linear shit with no feedback. If it's low power then a common-drain configuration might work though.

>>1762710
Is that a JFET? And why is there a resistance to ground from the transformer? Do you need a transformer in the first place? A 555 timer is a bad choice for PWM as not only does its duty-cycle change, but so does its frequency. The off-pulses are all the same length, so it's probably giving you distortion on the output. Make your oscillator frequency as fast as possible compared to your desired waveform (at least 10 times faster than it is atm), so you can cut more of it out with a filter. 555s are just shit in general, don't use them for anything. Probably couldn't even turn on most FETs since IRL the output only goes up to 3.5V or so.
If you know what you're doing, you can use negative feedback from the output back into the input in order to correct for aberrations in the signal, which is the principle behind a class-D amplifier. With a constant high-impedance load and good PWM control, that extra feedback shouldn't be necessary.

>$135
AHAHAHAHAHAHA

>>1762830
I even had to draw it out to make sure.
It might not seem like it at first think, but it's true. Consider it along with the rolling resistances of the car and trailer and you see. I even had to draw it out myself to make sure it was right.
see for yourself. If I'm wrong, prove it. What mechanically looks like a series circuit actually works out to a parallel circuit. The link is high impedance. Force with no movement. At steady state, the force across the trailer tongue is equal to the trailer's rolling resistance. when the car accelerates, that force is proportional to the trailer's mass.

My point in all this was, you can't get anywhere with that control system trying to measure speed. There's a reason why trailers with active braking do it with a pressure switch instead of measuring difference in speed in any way.

>>1762871
Not him, but it's confusing because forces add in parallel while movements add in series, which is the opposite of voltage and current. You could instead treat voltage as the movement analog and current as the force analog, which works better as far as Kirchhoff's laws, but it isn't quite as intuitive. I think the problem is you're both using different analogies.

Fuck it. Either way, he basically needs to put a material with some amount of give (springiness, elasticity, low tensile strength, whatever, i'm not a mecheng) between the trailer and the towbar, in order for it to move enough to be measured, since only displacement can be measured, not force. Even a load cell or piezo element requires deformation to give a reading. The standard load-cell devices are IIRC a load-cell bolted to a block of aluminium with cutouts inside the block to allow it to flex somewhat, but aluminium plastically deforms too easily and hence probably isn't the safest choice, and we can get away with more displacement.
I'm thinking ultrasonic distance measuring inside a cylinder that runs parallel to a (damped) spring of some sort. That or an optical or magnetic linear encoder. A rack and pinion with the encoder on the pinion might work well too, especially since the added gearing could provide extra resolution, but backlash in the gearing will cause hysteresis, which could cause oscillation. Mechanically speaking, the spring slider will need to only have give in one direction, forward/backward, and be perfectly rigid in other directions, which sounds like it's requiring linear bearings. Then again, a big leaf-spring held at either edge of the front of the trailer, with the tow-hook connected to the middle, probably wouldn't suffer from that issue, and would be rigid enough to stop sideways movement, only needing some simple guide brackets to ensure it doesn't travel up/down.

>>1762883
>a material with some amount of give
exactly. I've seen (in other applications) load sensors stuck directly onto steel. the steel itself provides the tiny amount of give to generate a (tiny) signal.

>>1759691
>reddit spacing

stop shitting up the thread two returns to separate greentext or paragraphs is ancient tech. Reddit spacing is every sentence or just about every sentence.

>>1762456
?
Try mill-max's pin receptacles maybe, they come in a variety of connection styles including solder cups.
https://www.mill-max.com/products?q=recs:0&r=recs:0;27986:14148,14153

Can someone confirm the value of this cap? Pretty sure it's .47 pF, but I'm getting conflicting google results.

>>1762924
I think it's more likely to be 0.47µF/470nF, because 0.47pF is the kind of capacitance you get from placing two wires too close together. It could be 0.47nF/470pF, but usually capacitor markings are in pF or µF, and I've never seen one marked in nF.

How big is it? Because you can make assumptions about its capacitance due to the dielectric it's using and that dielectric's theoretical energy density at max voltage. Good for an order of magnitude of precision at least.

>>1762930
Maybe a centimeter wide, it disintegrated, so I don't have it in front of me atm.

>>1762936
Size-wise, that sounds normal for a 470nF capacitor, perhaps a tad too large though. Have a look at some 470nF 100V film capacitors and check their dimensions, should be pretty similar.

What was the circuit context though? How did it fail like that?

>>1762938
It was inside a subwoofer. Not sure what caused it to fail, maybe a power spike.

Got my LCD setup!
Its my first time programming an arduino.
Now I just need to get a BT module that will take metadata from my phone and display it like pic related.
And then install it in my car.

Any suggestions for BT module? I hear I'll need one that supports AVRCP 1.3

>>1762939
Was it part of the crossover or some other filter? Could have shorted and absorbed a lot of energy from the amp. Class-X/X2/etc. caps are designed to fail more safely, perhaps replace it with one of those. They're usually rated for higher voltages too. If it was part of a crossover, you could measure the inductance of the coil in the filter and determine roughly what the capacitance should be by the intended corner frequency.

>>1762940
what if there isn't one? read the third section in >>1761645

Is this a good MOSFET driver? The comparator is open-collector. With up to 170nC of charge, and possibly as low a current pulled through the totem pole because of the 10kΩ resistor as 180mA, that's a ∆T of about 1µs. I'd prefer this to be lower, should I consider replacing the totem-pole's PNP with a darlington? Or can I slap a diode somewhere? Do I make another totem-pole to drive my totem-pole?

>>1762801
Yeah, it would, but velocity itself gives little information

>>1762962
it's aight. if you want the trannies to kick in quicker, add a diode between the bases of Q? and Q? to provide some voltage drop and speed the transition. you could take the 10k a lot further down. also put some resistance on Q?'s base so it doesn't saturate so hard
also consider Pic related
>putting a totem pole on your totem pole so you can drive while you drive
yo dawg

>>1762974
could you repeat that

>>1762978
diode between the bases of Q2 and Q3
also put some resistance on Q1's base

>>1762970
>I have to find a way how to get the speed of a vehicle

>>1762984
>resistance on Q1
Oh yeah, there's nothing limiting the current there, is there?

>>1762990
If he's using BLDCs he already knows the speed of the thing, because of the back-EMF or hall-effect feedback from the motors. But I don't think that data is actually helpful in providing feedback to the motor drive strength/current/duty-cycle.

Can you recommend some shop on ali that sells that transparent goo you melt all over a board with a heat gun before you work with smds
All i found are small syringes that have like half a shot glass worth of the stuff in them and cost like $20 a pop which is obviously bullshit since i watched some videos from luis rossman and he uses like a gallon of the stuff to solder even a single smd resistor so that would costs him like $10 worth of goo just for that one smd resistor

>>1762999
>But I don't think that data is actually helpful in providing feedback to the motor drive strength/current/duty-cycle.

Yeah, that's the crux of the issue. I tried to prod the guy into realizing but he was too dense. I mean you could have a naive open loop program that provides the right amount of force to overcome aero drag at a given speed, but that is stupid as fuck. What happens when you try and slow down or go over a hill?

>>1763032
Well even with a load cell he'll need an exception programmed in so it doesn't flounder about when he's parking. It should automatically not power the motor when the force on the load-cell is too small and when the speed is too small. Though maybe if he gets stuck in some mud the speed-based one could be toggled with a switch, or it could require a much larger amount of force or something.

I hope he doesn't plan on powering this off the car's battery, because that also doesn't make any sense. Considering the size of an average car, getting an electric motor or two to make an impact (say 50hp) will cost quite the fortune, so he's likely better off with some sort of petrol engine. Or you know, throwing a turbo on the car.

>>1763019
It's called no-clean flux, though there are likely names for particular formulations that you'd be better off searching for. Are you sure Rossman hasn't put a list of his soldering equipment in a video description or dedicated video? Because that sort of thing is often quite requested.

>>1763036
If the car is powerful enough to keep the speed up in a straight line, the motor could just be for acceleration. If it had regenerative braking it could work. Still seems like a lot of work for a marginal gain

I am buying a heat gun and chinks offer a version with what i assume is a replacement heating element.
Does that mean the heating spirals in heat guns wear out really fast?
I doubt it will be using it too heavily

>>1763037
Are there must clean fluxes? Because i don't see how you could clean them once you melt them and they flood the entire board

How should I go about designing an audio interface for a Raspberry Pi? Like, are low-latency adc chips all i2s? Can they be used interchangeably, with pi's OS figuring out how to listen from them?

I want to make something like pic related, but with a quarter inch jack for a guitar. Then, how could I go about making the Pi work as a DSP, then output the sound to a second chip to output the processed sound to a PC via USB? Would I be able to output the digital signal from the PI directly into a usb-audio chip, without converting it to analog, and then back to digital (which would be dumb)?

>>1763066
>Does that mean the heating spirals in heat guns wear out really fast?
It means it can be replaced and you'll know the exact kind to replace it with without asking on random forums. That is the same kind of hot air blower you'll find on stations such as 886D, meaning you'll also be able to buy the entire new thing if you want to. The blower inside it is a common 5015 blower, etc. The entire thing is very weak, you take a bit more than you should to remove IC's, but it shows the chinks think about repairability. Think it's going to be used not by someone who can get Hakko, but by a random Apu in India.

>>1763089
very weak but also very cheap, show me a better blower that actually works for $20 and i will be pretty impressed
I don't mind if it takes longer to melt those smds than some fancy $500 blower since i will be using it very rarely

>>1763093
>show me a better blower that actually works
Pic related heats the components a lot faster, it has a pump in the base, not a small blower

>for $20 and i will be pretty impressed
But it's more expensive

It's a step up from yours, BUT, yours will work

>>1763094
That is like high low end. Normally i always buy good quality middle end devices, but the problem is that i will be using this so rarely that it's not worth it to spend even the extra $40 if the $20 product will do the job.
Not to mention the cheap one takes up very little space which is great for something you almost never use

Whats the big deal with RISC-V?
I've seen a lot of fuss over chips coming out soon, but no one seems to have a solid reason other than its the newest thing. Is there something I'm missing? It uses a new instruction set, but that's all that I've really seen.

>>1762575
update: while fucking around with different fans i shorted something and shorted 12V through the fet
thank fuck i used thin wires so everything didnt weld itself together

the fan now runs 100% at 0 gate voltage and slows down when i increase it
did i just turn it into a depletion mode mosfet?

Do comparator inputs need some minimum current?
I tried using a piezo piezo plate that can output up to 5V or so but only a few micro amps and the comparator didnt even react to it
the comparator is an LM393

I can't find any shops that sell enhancement mode JFETs

>>1763165
>LM393
have you worked with the chip before? are you aware it needs an pull up resistor on the output?

>>1763180
>enhancement mode JFETs
hahahha, good one bud,

So I fucked up my headphones and one side is a lot quiter than other. Checked cable continuity and there's connection. Then measured resistance between mini jack rings and sleeve-tip is like 36 ohms and sleeve-ring almost 150 ohms. Does it mean I need replace the cable or one of the speakers is dying?

>>1763180
>enhancement mode JFETs
you mean UJT´s

>>1763186
could be both, better open em up and measure on the drivers themselves

>>1763159
I think you just need to shit in your hands and rub it on your chest, bro

>>1763183
>are you aware it needs an pull up resistor on the output?
well on the output I had an LED tied to ground and it lit up when I replaced the piezo with 5V from my power supply
on the inverting input I had 2.5V
this is confusing
I've never seen pull up resistors attached to an output before

>>1763197
ok I see it's an open collector
nevermind

>>1763183
thanks btw

Retard here.
Are there any current-production diodes that would be a suitable replacement for an AA143 germanium diode? They are evidently not produced anymore.
If nothing else I found some old stock on ebay but would prefer to avoid that if possible.

>>1763200
Just put some flux in there, bro

You are paid handsomely for doing your favorite electronics-related work
BUT!
you cannot use any kind of measuring device, and the datasheets you can look at are printed out, in a room full of binders.

>>1763201
me too, thanks

>>1763124
Open source ISA, performance similar to some ARM and less energy consumption

>>1763199
oh, no problem, a lot of my fellow students seem to forget that most comparators have an open collector output

>>1763200
for what specific purpouse???

>>1763205
yes, ofcourse

>>1763205
how did people even look for parts back in the day when you couldn't just google stuff and you had to print everything out or order specific catalogues?

>>1763200
>>1763219
disregard, solved

>>1763079
Rosin flux needs to be cleaned, especially if it has inorganic activators. You spray your board with isopropyl alcohol and scrub it off with a toothbrush. Usually rosin is used more for THT soldering, so there will only be a little on each pad, instead of everywhere and underneath your DIPs.

>>1763217
Not him but that one caught me up for a while a few years ago. It doesn't say it anywhere in the datasheet until you scroll down to the equivalent transistor diagram, and I was only formally taught about op-amps, not comparators.

>>1763220
i've no idea dude. I don't think I'd be exaggerating if I said it would've put me off electronics altogether.

>>1763220
you would order a catalog from one or more firms to look up the parts and design the schematic around the part which you ordered with a datasheet, or you could get parts through calling a distributor, who would actually have ´´sales engineers´´ helping you out.

>>1762794
Was going to be a imitation flashbang grenade, but a chemical fuse is the best way to go.

>>1763288
why the metal detector, to check if its still in the launch tube?

>>1763086
>Like, are low-latency adc chips all i2s
yes. there may be a little configuration you have to do to set word justification, but that's pretty much all there is to it as long as you have a hardware I2S port
>how could I go about making the Pi work as a DSP
look into the JACK audio subsystem
>to output the processed sound to a PC via USB
I thought you wanted low latency
>Would I be able to output the digital signal from the PI directly into a usb-audio chip
look up USB gadget drivers. if there is not already one for the usb-audio class, you may have to build one

>>1763124
ARM licenses are a bit on the restrictive side (can't add/modify instructions, etc.) and not free

>>1763220
National Semiconductor had a nice 3-part set of data books for their linear line (I think it was amps/regulators, audio/video, and special purpose). TI had a big fat databook for their 74 series logic, and I think they had one for their BJTs too. other manufacturers surely had the same
today we just pretend that LM317, LM324 and LM393 *are* the entire universe of linear parts

I have like 15 orders queued up in chink stores to resupply my lab
i hope coronachan won't kill them before they manage to ship them

>>1763373
soon we'll be ordering our electronics from india instead
or we would if they weren't catching the NCV as well
goodbye, civilisation

>>1763385
if both insectoids and poos go extinct it will mark an end to my electronics hobby because there is no way i'm buying insanely marked up components from local retails jews that literally cost 10 times more

>>1763386
but all the consumer electronics will also be marked up a bunch, necessitating that you diy your own stereo and shit all the more

>>1763387
That could be pretty cool because if prices of consumer electronics skyrocket where something like a digital thermostat now costs $500 instead of $50, it will be possible again to make decent income by dying simple electronics like that and selling to people in the area

>>1763388
It probably already is possible to make money off that stuff if you style them to hit the right market, while making them as non-proprietary/as universal as possible. Like TV remotes made with that secret wood + resin mix that work with all television types. Or thermostats that work with google home and the xiaomi one and the apple one and alexa etc, also with some sort of hipster aesthetic.

>>1763387
hi-fi > wi-fi

>>1763386
but aren't you glad you bought the soldering gear while it was cheap?

>>1763186
>>1763190
This is how it looks like inside. Multimeter shows the connection between cables of the same color and mini jack, so I suppose it's a driver failure.

>>1763396
>hi-fi > wi-fi
Not as far as the average consumer is concerned. Though on the other hand, you could make chromecast clones that use 8P8C instead of wifi in order to appeal to the "muh death radiation" crowds. Assuming they don't already come with that option, I don't have one.

Can you run smartphones out of dc current from some type of adapter rather than from battery? I want to connect both power and old disk to old phone to make shitty seedbox but i have only 1 usb to work with

What soldering station do I get? I have a really shitty chinese piece of shit and I can't do shit with it.
Looking for something capable of doing smd preferably with a hot air bit as well.

>>1763415
For a standalone soldering station, I can recommend the T12 OLED soldering station from Quicko or KSGER. There's also a hot-air setup by the same guys, but I'm not sure if I'd trust it as I don't have any hot-air myself. See this comment chain:
>>1763066
>>1763089
>>1763093
>>1763094
With any luck, you'll find power and air-flow ratings with which to select the right model for you.

>>1763415
>>1763419
>For a standalone soldering station, I can recommend the T12 OLED soldering station from Quicko or KSGE
NO!

Get TS100 that is the ONLY acceptable choice for any electronics work.
I have been using it to years and just the thought of using anything else to solder makes physically ill.
As for hot air get the cheapest piece of shit heat gun possible, all it has to do is blow hot air, only a retard would spend $1k on what is basically a shitter hair drier.
The only exception is if you plan to work on electronics that feature big heat sinks like pc motherboards (huge copper planes) then you will need more powah to heat that shit up

>>1763424
>>1763419
i remember why i gave up on this shit
no one sells this shit here
all i can find is chinese copies sold by incompetent pieces of shit for $100
i hate this fucking country

>>1763425
ts100 costs $50 and chinks sell them as well, legit ones
in fact that is where i purchased mine
make sure to buy from an ali store with at least 100 reviews

>>1763427
Wait I found 1 (one) person selling it

>>1763439
https://www.aliexpress.com/item/32860309312.html?

bullshit, ali is full of it, stop being slow in the head

>>1763443
Sorry I meant 1 person in my country

>>1763403
it worked on some old piece of shit i used(made around 2011), the battery percentage was just a grey "?"
newer phones might not turn on without the shit that comes with the battery

i do wonder what would happen if you charged the phone while running it off DC, sink current into the DC source?

fuck
>power goes out for 3 hours
>power comes back on but breakers trip, probably incompetent cunts at the ele company sent a nasty pulse
>computer no longer turns on
>other computers do though, so the surge couldn't be THAT bad
do computer power sources have like poly fuses that can take several hours to reset? do you think it i could just be a blown fuse or something?
i hope to god it didn't reach the 6 expensive harddrives that are inside

>>1763506
Your PSU is fried because you were to cheap to install an UPS and to stupid to unplug important electronics during a power outage.

>>1763507
fuck the psu, i don't give a shit about it, but do psus have some safety shit that prevents the AC electron bully from reaching the very sensitive DC insides of the computer like the motherboard and shit?
It is my NAS and i have like $500 of harddrive on it and tons of important data.
It's raid but that is useless if all hdds are fried. The important data is backed up of course but i would still lose at minimum two days of work.

What are those buttons on the bottom right called and how do you go about making them functional in a box for example? The other components are so tall too in comparison.

Trying to control a 6V lamp with a small relay & two switches. Would this work? Is there any way to to use normally off switch for the OFF button.

>>1763511
tactile switches, and some of them get pretty long (13mm)
Any more than that, you'd use a panel mounted switch and just run wires to the board

Electronics noob here, cross posting from /mcg/ hoping someone can help me out with an issue.

I've got a PCB I had fabricated based off a kicad schematic, excerpt pic related. There are 4 CD4021BE ICs, each with switches connected to the parallel input pins, forming a keyboad, as well as some additional switches.

SW102 doesn't register presses at the arduino for some reason, but every other switch on the board does, and I'm not sure what the issue is. On the board, multimeter tells me that switch operates fine and there's no shorts or bad solder connections.

Any ideas what I've done wrong? I'm at wits end, but also I expect the problem to be stupidly simple.

>>1763570
Footprint. I've removed the copper pours for gnd and +5v from kicad for that screenshot for easier viewing, but they exist on the fabricated board. You can see the trace from SW102 to R194 to pin 1 on P104, a CD4012BE IC.

Did anyone else ever have a li-po leak like an alkaline?

Opened up some old cheapo bluetooth receiver and noticed that the small li-po in it had some disgusting shit on it.
Lifted it up and saw that the stuff ate away the solder mask of the pcb under it.

Kinda impressive, never saw a chemical harming the solder mask.

>>1763403
smartphone power management ICs do system load sharing (aka power path management). if you have a beefy enough adapter and the phone knows how to enable its high power mode, you'll be fine. some PMICs do require that a battery or reasonable facsimile is connected during some parts of boot
but consider a $10 chinkboard SBC instead, which will have multiple USB ports, a distinct power inlet, and no need to fuck around with a battery safety controller which already hates you

>>1763415
plenty of 2-in-1 pencil+hot air stations on ali for about 50-100 burgerbux. fine for getting started. buy a set of "screwdriver" (aka double bevel) tips of various sizes, tip tinning compound, and the brass-wool tip cleaner. TS100 is a meme

>>1763500
yes, it would try to deliver current into whatever you have connected to the battery terminals

>>1763529
yes, add another relay with a NC contact as an inverter

>>1763424
>Get TS100 that is the ONLY acceptable choice for any electronics work
this is a shitpost, right?

>>1763642
obviously

I bought what looks like a counterfeit lt3045-1 which is a low noise regulator.
So now it looks like I need to spend $40 to make sure I get a real one or build my own board. Thing is I dont know much about board design for noise. Where does everyone else get there ultra low noise regulators from.

>>1763403

Even if the phone didn't like the lack of battery, you could easily fake it by attaching an extra regulator to your power supply and feeding it to the +/- terminals on the phone. A resistor might also be required to fake the NTC thermistor that normally monitor's the battery's temperature, but, again, simple stuff.

Source: Recently made an adapter to make my camera run off a power tool battery for extended recording times. Basically the same thing you want to do, worked out fine in my case. You may not even need the regulator, as the 5V available from the average USB adapter is close enough to the-fresh-off-the-charger voltage of a li-ion battery that most equipment won't have a problem with it. The camera didn't, at least.

Here’s a big dummy question.
I have a power brick that puts out 90W. The original wire that I’m replacing is rated for 7A 120V. The new wire is 2.5A 250V. Am I gonna see this wire light on fire if I plug this in? It’s for a laptop charger btw

>>1763713
Digikey, Mouser, Arrow, Farnell, etc. are more than likely able to give you legitimate parts. You could also try applying for free samples from the manufacturer itself, and if you're ordering a large enough quantity you could order through the manufacturer normally. Like with those links at the bottom of TI datasheets.

Or instead is the question "what parts to do you use when you need a low-noise regulator"? Because if that's the question, I can only hope that Mouser or one of them allows you to sort for parts by PSRR or some other noise stat.

It fuckin works friends!
https://streamable.com/zsz0c
Bluetooth added to a stereo from 1995 and it sounds better than my fancy kenwood deck I paid $130 for
Thanks for the help /ohm/!
Now to get working on the media keys and metadata display...

>>1763740
Ive read the datasheets for multiple ldos. But ive also read a lot about how board layout is responsible for like 75% of the noise. The parts arent what Im concerned about.

>>1763730
>Am I gonna see this wire light on fire if I plug this in
Well 90W at 19V means 4.7A. Putting 4.7A through a 2.5A cable won't end well, but probably not to the point of catching fire, just getting hot enough to soften the PVC insulation. If you're not careful, this could cause the conductors to push through the PVC and contact one another, causing a catastrophic short-circuit or at least blowing a fuse in the PSU. Even if the insulation is rated to 300C or higher, I'd still avoid it and go for a better cable.

Unless you're putting the cable on the AC side, in which case it's perfectly fine.

>>1763748
So by "Where does everyone else get there ultra low noise regulators from" you somehow meant "how does everyone design low-noise PCBs"?
Bypass caps everywhere, beware of inductance of traces (put current-carrying traces close to each other so their inductance cancels out), avoid microphonic capacitors and non-toroidal inductors. Make jumper wires twisted pairs or coax, and if you're dealing with higher frequencies, look into microwave board design techniques and transmission lines. Put ground fill everywhere. If you're really bothered, a grounded ferromagnetic shield soldered over your board would do good too.

>>1763762
>Unless you're putting the cable on the AC side, in which case it's perfectly fine.
It’s the wall plug, yes. Burgerland voltage.

>>1763766
I must add: But why is having that wire on the AC side a non issue?

>>1763767

Because power (Watts) is the product of voltage multiplied by current.

Assuming the charger is 100% efficient (it's not, but whatever), this means that, at 19V, you have 90W / 19V = 4.74A of current on the low-voltage, DC side. However, on the AC side, you have 90W/120V = 0.75A of current, well within the 2.5A rating of the new cord.

Even if we assume that the power supply actually only ends up being only 50% efficient, that brings up the required current to 1.5A on the input side (ignoring power factor, which is beyond the scope of this tutorial), which is still acceptable.

>>1763770
At last I truly see. Thanks.
I found my reading glasses and managed to make sense of the pea sized print on the power brick. It says:
>INPUT: 100-240V~
>2.5A-0.5A 50/60Hz
>OUTPUT: 20V (line and dotted lines below) 4.5A
What does that -0.5A mean?

One of the pins on my micro suddenly developed a +300mV offset. Did I fry it? I've been messing with PWM that I fed into RC and LC circuits. I think it happened after I messed with inductors. I am always afraid of them. All other pins are just fine. And this one measures +3.3V high and +0.3V low.

Nah I don't think the inductor could do anything, just ran a simulation and thats the waveforms I've also seen on the scope. THe voltage drops on R and L cancel each other out and I don't think any negative voltage could appear on the pin?

>>1763779
>Did I fry it?
Maybe. Were you using diodes for safety?

>>1763783
Look at your MCU's datasheet and look for an equivalent output and equivalent input diagram, because it doesn't just act like a voltage source.
>negative 4.5V
>I don't think the inductor could do anything
Look for the absolute maximum input voltage range. Chances are, it's Vcc+0.3V to -0.3V or so, hence -4.5V could definitely fry it.
What MCU are you using? Why would you use it instead of a function generator?

>>1763776
>What does that -0.5A mean?

you know how a regulator can be labeled 5V +/-5%?
your shitbrick is rated 2.5A +0%/-20%

PSU noise anon again. Both C4 AND C7 had dry and broken solder joints that I reflowed. I then threw a couple of 10nF caps on the bases of Q1 and Q5, which dampened down the higher frequency stuff a bunch. No clue why they have both R2 and R3 with a cap between them, it looks like something intended to dampen noise but it obviously doesn't work. Turns out some extra noise I was seeing was from the built-in 100kHz 555 oscillator, which I turned to the 1Hz setting so now that's not confusing me. The trick to figuring out the noise source was throwing an alligator clip to ground with a ceramic cap on the end that I'd just probe about, looking to see what improves my noise.

Now my only real problem is that I've got varying levels of 100Hz mains noise permeating all 3 of my rails. My scope won't trigger on it properly because it's really fuzzy, but at least I have a line triggering function. I can dampen this 100Hz by slapping a larger electrolytic cap along with the new ceramics, but I'm starting to suspect whether C4 and C7 are the source of the noise in the first place, removing or replacing them might fix my problem.

Also now that I'm probing the voltage to ground on either end of my new ceramic I'm seeing the 100Hz noise on both ends, which suggests that the caps aren't tied to ground but to some other rail. But turning it off and measuring continuity they are connected. What the fuck.

I have an oxygen sensor which is basically a current source providing 80-120 uA at a normal O2 concentration. The recommended load resistor is 47-100 Ohms, which provides 5-10 mV. Now I have different ways to read it out. I want to be able to see when I have 0% O2, i. e. 0V.
What would you choose?
- take a random 5 cents op-amp, amplify by ~300 and read it out with the built-in DAQ in an arduino
- use an instrumentation amplifier like the INA188 and digitise the signal with an arduino
- use something dedicated device like an INA226, that's read our via i2c
- a better solution one of you provides

For 1 & 2 I probably need -5 V if I want to be able to see an input of 0. What's a space-saving, elegant way to generate that?

>>1763642
watch a review on it from anyone that matters and you will find out how stupid you sound just now

>>1763813
Since you need a DC output, I'd lean towards 2 or 3. The input bias required by normal op-amps could skew your result by a percent or two. But with the right compensation, a cheap op-amp might do the job. I'm no expert on low-noise circuits, but I can only imagine that a purpose-built ADC will have better linearity than the arduino's ADC, and better small-signal and DC precision than even an inst-amp circuit.

>>1763812
>accidentally short heatsink to ground
>alerted by crackling noises
>heatsink was at 18V and directly connected to the rectifier
oh fuck oh fuck
>still works
Nice. I just removed C4, and the circuit behaves identically to how it was before. Still haven't worked out the strange ground issue. I don't think it's inductive coupling of the circuit to the transformer, and there's no resistance along that trace, so I'm stumped.

>>1763506
This is so weird.
So, the PSU is working just fine i measured all the voltage outputs and they are there and the fan spins as well.
Is it possible for a power surge to go through a switching power supply, fry the motherboard and do all that without damaging the PSU at all?

>>1763849
I watched Dave Jones review it and compare it to the TS80 and decide that the TS80 was slightly better. But why should I take what >influencers say over my own opinion constructed from multiple testimonies? The TS100 isn't particularly bad, but it's not stellar either. It suffers from a few issues, notably:
>no front finger guard
Everyone 3D prints their own because the retard designers of the TS100 valued aesthetics over functionality. When you've been soldering for a while and have sweaty hands, a solid plastic piece without a ledge isn't the best situation to find your fingers on, compared to the rubber grip on my T12 and its solid 2mm ledge. Sure this can be mostly fixed, if you have a 3D printer or other fabrication setup, but that's just an if. Still no rubber grip, though for some people that could be a downside.
>doesn't come with its own power supply
The "omg only $50" selling point doesn't count for much when it still requires a $30 power brick. Sure if you have a laptop power brick lying about then you can use it without buying a new one, but like the previous point, not everybody does. And even if you do have a laptop power brick, those things only put out 19V, not 24V. Meaning you're only getting 63% of the maximum power output, which is a pretty big deal if you're going at a ground-plane. Since barely anyone will just have a 24V adapter lying about, the $50 pricetag is pretty disingenuous.
>bulky grip
The whole point of the TS100 design is to be portable, I get that. If I were buying a new iron for soldering on the go, I'd definitely want a TS100 or TS80, the form-factor is great for that. But I never do, because I won't have my helping hands, my vice, my magnifier, my tub of flux, my oscilloscope and function generator, and my boxes full of parts. 95% of the people using this iron will be using it stationary. The result is a handle that's bigger than it needs to be, with uncomfortable screen and buttons on it.

The T12 fixes all these.

>>1763860
>>no front finger guard
Allows you to 3d print a custom on that is any shape you want, or leave it in this sleek super portable shape so you can carry it in your pocket in case you need to solder on the road and also
>i don't have a 3d printer
HAHAHAHAHA pleb
>$30 power brick
lel WHAT?!
I use mine with a high quality 24V 6A power brick from chinks that costs $6

Also because the TS100 has plug and play soldering tips i can easily switch them with my bare hand WHILE SOLDERING without even having to turn the power off and the newly inserted tip is at 300°C and ready to solder in about 5 seconds

Not to mention it works perfectly with a 12V car battery or lipo pack which will come incredibly handy when the society collapses in the coming weeks due to coronachan and you will need to do survival soldering on the road. Good luck drugging around you table sized hakko station and finding a working power outlet for it. But you clearly don't prep so you will be long dead when that happens, so that is not a selling point i guess as the dead have no need to solder anything.

>Also because the TS100 has plug and play soldering tips i can easily switch them with my bare hand WHILE SOLDERING without even having to turn the power off and the newly inserted tip is at 300°C and ready to solder in about 5 seconds

You can do the same with a T12

>>1763860
In addition, if I wanted to make my T12 setup portable, I'd just have to drill a hole in the back and install a DC barrel jack and wire it up to the output of the power supply (after the diode that's on the flyback transformer but before the ∏ filter). Then plug in one of those DC plugs you get on caravans, or wire it up to a car battery if you don't mind the 1/4 power.

Now the T12 isn't perfect. While it does have a pretty good interface (can see set-temp, current temp, duty-cycle, battery/psu voltage, time, and ambient temperature all at once) that lets you set the standby and boost temperatures, sleep and standby time, the size of temperature steps, low-voltage protection thresholds, and a bunch of other features, it isn't particularly easy to program. You have to unscrew the front and plug into the ICSP header yourself, no easy USB stuff. But I'd bet that 99% of people using the TS80 or TS100 will never update their firmware at all, and I've never seen any reason to do so myself.
The soldering handle is also relatively long, and the tips don't stick out as much as they could, but both apply to the TS100 too.

>>1763862
T12 has a finger guard and it isn't exactly bulky. Do you know what 2mm means?
>HAHAHAHAHA pleb
So now i have to buy a TS100, a PSU, AND a 3D printer to solder? Sure showed me pal.
>$6
You mean those chinky deathtraps that come without an enclosure? How portable.
>plug and play soldering tips
Doesn't it have a little set-screw you need to loosen? You're forgetting that the T12 has the exact same tips and hence you can do the same thing with it. And no set-screw.
>bulky hakko
The T12 is the opposite of that, it's a small aluminium box with a SMPS inside it. An aluminium box that's bound to be smaller than your chinky deathtrap.
>ignores the "If I were buying a new iron for soldering on the go, I'd definitely want a TS100 or TS80" part
Have you ever been told that you're not a very fun person?
>society collapses
>prep
HAHAHAHAHA pleb

>>1763866
>You mean those chinky deathtraps that come without an enclosure? How portable.
I believe he's talking about the exact same PSU that's used in the T12 stations.
>The T12 is the opposite of that, it's a small aluminium box with a SMPS inside it. An aluminium box that's bound to be smaller than your chinky deathtrap.
To be fair, you have a dodgy chinky power supply in the T12 and the small aluminium box isn't even grounded. They use taped plastic standoffs to fix it on inside the case. There isn't even a screw hole too fix a ground terminal on.

If you don't want a Chinese death trap, you should go with an original Hakko.

>>1763868
If you already own t12 then there is no point to buy ts100, but if you are that guy who is looking to buy his first iron then ts100 is the best choice (ts80 is slightly better hardware vise, but also costs like a third more so it's not worth it yet)
As for the power source you already have it at home, it's that old brick from that notebook you don't use. Or literally go to like a junkyard or a thrift store and get one for $1 or even free)
I hope he won't let the t12 shill trick him, but oh well, can't rescue everyone i guess.

>>1763868
Well to be fair, the aluminium case is anodised and hence not conductive on the surface. Pushing my DMM leads into it as hard as I can doesn't give a reading lower than 60M, so I doubt any loose wire that touches the outside will cause it to go live. Not as good as plastic, but still not as bad as an american ungrounded toaster. When I put a DC jack in mine I'd consider throwing a ground on it too, if only I wouldn't need to put 4 bolts through it in order to ground all 4 pieces of the enclosure.
As for the power supply itself, the one inside the T12 goes for about $12 not $6, and as I said it isn't enclosureless, it goes inside the enclosure that comes with the T12.
>original hakko
But don't they all use transformers instead of SMPSs? Makes them more bulky and what. I'm not opposed to buying a higher quality station from a proper brand, and I likely will do so eventually, but compared to these new open-source designs they look a bit behind the times. I saw that video where Dave accidentally plugged a 120V Weller into 240V and fried the windings, and looked inside to find no primary-side overcurrent or overtemp protection. He then checked a bunch of other soldering stations to find the same thing in a decent fraction of them.

>>1763871
>ts100 is the best choice
Have any reason behind that excluding ad-hominem? Anything I haven't already addressed?
As for notebook power bricks, if you'd read my point above, the 19V laptop brick only gives your iron 63% of the maximum power output it would have from the proper 24V power supply (19^2/24^2 = 0.63). For ground-planes, this is a big deal and means turning your temperature up even higher, meaning you're more likely to burn parts and have less time to work with them.
>shill
If you'd been reading, the TS100 guy has been using less sound arguments and more emotive rhetoric, making him a higher class shill than I.

Is there a name for a relay that can switch both + and - wires at the same time and also has larger gap between contacts to protect against ´high voltage surges?
so when the relay is off the device is completely air gapped

>>1763876
Relays use the same terminology as normal switches, so you want double pole (DPST or DPDT)

https://www.aliexpress.com/item/4000495151872.html
d-does it have actual temperature feedback? or just open-loop duty-cycle? if it does have temp feedback it's pretty good value for money, since it fits basically all the talking points of this guy >>1763862 but for $16 instead of $50

>>1763876
how high voltage? because if it's higher than a 240vac relay or a hvdc relay i do not know if you will be able to find anything. as for the + and - thing, you want a dpst or dpdt relay

>>1763879
>how high voltage
lightning strike
so only very short voltage increase, not a consistent high voltage

>>1763881
So you want surge protection? Just add some MOVs.

To survive direct lightning strikes you'd need something ridiculous beefy though.

>>1763883
if the voltage won't kill your shit then the emp certainly will. just ask the squids from matrix

>>1763881
>>1763883
why not just have a normal relay and a spark-gap to ground that's shorter than the relay seperation?

>>1763887
>spark-gap to ground that's shorter than the relay seperation
oh shit, that's actually a really smart idea, you should have that shit patented, i will definitely do that

>>1763879
>Temperature Stability: +/- 2%
Regulated, T12, real knob, no digital bs.

>30dB ripple rejection through linear regulator
Ah, that's awful! That UA7805 should be rated to a minimum of 60dB, maybe I'll have to replace it.

>>1763895
I'd like to see a teardown of that, and see whether there's an MCU in there, some sort of dedicated thermocouple pwm (PID?) IC, or just an analog control loop. If dedicated PID ICs do exist, I'd buy a dozen right now. If it's a control loop, I'd like to see it anyhow. It's probably a micro isn't it?

number 2 mains shock
shouldn't be doing electronics at 1am

>>1763896
>I'd like to see a teardown of that, and see whether there's an MCU in there, some sort of dedicated thermocouple pwm (PID?) IC, or just an analog control loop. If dedicated PID ICs do exist, I'd buy a dozen right now. If it's a control loop, I'd like to see it anyhow. It's probably a micro isn't it?
Just an analog control loop:
>https://hackaday.com/2020/01/27/review-saneryigo-sh72-soldering-iron/
>The two main semiconductors are a TPC8107 MOSFET and an LM2904 dual op-amp, both in SOIC-8 packaging.

>>1763899
Oh man, that's pretty neat. But "using the element resistance as a temperature sensor" sounds a little suspect, since that would require measuring only when the element is on. Plus there's a thermocouple that could be measured with one of those dedicated MAX-etc ICs for more effectively feeding a single-comparator PWM circuit. I simply can't see how you'd get amplification AND an oscillator AND the PWM from a single dual op-amp. I suspect instead it's just the oscillator and PWM and the pot is just for duty-cycle selection. But he says the temperature scale was somewhat accurate, so maybe it does have a feedback loop.

In any case, someone needs to do an actual teardown, with a circuit diagram reverse engineer.

I have recently become the owner of a 1500W, 230->120 V autotransformer: https://www.coleparmer.com/i/todd-systems-sd-14-g-inc-step-down-transformer-1500-watts/0157809

it's too bad the output is fixed, but what are some projects ideas?

>>1763904
use it to make a really powerful boost converter

>>1763879
https://www.aliexpress.com/item/4000319749328.html
here is another model again, with extra stick out

>>1763905
interesting. Could I drive a DC motor with that?

>>1763913
depends on the motor? would need to test if the mutual inductance or coupling efficiency changes as a function of frequency though

Hi /diy/
I am making for myself analog video comb filter. As a base I want to use SAA4961 Philips chip. I already know how to connect every pin, except pin FSC, which is basically subcarrier frequency input (3.57 MHz). Now I don't how the hell I should connect it. I thought about Pierce oscillator with correct parameters and a 3.57 cristal, or even making a bandpass filter for the input video signal to extract the goddamn subcarrier frequency. Any ideas? I've never dealt with video stuff.
Datasheet: https://pe2bz.philpem.me.uk/pdf%20on%20typenumber/S/SAA4961_1.pdf

>>1763786
>Were you using diodes for safety?
no

> Chances are, it's Vcc+0.3V to -0.3V
STM32, so yes of course it wouldn't tolerate -4.5V but I don' see how it would result in -4.5 on the pin. on this simulation at the falling edge: >>1763783 -4.5V is across L to ground. at the same time V across R is 4.5V and the pin is 0V. So how would the pin "see" that -4.5V?

>>1763946
I think I know where I am wrong, since I am modeling a voltage source that is alternating between 5V and GND, that is not how the real GPIO port works.

What's a good cheap way to measure current with higher frequencies?
I got a 40kHz AC current and none of multimeters can keep up.

So, should some thing like this fix the problem? By placing the diode before the resistor instead of across the L so I still could see full charge/discharge cycle without damaging the port?

This is retarded.
Whose brilliant idea was it to combine square waves with inductors?

>>1763979
man i am so confused. the GPIO already has internal protection diodes, wtf am i missing? why can't GPIO tolerate negative voltage then???

>>1764009
ah maybe it may exceed the current, it may be a puny little diode or maybe some kind of a fet that can only tolerate -0.3V.

What the fuck

>>1763859
Yes, it's very possible.
I've dealt with your situation a few times, check your RAM. More times than not, I've found a back stick of RAM after a power surge.

I was given an alico dj-v27 ham radio by a coworker. I need to get the battery working. I see a few online but was wondering if there is any way to just repair the old one?

Battery is pic related, looks glued together.

>>1764009
>>1764012
Yes, you are most likely trying to pull way too much current. It would be better to use some buffer to drive the inductor. This is a prime example of why motor drivers exist (H-bridge for example).

>>1763813
mostly depends on what I have on hand, but I generally favor the integrated solution I can trim digitally or don't need to trim at all, so INA226 probably. you might also look at the cheap, slow but high-res digital weigh scale chips like the HX711. for the inst amps you should only need some stable voltage source for the 'ground' side of the sensor, which some ADCs might provide for you, and not need to worry about a -5 rail
if you're on the cheap you could also look at some of the decent low-voltage precision opamps like the TP156xA from chinkland

>>1763859
anything's possible

>>1763904
>I have junk, what do?
sell it to someone with a use for it

>>1763937
yes, feed the output of a Pierce or other oscillator in there. check the datasheet to see if it wants a sine or if it's okay with a square wave

>>1764012
yes, it's a puny little diode intended for minor ESD events, not absorbing inductive spikes. always bring your own protection

>>1764030
holy /csg/ batman

>>1764038
you can replace the LiPo pouch inside (if you can locate a suitable replacement), but it probably isn't worth the time unless you're doing some sort of period piece restoration

>>1764043
>junk
who said it was junk, apparently it's worth 100 dollarydoos

>>1763977
So I ended up using a toroid core for a makeshift current probe and hooked it up to a scope.

Not very accurate but the readings seem about right, plus I get a nice visualisation when ever the resonant frequency is just right (nice sine wave - garbage otherwise)

>>1764067
this is engineering, not found object art class
sell it to someone with a use for it because clearly you have none and lack the ability to come up with any

>>1764099
yep, pretty much just add a probe of your choice, optionally filter, and read the voltage with an instrument of your choice. Allegro Microsystems makes some nice little Hall-effect current sensors that are good to have on hand. a resistor with a floating amplifier might have worked too

>>1764118
lmao what are you even saying dude
>this is engineering
you're posting on a hobby board of an anime website. Stop larping

Is this beautiful or what?

>>1764152
>component name obscured by wires
0/10 would not place and route

>>1764154
Shit you're right.

>>1764128
On this board we get too many retards saying "I bought X, what should i do with it?" where X is a raspberry pi, arduino, vacuum tube, motor, all sorts of stuff. And they never do anything with it, at least not that's ever seen here afterwards, it's just a waste of everyone's time. If you don't know what to do with some thing you bought, then how the hell are we to know what would be useful to you? Our default suggestion is "sell it", hence that bit at the bottom of the OP. I think that guy's response was perfectly valid

>>1764152
>bunch of high level digital ICs plugged into one another with barely any passives
Truly a circuit of the 21st century.

Why is everything simple so complicated??
All I want is to measure the time constant for an LC circuit. Without killing my GPIO ports. I tried adding an op amp like in the pic. But it distorts both output AND input waveforms, so measuring anything is pointless. It makes the input PWM look like a trapezoid. And I am talking periods like 40us, nothing too fast. What the hell? Do I need a special buffer amp? This is LM358. It is rail to rail and not particularly slow.

>>1764177
Why not use a MOSFET or totem-pole? An op-amp really isn't necessary, not to mention not terribly good at rail-to-rail work. Also I'd instead throw a cap in parallel/series with the inductor and measure the frequency of the ringing, as opposed to measuring the RL time constant. Or make an RL/LC comparator oscillator and measure it's frequency (replacing the RC low-pass filter with an RL or LC filter. The act of putting a diode there will dampen one half of the waveform, hence distorting your measurement.

Or check out how that chinky $3 transistor tester does it, I think that's an open-source design.

Is there anything wrong with using a TPS61022 boost converter to get 5V out and then use a LM3940 LDO regulator to get a 3.3V line? Could the frequencies interfere with each other?

>>1764187
yeah, regarding the mosfet i also simulated something like this. it looks better but there are some other issues there, since now that inductor is in parallel with the load resistor and <something happens but i forgot but it only works well for very high inductances>. as for the resonant circuit, yeah that is a totally different approach, I did something like this, but I am not building a meter. I specifically need to be able to observe the charge/discharge waveforms for 1st order diff eq for RC and RL, and be able to measure the time constants looking at the waveform (63% of Vcc). Seems easy for capacitors but not for inductors. Inductors are just nasty creatures. To be able to measure R/L in the range of single mH and below requires high frequencies or R would have too be too low so it would draw too much current.

>>1764190
LDOs don't have frequencies. if you use sufficient decoupling and bulk capacitance, you should be in good enough shape and very unlikely to experience any catastrophic conditions. in the event you need extra filtering because of some sensitive analog circuitry, feel free to add an LC filter to quiet things down

>>1764197
Thanks anon. I don't think I should need any filtering as everything is digital.

>>1764192
There a reason to use an NPN to switch the FET? The built-in output drivers shouldn't be awful, better than that pullup resistor in any case.

What's wrong with high frequencies? 100kHz shouldn't be an issue, no? But ensure you use doides with low capacitance like the 1N4148. Use a fast enough comparator and a time-digital converter and you could go well into the MHz range. Even without a TDC, just using the MCU for timing and using the comparator and some precision resistors to test when it's over the threshold would be my method. With an N-FET, because I don't see a reason to use high-side switching. You also avoid the SAR's delay doing it this way.

Something like pic related

>>1764177
don't forget the 1-100Ω in the GPIOs
>LM358
not actually rail to rail on the output
I suppose that's one more reason LC meters have converged on the AADE LM311-based LC resonator design

>>1764152
6.5/10 weird SWD connector but ok. maybe share the display (not shown) with the nRF chip instead of the SD card. I'd think about putting a couple of 10MΩ resistors in a voltage divider on PA0. otherwise your TRNG might be influenced by local periodic emissions (like the nRF module crystal) more than that sweet, sweet thermal noise

you know what would be handy? sr latches with schmitt inputs that have precise trigger voltages. for timing purposes, and so you could fit 3 of them on a dip14, or even 2 of them on a dip8 or 4 of them on a dip14 if you ditched the second output. so if i wanted a pulse that turned on 1s after pushing a button and lasted 20ms, i could just use a single dip8 ic plus some resistors and capacitors
555 timers arent all that great

>>1764205
Hey I see you have the same problem with unequal font sizes? I've been pulling my hair trying to figure this out. I got a new laptop and installed a newer version and now this shit drives me nuts. Capacitors are the only ones that don't have this issue. L and R are fine when placed horizontally but for the vertical ones the font size is set to 0.625 instead of the default 1.5. And it is 0.625 for all other components no matter the orientation. Wtf.

>>1764253
I ignore it, ltspice was never a terribly polished product

>>1762999
Why is it that whether or not I use a darlington, the pulldown-stage is still just as slow? Where's the current limiting coming from? Changing the bottom resistor to a smaller value does work, but I don't like to stray too close to the 1/4W limit. Theoretically, using a darlington should increase the pull-down current by a factor of the hFE, and that's not happening at all. Shorting the bases of Q2 and Q4 just increases the amplitude of the resultant wave slightly.

I'm testing it by driving a 22nF ceramic cap with that circuit, and I am using an LM393 after my function generator. Up around 10kHz the voltage on the cap only gets 3/4 of the way to GND before it gets pulled up again.

Also belated checking my own trips on the board where it matters least.

>>1764272
Measuring the slope of the voltage at the capacitor:
>∆V = 0V - 2.4V = -2.4V
>∆t = 1.6µs - -3.2µs = 4.8µs
>∆V/∆t = -500,000V/s
>I = ∆V/∆t*C = 11mA
That's not even 10 times the current going through that 10kΩ resistor, what the shit?

>>1764280
>pic related has the same curved pulldown delay on Q1's collector
Wait, is the transistor junction acting like a capacitor? One big enough to matter with respect to the 10kΩ resistor? Or is this some sort of delay caused by the transistor being oversaturated? Increasing the base resistance doesn't help. Either way it turns out that the totem-pole was working without a hitch.

>>1764287
>replace Q1 with an NPN in common-collector format
>use resistor as pullup instead of current limiter
It's not awful anymore. Now, even outside of totem-poles, I have a newfound respect for common-collector transistors.
Pic related is at 150kHz with the darlington pulldown, I might want to adda darlington to the pullup side too, it's still a little slow but it's a lot better than it was.

>>1764291
And now with the double darlington we've got this when at 200kHz, which should be good enough. Note that I'm currently short on 1N4148s, so this is without the diodes. Adding the diodes might improve the characteristics further, but we'll wait and see. If I was actually switching a MOSFET at 200kHz I wouldn't want to do it with this waveform, especially not switching potentially 60W of load, so I'll actually be switching it closer to 20kHz, maybe as high as 50kHz.

Oh and there's also the whole "it isn't very close to the 12V rail" issue to take care of. Putting a big filter cap from +12V to GND raises it from 6V to 7V, using 5k pullup instead of a 10k pullup increases that to 9V. All I need is for it to be above the PFET's threshold voltage, and -3V isn't quite at the -2V minimum (or is it maximum). So I'll see about it when my diodes eventually arrive.

>>1764253
y u no Micro-Cap?

>>1764295
small Schottkys would do well too, possibly with a lower Vf and lower reverse capacitance. but you actually want the Vf on D? to be relatively large
any readings on the driver transistor currents? if they're too low you have little choice but to turn up the bias
there is a finite recovery time in which a saturated transistor's charge is pulled out of the base so that it can return to the active region. active pulldown is quite pleasant and easy to use
if Vgs(max) is too high, just bung a zener in there and you're sorted

>>1764315
>you actually want the Vf on D? to be relatively large
Yes, in order to decrease the voltage difference between the PNP turning off and the NPN turning on, similar to a class-AB amplifier, IIRC.
>any readings on the driver transistor currents
No, but I believe my most limiting feature at the moment is the current sunk by the comparator through the pullup. The comparator is open-collector so I'm not too sure what the best way of advancing besides just using a power resistor is, but it's only rated at 20mA anyhow. Is it darlington time again?
>active pulldown is quite pleasant and easy to use
I can't for the life of me figure out how I'd do that without a pullup or pulldown somewhere in there. The comparator output is a current sink, and I already tried using it to pull current directly through the base of a PNP and it gave me this: >>1764280.
My transistor datasheet says nothing about recovery or slew time, but it has a gain-bandwidth product in the realm of 300MHz, while the comparator has a response time of 1.3µs, so going up to 100kHz is likely not a good idea judging from that alone.
>just bung a zener in there and you're sorted
Wait, a zener between the final two transistors, with the FET gate on the upper side of it? That makes some sense.

I actually have some RF BJTs lying about that I could potentially use for this, but I don't have complimentary pairs and I'll need to mirror this circuit for the other rail, so I'd really rather not go for that route.

What kind of protection is important for measuring residential mains voltage?
I want to do a project that would keep measuring it 24/7.

I'm not sure if I even need any since it's high impedance.
I often see people complain about cheap multimeters lacking metal oxide varistors but barely any mains device in my house has them and I never had a problem with surges.

>>1761861
Just use Kicad, creating parts is easy peasy in Kicad. And you don't have to pay for a subscription and even get step export for free. Learning Eagle isn't worth it at this point in time.

>>1764323
I'd just throw it through a transformer

>>1764331
I'm not worried about electric shocks since I'm not gonna be touching it.
The data would be send to another device wirelessly.

I just don't want it to burn the house to the ground while I'm sleeping.

>>1764334
transformer is still a pretty safe method, but capacitive coupling might give you lower q-current, and if you use mains-rated X caps it should be safe enough

>>1764334
>I just don't want it to burn the house to the ground while I'm sleeping.

if you're a worrier, just put two small (i.e. low-current) fuses inline with both power lines and sleep soundly.

i want to make a board for programming dip atmega328ps, has anyone here done this before, and is there anything to watch out for? can i just leave most of the pins not connected? i want to program fuses too

>>1764343
It's very simple.

pic related is all you need (you can also drop the cap for AREF, it's only necessary when you use the ADC)

Audio output on my computer puts out like 2V amplitude wave to power mz headphones. But what if i want to connect it to a big speaker which has no amplifier.? Will i need bigger amplitude or still only` 2V but much more current?

>>1764346
that voltage will result in a certain power when connected to a speaker with a given impedance, per P=V^2/R. it'll work as long as that P is within the mobo's limit, and also loud enough for you. otherwise you'll need an amp.

>>1764346
You'll need both a higher voltage and a higher current-carrying capability. Your computer's audio output will drive 32Ω headphones just fine, but 8Ωs will likely be hitting some sort of current limitation. They'll probably still sound fine, but essentially they won't be being driven at the full 2V. There are a few amplifiers you could go for, ranging from a simple LM386 and a few passives, to one of those plug-n-play 50W class-D modules for $10 from ali, to an enclosed home-audio amp with built-in power supply. I'd personally lean towards the class-D just for value per money, and I've heard they sound perfectly fine too, at least some of the models anyway.

>>1764348
Pretty sure the built-in class-D of the computer will have a current limit that's lower than the mobo's total power limit, if that's what you were implying.

>>1764207
Thanks for the tips.
> maybe share the display (not shown) with the nRF chip instead of the SD card.
Not possible, nRF is has an interrupt pin, there could be a problem when the mc could be interrupted to process radio message while it sends data to the display, I'd have to do some trickery to alleviate that.

Could a 100nF or so cap get rid of any glitches at the spots marked with a red circle?
Propagation times shouldn't be more than a 100ns.

>>1764540
I hate that fucking IC, every time I ordered a board with it I had to add a different component to make it work. CLK nad to have a resistor in series and a capacitor. Then I had to add a big-ish capacitor for the power, then I had a bleeder resistor to reset the circuit while it was powered down. On the other hand, that "feature" could be used to implement a type of memory.

>>1764319

100kHz might be a little bit aspirational for the components you're using. 60kHz is still relevant and useful in an application like this, but you'd want to reconsider your inductor current ripple
bear in mind that driver transistors are a medium-power application, not something that the jellybean 100mA Ic transistors are going to handle well. by now I would have opted for an integrated MOSFET driver
>Wait, a zener between the final two transistors, with the FET gate on the upper side of it?
sorry to be unclear, I meant a zener somewhere on the base side of the complementary pair, maybe in parallel with Q1, to limit the Vgs of the MOSFET to acceptable bounds, not necessarily to help your speed. not very relevant if your V+ supply is smaller than Vgs(max)

>>1764334
if you can guarantee adequate overvoltage protection some other way, you might not need MOVs. e.g. if you use a 2.49MΩ/10kΩ 1/2W resistive divider across the line, it's important that the line input stay below 3.3 * 250 ~= 830V in a 3.3V system. you could use TVS diode? zener? just Schottkies to the rails? to adequately protect the low-voltage circuitry. protection requirements are different for a power supply input

>>1764363
>nRF is has an interrupt pin
is also has Tx and Rx FIFOs, each 3 packets long. you can probably afford to be a little bit lazy about servicing it
how's the display's maximum SCK speed? you could run the card at up to 25Mbps according to spec but the nRF maxes out at 10Mbps. I think the blocks being sent to the flash card at a fixed 512 data + whatever command would be longer than those sent to the display unless you're sending very many full frame full color bitmaps
with all that asynchronous periphery in the mix, I'd reach for FreeRTOS pretty quickly, giving the SPI port its own task to listen on one or more queues, run bus transactions (with DMA if you're feeling ambitious), and dispatch response data to requesting tasks

>>1764540
do a Johnson counter

>>1764543
>do a Johnson counter
It is one.
And i think an actual Johnson counter in a single IC would have the same glitches if i ANDed outputs next to each other.

>>1764541
You wat mate?

>>1764547
>You wat mate?
I hate the 74xx74. I used it in a motor controller. The motor drives something all the way back until that thing touches a sensor, then you push a button on the controller and it brings that thing back until it touches a sensor.

Got my ESP32 to display text!
Now how the fuck do I set up AVRCP?
I searched around and didn't find any libraries to #include: or anything like that.
Am I missing something?
Or are bluetooth profiles done in a completely different way?

>>1764543
>100kHz might be a little bit aspirational for the components you're using
As I said in an earlier post, I'll likely go for something between 20k and 50k.
>medium power
>100mA Ic
I figured that since the current will only be high for such a short time that it won't matter, and since the datasheet doesn't say anything about maximum peak current, or have a current-duration graph of some kind, I suspect I'm not out of spec here. Still, these aren't the fastest transistors, and with my existing circuit I'm pushing somewhere between 0.6 and 1 amp on my pull-down stage, for a few hundred nanoseconds.
With the 5 transistors I'm currently using instead of the 3 I thought I'd need, a dedicated MOSFET driver IC isn't a bad idea at all, since I'll need to order half my parts anyhow.

>>1762448
can you spot the contradiction in this post? :)

>>1764621

is there a prize?

>>1764345
and done
but I can't find my arduino nanos in order to prototype and program with this
wait fuck i soldered vcc and gnd backwards

>>1764687
ok could be worse
as well as the arduinos i also cant find my colored wire

>>1764689
shit nevermind it was the right way around

>>1764256
yay just got the latest release from Feb 6 and it is 99% OK. the only thing i noticed that's still in a tiny font are power source labels.

brainlet here

could I get these solar cells, cut them to my dimensions, and it will just work? https://www.aliexpress.com/item/33001187620.html

>>1764706
I wouldn't, unless you had a waterjet. The silicon wafers are somewhat easily cracked, plus they have wiring near the edges.

>>1764707
rip then, just wanted a ~40mm length, ~10mm width solar cell :(

>>1764693
everythings at the right voltage, but avrdude + chink avrisp just isn't working
it wants me to update the firmware on the avrisp but i dont think i can for whatever reason
guess ill use arduino as isp like i always do

>>1764708
check alibaba maybe
perhaps go for amorphous solar cells

>>1764709
oh so thin film is good to cut up? noice

https://www.aliexpress.com/item/33020111462.html

>>1764710
maybe? it's 2v so theres at least some series cells in there and you cant see the connection wires from the images
if you cut the wrong wire you could render it open circuit

>>1764714
what happens when it becomes an open circuit? shit just won't work?

>>1764715
it wont work at all

Absolute noob here, looking to buy some equipment to get the hobby ball rolling.

What is the difference between these two multimeters considering the price difference?

Will the cheap one be alright for me?

Which one do I choose? I don't know the difference.

>>1764731
Get a t12 ks-ger unless you have a specific reason for the meme pen. It will come with more tips and a better psu.

>>1764734
Nah I want the meme pen. I'm typing from a 2019 Macbook Pro and still defend Apple, I am far past the point of being reasoned with.

So - tip suggestion pls senpai.

>>1764736
KYS

>>1764737
Listen, if I don't get a suggestion from this board I'll have to get it from a memetuber.

Do the right thing.

>>1764734
>>1764737
Memetubers came to the rescue. I made sure to SMASH that like and subscribe button.

And I unironically did in fact hit that bell button to receive future notifications of their videos.

>>1764723
Top one lacks autoranging, and looks a bit shit, while the bottom one is more full-featured with autoranging. But none of them have any brand to speak of.
So buy an ANENG AN8009, it shouldn't be more expensive than the bottom one and it's a tried and tested meter.

>>1764731
K for soldering on larger stuff, D24 for smaller stuff. KU might be a reasonable choice for medium-small stuff too, but I'm not sure. Personally I'd look for a D12 since I'm wanting to get into SMD soldering, but that's only because I already have a K. Avoid conicals though.
Considered getting a TS80 instead? It runs off USB C, meaning if you have a USB C macbook you won't need to get a new charger. Might even be able to run it off the macbook itself, I'm not sure.
Tips are more expensive though.

Still, the T12's smaller pencil handle and full-featured display is more convenient. There are some comparisons between TS100s and TS80s and T12s on youtube, if you'd like to get some actual data.

>>1764748
A hero appears. Thank you for your consideration.

The TS80 is £15-£20 more expensive which I could use on other components/tools to bankroll this hobby.

Am I being shortsighted? I have watched comparison videos but a lot of the differences come down to peoples preferences and my inexperience means I don't yet know my preferences.

>>1764750
Well the T12 with PSU costs as much as the TS100 without a PSU, if cost is your driving factor. If you never intend on using it on the go, there isn't any particular reason to go for the more portable designs over the T12. And again, without the power supply a TS80 might be more reasonably priced.
I just watched a comparison and saw that while the TS100 was 60° under the set temperature, the T12 was 30° over temperature, which reminded me that I've never actually tested my station to see how well it's calibrated. I've got a thermocouple module and a thermocouple for it lying about that I could use, so I might do that soon.

Are wide bandgap (SiC, GaN) power devices a meme? Have you ever used them?

>>1764723

The first one is infamously bad meter, don't even look at it. It also doesn't have as many features as the second one. But the second one is also probably crap. There are a hundred thousand shit multi-meters out there with the same junky circuits inside.

The ANENG AN8008 fine for a beginner

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

>>1764757
I've never used them, at least I don't think I have. Also aren't BJTs better at high frequencies than MOSFETs? I never see MOSFETs in RF circuits but I see BJTs all the time, along with JFETs. Like the 2N5109.
Also the graph doesn't mention thyratrons or other high power tubes that are still used in substations and the like. At least I think they're still used.

>>1764709
>everythings at the right voltage, but avrdude + chink avrisp just isn't working
>it wants me to update the firmware on the avrisp but i dont think i can for whatever reason
Those 2$ USBasps? That's normal. To get it to work with avrdude you have to flash them with a fix.
Look up the exact version you have (should be written on the PCB somewhere) and google for it.

I bought this LED bulb but I have a problem, maybe one of you will know what to do?

I thought the light would have some ability for me to choose a colour from the entire spectrum rather than be limited to the 12 preconfigured colour choices. It does not.

The thing is, there is a button that slowly changes the colours throughout the spectrum, so the LED is capable of displaying the colours that I'd want, it just happens to not have the ability to be programmed into the remote or an ability to pause.

An example: There is no warm cozy light button, but I can see the exact warm lighting whilst it's going through RGB. If I could freeze it, I'd use that light as the primary light in my room.

What do?
How do the presets even work?
Is it hackable?

>>1764763
8009 is the same but upgraded

>>1764775
i tried once before but ill give it a shot again i guess

>>1764776
Depends on whether the control circuitry is in the bulb itself or in the remote. Since those things usually remember their past state even when turned off, I'm guessing it's all in the bulb, which means the bulb's control circuit will likely only be able to receive signals from those preset colors, and it will cycle through the colors smoothly with or without the remote being turned on. So there's nothing you could do with the remote to pause it in the middle of the cycle.
But it's possible that the remote isn't using all the possible presets for the built-in controller, or even that the remote sends arbitrary RGB signals. I'd get an IR receiver and watch what the remote spits out with a scope (or $7 USB logic analyser) to see what format it's transmitting. Then if there are any unused codes that fit in that format you could try sending them with an IR LED. Pulseview (the software I use with my logic analyser) might even have an IR remote decoding mode or two on it, it's pretty good software that can decode plenty of digital standards.

>>1764776
>>1764780
Oh and if you don't have a dedicated IR receiver (or have one but for a different frequency) you could just use an LED as a photodiode and use a couple transistors and/or diodes and a low-pass filter (or monostable circuit) to demodulate it.

>>1764780
That sounds fun - you've unlocked a new learning tree for me. Thanks.

>>1764571
Dude.

Imagine. You can write your own libs.

>>1764547
the inverted feedback defines the Johnson counter. yours is a one-hot counter. if you used a proper Johnson counter, CLK01 would be the output of one of the two FF stages and CLK23 would be the inverted output of that same stage, no diode fuckery required, two less pins on the quad gate you use to decode CLK0..CLK3
a 100nF might be too big, might be too small. "any" is a big word, depends on logic family you are actually using, its switching speed, its drive strength, the xx74's Q to !Q skew,

>>1764551
nothing wrong with the chip itself. maybe long switch wires and too-small pullups

>>1764584
assume the limiting Ic value is instantaneous, not average, if they don't call it out specifically. hot spots are no good. you could improve your luck a bit by heatsinking the body and using extra copper at the collector/drain terminal

>>1764571
possibly related API docs https://docs.espressif.com/projects/esp-idf/en/latest/api-reference/bluetooth/esp_avrc.html
a functioning sample project into which you should be able to add your metadata callback code https://github.com/MhageGH/esp32_BT_Speaker/
the patch to the SDK that implements AVRCP metadata functionality, for reference https://github.com/pufstudio/esp-idf/commit/645588c173ab9b4d45d16c3210f2afa3baedb6bd

>>1764798
>you could improve your luck a bit by heatsinking the body
The final NPN is getting up to ~50-60C, so you're definitely right about it being a medium-power application. I'm going to be buying some complimentary power BJTs anyhow for my emitter follower section (lower dropout than a source-follower), so when I get those I'll see whether it works fine with them, or if I should go for a dedicated driver.
>extra copper
I may or may not be testing this in a 50 year-old solderless breadboard.

As for that Johnson counter thing, I've always found it confusing that the inverted ring counter and the single-bit non-inverted ring counter are referred to as "ring counters", and sometimes both as "johnson counters" too. This is the first I've heard the term "one hot". The non-inverted ring counters rely on bit presets, which I've always kinda ignored in digital logic circuits as they're not part of the basic gates.

>>1764813
Oh yeah, are the TIP41 and TIP42 good choices? Or are they ancient? With their subpar hFEs I'd maybe need to use them alongside two other transistors each as triple-transistor darlingtons.

And a datasheet I'm finding for the BC547 says: Ic = 100mA, Ic (continuous) = 2mA. So I'm probably well out of spec. I figured that all TO-92s could handle the same kind of punishment, but I was quite wrong. The same source tells me that the 2N3904 and 2N3906 are rated at only 1mA continuous, for what that's worth. I guess it's the die size and hot-spots, like you said. Should get some BC327s and BC337s instead, or equivalent.

II
W

>>1764827
i will now draw my LEDs as faces
I I
W

>>1764820
TIP are ancient, kinda slow too iirc. those BC327/337 would be fine
that site might be on drugs. just pulled up the ON MMBT3904 datasheet and they're rated for 200mA

>>1764830
owwwo
>resistor with happy terminals

>>1764831
>that site might be on drugs
you're expressing a form of prejudice

>>1764820
Get 2N7000s instead

>>1764834
no?

>>1764834
2n7000 is weak as shit, they're for logic level conversion stuff

AO3400 and AO3401 can motors

>>1764873
>can handle motors

>>1763570
try swapping p104, the ic could just be broken

Years ago I worked for a company where I did schematic capture and PCB layout for boomer EE's who couldn't use computers. I want to get back into it.
Anyone have any designs they want made into circuit board manufacturing files for free?

>>1764834
>>1764873
Why are you suggesting MOSFETs, I want BJTs, both for driving FETs as quickly as possible (though I'll probably just use a TC4427 dedicated driver or similar) and for making an emitter follower, for which low Rds-on means nothing and gate threshold voltage means FETs are innately less suited. As was implicitly stated in the previous post >>1764813. Not to mention posting SOT-23/TO-92 MOSFETs, which likely have higher junction to case and case to ambient thermal resistances than the TO-220. I'll be using the power transistor as a linear regulator, so I'll likely need at least 2.5W of dissipation.
I'm assuming that those FETs will switch my main FET's gate slower than a BJT making them subpar as a driver, but I might be wrong on that.

I plan on using FQP30N06 and RFP30P06 as my main switching FETs, and while the AO3400/3401 could probably replace these if I wanted to, I'm a primitive THT boy who sees a TO-220 as a D2PAK waiting to be trimmed to match the footprint, hence there isn't really a reason to buy dedicated SMD FETs. Unless I wanted them small, but I'm using like 5 quad op-amp DIP14s so that's not really relevant.
But I'm open to a switching FET that has a lower gate charge than the 30N/P06; I have been testing with a 22nF cap as my mock gate after all, and I haven't ordered them yet.

>>1764905
kicad is fun and engaging, why outsource it?

I want to make a solar battery backup for my phone and gopro that I can take backpacking and camping. I am new to electronics but have been reading as much as I can.
This is what I'm thinking so far:

Solar cells (~7-8V normal conditions) ---> li-ion charger ic ---> li-ion battery (2 3.7V cells in series) ---> linear voltage regulator with USB output ---> device (phone or gopro)

I selected the cells because they would fit in the package I want to use and I plan to put them in series to meet the 5-5.1V, 1A output needed to charge my phone.
1. do these numbers make sense? i.e. will it work?
1a. do I need more voltage to charge the li-ion battery if they're in series?
1b. does keeping the two cells in series complicate their charging, assuming i have enough voltage from my solar cells?
2. is there a serious concern about housing this in a short PVC pipe? I'm thinking static e- buildup? (I want to wrap the flexible solar cells around the outside so it's easier to hike with)
3. if I need a heat sink on the voltage regulator will it be a big deal if it's enclosed in a water-tight PVC housing? will it get too hot if I have to shed 2.4V?

>>1764205
Giving it another thought... Yeah this makes sense, I used high side since it is easier to think about measuring things relative to the ground with nothing in between. High side works well for RC in series (not sure if the low side will even work for RC?). As for the comparator, that sounds like a good idea: instead of continuous PWM, can generate a one shot pulse, bias the comparator to 0.63 VCC and measure the delay by feeding the output to ADC or even to a digital port and detect the edge via EXTI, and do that in a loop a few times and calculate the average. The only problem is this method can only work for relatively high inductances, say >1mH. Anything lower requires a much faster ADC since the R value has to be high enough to keep the current low. Or have to measure visually on the scope. Or maybe measure 2t or 3t instead of t but that may result in a significant error or the comparator should be super precise.

>>1764757
>>1764768
That table is garbage, those ranges are way off. BJTs alone take up a much larger space than what's shown.

>>1764955
Do not wire liion batteries in series. Use a booster circuit to go from 3.7 to 5v.
They make liion chargers that support a wide input range. Get one that supports the voltage your stomach cells put out under load

>>1764962
*Solar cells under load
Autocorrect blows

>>1764963
No worries
>>1764962
Can I put them in parallel to get more mAh? Also wound that in any way waste the extra voltage from the solar cells?

Thank you for the reply btw

>>1764957
Here is what I am thinking. Don't even need to care about syncing with the input pulse. Just measure the output pulse width and thats your time constant. Swapping L and R made it a bit easier. No need to mess with ADC, can do all digital EXTI raising/falling edge trigger/measurement.

Alright, so i want to be able to save data on my micro when a power out happens without doing some battery shit, so i was thinking
Put a 10k uf cap across the + and -, and considering the micro draws 10ma that means when the power goes out the voltage will start to drop at a 1V per second rate, i will then use a voltage divider, to turn the input 5V into 1V and measure that with the micro and compare it to an internal 1V reference.
So when the power goes out i will be able to detect the falling voltage and do a quick data save, in about the 1 second of time before the voltage dips too low
Not to toot my own horn, but that is pretty fucking smart

Just bought a cheap (but much better than my current) multimeter (AstroAI DM6000AR). Would it be worth it to go ahead and get some good leads? The ones it came with seem ok, but if I got a good set I could at least be sure

>>1764881
Yeah, i tried that, swapped p103 and p104, same results, so it can't be the IC

>>1764957
Yeah, that's just the limit of an RL constant measuring method instead of using an LC oscillator. But at least you subvert the ADC's speed and use the speed of the digital input, which should be much faster. Plus you could even calibrate it with known inductors and get even higher resolution.

>>1764996
But measuring the output pulse-width is no more precise than measuring the single time delay. You get the bonus of it not being synchronised to the main clock, hence theoretically all your measurements will average out, but I'm pretty sure you'll get inaccuracies as well. Just look at the slope on Vcomp_out.
I'm almost certain that your system will be more than fast enough with the first method because the GPIOs on an stm32 are 10s of MHz fast.

>>1765113
I'd be more worried that the switch is broken.

>>1765113
>>1765172
Never mind you measured the voltage and it said the switch was fine. Did you measure the voltage at the IC's pin1?

I asked this in /sqt/ but was told I'd have better luck here:
I have an old amp I got used for cheap. I noticed some of the componemts get very hot with use, specifically two transistors. They are the ones marked in red in the schematic. After using the amp for about an hour, their heatsinks are so hot that I can't keep my finger on them for more than a few seconds (I have no other way of measuring the temperature, my thermometer only goes up to 43°C).
Some more pictures:
Whole Amp: https://0x0.st/i-Hf.jpg
Q801: https://0x0.st/i-HV.jpg
Q802: https://0x0.st/i-HW.jpg
Top case: https://0x0.st/i-H4.jpg
The bigger of the two discolorations is above Q801 and the smaller one is above Q802. Both were already present when I got it.

I did replace the thermal compound on the main amplifier chip (IC701, a STK4191) and on Q801 when I was cleaning the amp (it was full of dust) so that can't be the problem (also, it's the heatsinks that are hot, not only the transistors themselves)

Is it normal for them to be so hot, or should I be looking to buy some replacement parts? Do I maybe just need bigger heatsinks?

>>1765174
Those two transistors make up the + and - rails of the linear power supply, it's understandable that they get pretty hot. I'd measure the voltages across the zener diodes ZD801/802, the DC voltage coming out of the bridge rectifier, the voltages at the two transistors' emitters, and (when off) the resistance of R802 and R807. The reason for testing this is to ensure that the transistors aren't dropping more voltage than they should be. Theoretically you'd only want them dropping 1-2V, meaning they shouldn't get that much hotter than the bridge rectifier itself. If the zeners or power resistors are damaged or have otherwise changed their values, they could be causing the transistors to dissipate more power was designed. But for them both to do the same thing makes me think it's just working as intended, it's not uncommon for transistors to work hard enough to get too hot to touch, and this is definitely an application where it's possible. But it's also possible that something further down the line is drawing more current than it should, so keep an eye out for anything else getting too hot.
The transistors are there in order to remove ripple from the power supply, so measuring the ripple voltage at the output of them would also serve to be some reasonable troubleshooting.
>it's the heatsinks that are hot, not only the transistors themselves
Yes, that is the point of a heat-sink.

>>1764966
Lipos of your choice may be able to deliver 10C. Some can do retarded 30C

>>1765196
Voltage across both diodes is 15.1 V
The outputs of the bridge rectifier are -43.7 and 43.4 V respectively
Emitter and Base and Collector voltages are 14.4, 15.1, and 38.8 V on Q801 and -14.7, -15.2, and -40.5 V on Q802
The resistors have 56 and 150 ohms, as they should.

Also, in the 2-3 minutes the amp was turned on both heatsinks got warmer than 43.9°C.

>Yes, that is the point of a heat-sink.
I mentioned that so it was clear that the thermal coupling wasn't the issue.

>>1765215
Oh shit on closer inspection they're using 46.2V as their ± rails and dropping up to 32.1V across the transistors. No wonder they're getting hot. If anything the 40.5V and 38.8V are a little low, but that might just be a difference in measuring since it will have a substantial ripple on it.
If it's 40.5 down from 46.2 across a 150Ω resistor, that's 38mA, so 0.98W dissipated in the transistor. I'm assuming the speakers themselves are driven off the ±46V rails, so this power burnt through the transistor won't go up appreciably when you're playing music loudly. Because of that, if the transistors haven't burnt themselves to death, I'd say it's working as intended. If you really wanted to you could add a pair of SMPS before the transistors in order to decrease their load, or add another lower voltage winding to the transformer (don't know why they didn't do that in the first place), but it's a big hassle. Larger heat-sinks might increase the longevity of the transistors, but since it's been like that for so long it's probably going to be fine, and it looks like it might be hard to fit them in. If you're still worried you could check the datasheet of the transistors and see what kind of power they're rated for, but accounting for the heat-sink means measuring the temperature of the system or getting a good estimate of the heat-sink's thermal resistance to the case and to ambient. What vintage is it anyhow? For them to skimp on both extra windings and a switcher leads me to believe it's old and cheap.
>I mentioned that so it was clear that the thermal coupling wasn't the issue.
Oh right, sorry.

>>1765220
>SMPS
It took me an embarrassingly long amount of time to figure out you meant switch mode power supplies, but I don't think there is any room.
>another winding
there is another winding that has ±1.6 V for the VFD (no idea why it needs that), but I doubt that's useful.
>heatsinks
I.was thinking more along the lines of connecting them up to the giant heatsink the STK4191 has, because that one never seems to get hot, only slightly warm. I'd just need a good thermal conductor, bend it into into shape and maybe drill some holes. Or would thay be a stupid idea?
>what vintage
It's a Blaupunkt A-5600 which is just a rebadged Hitachi HA-2, made 1983-85. According to Hifi-wiki it cost 1000 DM when it was new (weirdly the Hitachi only cost 690 DM). I got it for 5€ and I don't really plan to stick a lot of money into it.

>>1765240
>Or would that be a stupid idea?
It's not stupid, but if you drive the amplifier hard heat might end up flowing back the other way. Still, if you've got some copper pipe or whatever lying about you could give it a shot.
>83-85
I expected earlier, they really should have put another winding on there. Still a bit early to jam a switcher in there though. Also what's that price in freedom units?

>>1765252
The exchange rate fluctuated between 2.5 and 3.15 USD/DM in 83-85 so 1000 DM would be anywhere from $320-400.

>>1765255
That doesn't sound cheap

>finally run out of lead-free solder
>swap to my big reel of leaded solder
>turn iron from 350 to 320
>solder doesn't melt quickly and puddles grow slowly
>turn back up to 350
>feels like normal
wait what the shit was i alway soldering with 60/40?

D-did I order the wrong thing? It doesn't fit in from the other side because the bore is too small, and from this side only 2 of the 4 contacts actually make contact: the front two of the socket on the middle two of the jack. The second-to-the-end pin of the socket is too high up to touch into the notch in the jack, and the one on the end is nowhere near the jack at all. The jack's sleeve is left unconnected because that front plastic cylinder covers it all. I think it would work fine if I sanded down the cylinder of plastic, but that feels like a really janky workaround and I can't be bothered to use it on all 10 of these, especially since I don't have a belt-sander or dremel.

And I can't even leave 1 star because then chang would know that i received it and ask for the refunded money back because it arrived 130 days late.

>>1765240
>I.was thinking more along the lines of connecting them up to the giant heatsink the STK4191

this sounds like way too much work for unknown gains. what i'd do is pop in a small thin 12V fan from, let's say, a 15yo graphics card. you couldnt power it from the +14V coz that would just heat the transistors more, and you couldnt easily power it from 43V coz a dropping resistor or regulator would get too hot. so i'd just run in a wire from a 12V charger thru a hole in the case. that means your vintage amp stays vintage, as no permanent mods were made to it.

>>1765313

betcha it'd work if you cut off or ground off the lip at the front of the jack.

>>1765313

betcha it'd work if you cut off or ground off the lip at the front of the jack.

circumcision: not just for jews anymore

>>1765315
>feeding a noisy 12V source into a noisy fan in the middle of an audio circuit
It better be a brushless fan too, not sure what gpu fans were like 15 years ago, else that brushed commutation will play horror for your circuit.

>>1765319
The lip on this jack is already comparatively small, there's no way the socket was designed for thinner aux cables than this. Especially since many sockets rely on the jack's lip to prevent overinsertion; they don't have end-stops on them.

>>1764955
1. meh
1a. yes
1b. aka 2S, yes, you will need protection circuitry specific to your pack layout. better to use 1S2P
2. yes, Li+ batteries really hate heat. make sure you use the thermistor temperature-limitation option on your charge controller
3. the mere fact that you're even considering using a linear regulator on a battery-powered project is a tell that you're in way over your head. you should probably look at pre-made power bank controller boards on ali instead
do not try to power a device through a charge controller's battery-output leads
do not place loads on a charger's battery leads unless you will never be using and charging at the same time. if you want to charge during use like a phone does, look up power path management aka system load sharing and use a charge controller capable of that
do not build packs out of cells which have seen different wear patterns

>>1764960
for switching? doubt it

>>1764998
9/10 nobbad. look up dying gasp controllers

>>1765313
use nippers to remove the vestibule. that's much less janky
TRRS jacks with shorter front clearance do exist. a designer might want to use those to allow for a challenging design with a front panel of nonzero thickness but no jack well

>>1765329
>use nippers to remove the vestibule
Won't that crush it? Also there are sockets with similar form-factor that have the pins in positions that my jack can use:
https://www.aliexpress.com/item/32788968290.html

Here's the one I bought:
https://www.aliexpress.com/item/32842525190.html
Just scrolled down in the reviews and saw this:
>The connector is not suitable for most headphones-the left pin does not even touch the tip of the headphone plug.
>You have to cut the belt, so that the headphones are correctly connected to this socket.
Shoulda kept my eye out I guess.
I found a datasheet for a similar model:
https://www.quick-teck.co.uk/Management/EEUploadFile/1466168136.PDF
But it doesn't provide much information besides using the word "audio" once. I think it's for some other sort of TRRS, one with a sleeve as long as a normal TRS. I know I've seen TRRRS pictures somewhere, so the longer form-factor does exist.

>>1765172
>>1765173
The switch is fine, swapped it out with another and no change. Assuming I measured it properly, voltage on the IC pins looks fine.

A few thoughts after some more troubleshooting:

Do all outputs of the CD4021 need to be hooked up? Currently pin7 on p104 is not connected to anything.

I've used 1k ohm on the pull down resistors, but the arduino guide for shiftin that details how to use the cd4021 has 10k ohm. Not sure if that might be an issue, as I said, all the other switches work fine, so seems unlikely.

>>1765397
>Do all outputs of the CD4021 need to be hooked up?
Well I doubt 4000 series CMOS logic has internal pullups, if that's what you're asking. Whether or not the erroneous readings from one of the inputs will matter to the circuit is anyone's guess, but considering CMOS inputs left open can behave quite unpredictably I'd give it a pullup/pulldown resistor just to be safe.

Any reason you chose the 4021 over the 74HC165 while we're here? I bought a boatload of 165s for a similar (musical) keyboard project, though I haven't gotten around to making the boards or figuring out the software end. But I was making a modular system so you could have as many octaves as you wanted just by plugging the boards together, with 2 SIPOs on the primary boards and only 1 on the secondary boards, alternating P,S,P,S, etc. IIRC, there were pullups on all the buttons with the initial values being high and a pulldown on the shift-input, which allowed for the sampler MCU to run the clock and figure out how many shift-registers (and hence octaves) were connected to the system, for auto-timing reasons. No debouncing either because the buttons aren't clocking anything and the bounce-time is shorter than the time between each sample set.

>>1765402
> I'd give it a pullup/pulldown resistor just to be safe.
I have to do another version of the board and get it refab'd anyway for other reasons, so I'll add it in.
> Any reason you chose the 4021 over the 74HC165 while we're here?
Lack of knowledge I guess? The arduino guides have the 4012 in the examples so I just went with them.

>>1765405
>so I'll add it in
Test it with a bodge now before you order the new one, so you can be sure that you won't need another fix later on.

>Lack of knowledge
I wouldn't say that the 4012 is any worse than the 165, since the 4000 series logic is more robust to different voltage power rails, but I think its propagation speed might not be as good as it could be. When I was learning digital logic I learnt it all with 74LS stuff, so I guess it's just a difference in origin. I still hardly have an encompassing knowledge of IC families, be they digital or analog, so I too stick to what I know.
I'll just note that 74LS logic is probably the most beginner-friendly, as it has internal pullups (or pulldowns?) on the inputs and current-limiting on the outputs, and isn't nearly as ESD sensitive. Not that you can buy everything in 74LS series that is.

i want to add a battery as a backup power source for my raspberry pi in case the power goes out
I will be using a two cell 7V lipo, with a buck to get 5V, the back output will be connected to the 5V rail and GND pins on the raspberry.
The main source of power will be coming from a typical wall wart with micro usb.

If i simply hook the buck output to the pins it will probably melt since the wall wart will back feed into it right?
So i was thinking, will adding a single diode on the + wire between the buck and 5V pin do the job?

The 5V outputs clean 5V when the raspi is powered by the wall wart and the buck will also output 5V, and the diode will refuse to conduct anything if the drop across it isn't at least like 0.5V right? So as long as the wart supplies power the diode will keep the buck cut off.
Does that sound good so far?
The problem is that i can't actually use a diode since the raspi can draw 3A which would melt the diode.
So i probably need to change it up so the diode perhaps opens a mosfet?

I'm trying to understand what kind of transistors to use for certain projects, having trouble with getting the datasheet characteristics. I.e. I want a 16v potential to pass when given a 5v base potential. What kind of transistor should I look for? Is there any location where I can find transistor types based on characteristics filtering?

>>1765592
BJT? FET? What kind of currents? Small signal or switching?
16v is no big deal with a 2n2222, but your input/output will be inverted.

>>1765592
trannies are current driven fren

Where the fuck do i start with this

>>1765541
draw picture, /ohm/ RULE 2
>The problem is that i can't actually use a diode since the raspi can draw 3A which would melt the diode.
3A Schottky diodes are fairly cheap and common, and relatively low in voltage drop. raspbian has some hacky fixes that allow it to deal with a shitty wall wart by shedding load by lowering core clocks and/or disabling cores
a synchronous buck converter could be destroyed by being back-fed by the wall wart supply. however, in a non-synchronous buck converter, the inductor will just pass current through to the input and survive, maybe even start running, which could destroy your battery and your charge controller
and the buck is still possibly running while charging, draining the battery as it fills and disturbing the charge controller's view of the battery's condition
5V UPSes are a popular project and are much harder than they look if you want to do them well and safely. seriously, google for other people's designs and learn by immitation crab meat

>>1765677
when in doubt, recycle it

>>1765616
In this specific case, 16v signal to power small model train traffic lights (alternate current), and 12v dc to power the (small) model trains (I'm not sure how much current they use). I'd like to power them by either an arduino/pizerow (5v triggering signal) or an esp (I believe these are 3.3v ts).

The main thing I'm struggling with if for example I take the 2n2222 datasheet is you have the collector-base voltage (60v), and the collector-emitter voltage(30v), when will this trigger?

>>1765541
>raspi can draw 3A
Not off the micro USB supply it can't. The output of a buck converter doesn't have a diode like a boost converter does, so if you used a 1S2P instead you could get away with it just fine. Just adding a diode rated for 3A is probably fine, a raspi runs off a 3.3V linear regulator, right?

>>1765677
Give it a scrub with a toothbrush and IPA, then apply flux and reflow all the scabby solder joints. Did water get under the potting compound?

>>1765693
Do the traffic lights want 50/60Hz AC? If so, just use a conventional mains transformer. To turn the transformer on or off you'd probably want to use a relay, but a TRIAC might work fine. If it isn't mains frequency or if you don't have access to mains power, would a square wave work instead of a sine wave? Because if so just use a half-bridge or full-bridge, which should be trivial to toggle if the FETs are driven by the arduino itself. If it needs a strange frequency and NEEDs to be sinusoidal, unironically use a prebuilt Class-D amplifier board connected to some sort of sine oscillator. As for DC, you'll probably want a standard logic-level MOSFET acting as a switch on the high or low side, whichever is more convenient.

A 2N2222 isn't really suited to switching power loads, but they "turn on" when the base-emitter voltage gets above 0.6V, and the collector current will be equal to a factor (~50-200) times the base current, provided the load can supply that much current. This leads to a BJT operating in either the linear or saturation ranges, linear being with the collector current being limited by the base current, and saturation being with the collector current being limited by the load. While MOSFETs simply have a threshold gate-source voltage that they near-instantly start to conduct once the voltage at the gate gets above this; no net current flows through the gate.

>>1765701
Thanks for your response, it makes this a lot more clear for me!

So if I understand correctly, (searching for the 0.6v on the datasheet), the base-emitter saturation voltage is the one to check whether or not you can trigger it using a specific voltage?

>>1765693
absolute maximum ratings are never to be exceeded if you want the thing to keep working
transistors are polarized and not easily useful to switch ac. you should rectify it into dc then switch it
logic-level MOSFETs are superior for micro-driven switching, picrel. IRLx024 (x = your choice of package type) is a decent part for general purpose 5V-controlled low-voltage work. for simplicity you will probably want to put your switch on the low side of the load i.e. + power to + load; - load to MOSFET drain; MOSFET source to - power (and micro ground*); MOSFET gate to your micro's GPIO output
figures of merit that are of interest to you include: Id, Vdss and Vgs(max) are limits you must not exceed. Rds(on) tells you how much resistance the drain-source channel exhibits when the transistor is in the fully on state (often specified at multiple Vgs). the typical transfer characteristics curve (Vgs vs. Id, picrel at bottom for IRLx024) will tell you if your choice of GPIO output voltage can get enough current going through the device to satisfy the load
* if you can't connect grounds of the supplies together, for whatever reason, use an optoisolator to drive the switch, picrel at right. the ubiquitous PC817 is a decent opto for low-speed control signal isolation

>>1765701
>While MOSFETs simply have a threshold gate-source voltage that they near-instantly start to conduct once the voltage at the gate gets above this
almost. MOSFETs do have a "linear" region. Vgs(th) is specified at a particular (relatively small) drain current. it's mostly useful to think of it as the maximum g-s voltage which will keep the channel from conducting *significant* current (as the manufacturer defines significant), above which the transfer function is steep but continuous. e.g. the IRLx024 under consideration here with Vgs(th) rating of 1.0-2.0V is "guaranteed" to not pass "significant" current as long as Vgs < 1.0V but is not "guaranteed" to turn on until Vgs >= 2.0V

>>1765709
More or less, yes, but all BJTs will have a voltage around 0.6V. Maybe 0.3V for a germanium transistor, if you're into that. So you basically never have to check for it. Instead what you'll want to look for is the hFE value, the current gain of the transistor.

>>1765711
>almost
I said near-instantly, not instantly. For the casual user of the common-source configuration, you don't need to worry about the linear region at all. You're much more likely to run into the gate capacitance than the linear region if you're trying to switch with it.
Also I'm grabbing your FET recommendation.

was meaning to get into electronics more so that I can fix stuff but I've nothing that needs fixing.
I guess I'll just have to buy broken shit and hope it's fixable considering there aren't many places to check for electronic waste here

>>1765693
For 12vDC, get a logic level MOSFET with a heat sink. If you can't find logic level, you'll need a driver too.
For the AC you might want a relay instead, transisters can be a bitch around high powered sine waves

>dave makes 30 minute video because he didn't understand reactive power
wonderful

I am still messing with LC measurements. Here is the waveform on the inductor. Why doesnt it decay all the way to zero on the positive cycle? And the negative one does go to zero. Wtf? I am measuring a 1mH ceramic inductor in series with a 33ohm R. It is as if that inductor has a huge internal resistance. Or at least I can get the same result in the simulation if I set the inductor's series resistance to something silly like 10 ohm.

>>1765793
With what circuit?

>>1765797
This one >>1764996 But without the comparator, just checking the waveforms.

But I just measured its DC resistance it is actually almost same as the resistor lol. That explains it. It is about 20ohm and R=33 ohm. So it'd need to make R higher, but not much higher or I would need to increase the frequency. And I am using a power mosfet with a very low Rdson so it probably has a huge capacitance. It distorts the square pulse quite a bit unless the current is high enough, say 100mA, then the waveform is perfect. So R should be low enough. Damn everything is mutually exclusive.

>>1765799
I think it might be the diode, try swapping it to just being across the inductor and see if that makes a difference

>>1765800
i tried without the diode, it makes no difference, the mosfet has its own parasitic diode, it is slow as fuck so there is quite a bit of ringing but that doesn't change anything. I am pretty sure the ceramic inductor's huge DC resistance has something to do with it. But I don't understand why it only happens on the positive pulse. Or it could be the mosfets diode which I can't get rid of. Regardless I think it is probably a wrong mosfet for the job. Too many issues with it. I will try some other nfet tomorrow.

>>1765803
>the mosfet has its own parasitic diode
But that won't do anything unless current can flow back into your PSU from the inductor, which depends on the size of the output caps and the quiescent current of your system. Also you're measuring the voltage on the opposite side to your circuit, maybe that voltage drop is just the natural drop across the resistor, so try swapping the resistor and inductor around.

>>1765701
>>raspi can draw 3A
>Not off the micro USB supply it can't.
Yeah it can, the new raspberry pi 4 uses usb c and can slurp up to 3A under load but usually it tops around 1.8A

>>1765828
>the new raspberry pi 4 uses usb c
Oh I didn't know.
>3A
At 5V? I'm not familiar with the USB 3 standard. Does it have a built-in buck converter?

>>1765833
usbc cable has like 20 wires so it can handle the juice

>>1759651
Any place where I can just find simple circuits with just a handful of active components and an explaiination?
the related links are all gay and I don't care about microcontrollers.
I just want to learn by doing so to speak

>>1765902
I am also about to read the art of electronics and stuff I just also want to do something practical as I have all the equipment and components to do so

new thread
>>>1765922

>>1765902

read first post: >Mims III, Getting Started in Electronics

actually Mims did a whole series of booklets. search for a torrent that contains more than a dozen.

>>1766034
thanks!

Hello lads - can anyone reccomend a quick and cheap PCB manufacturer in the US? I made my first board a couple of weeks ago, ordered it through SeeedPCB - well it's been 3 weeks and they haven't even manufactured the board yet, let alone shipped it. I hate chinks. Anything local that's faster? Since I'm new to this I'd prefer to have the board done quickly as to learn from my mistakes and make improvements.

>>1766226
probably because all of china is in lockdown at the moment
i think there was an american pcb manufacturer last i heard, google for it i suppose

>>1766231
Kek this was my thinking as well.
There are a few american ones, but I'm not sure which if any are good -

https://www.eprotos.com/contact.php

this was one I was looking at, but I'm not sure, no reviews on it really.

>>1766226
please understand fren due to corona outbreak your order has been cancelled

>>1766332
fug

can i download datasheets for transistor/IC/diode packages? i want them for thermal data and dimensions, things that dont change from part to part
though die to case thermal resistance could vary i guess