/diy/ - Do It Yourself

>>1743046
Anyone ever used one of those e-paper displays? The ones that barely draw any power?
What did you use it for?

>>1743036
it's an engineering matter, and so there are tradeoffs to consider, not least cost. some uses have clear, well-established choices that you don't have to think about much. e.g. for bypassing a chip you generally use ceramics because they're small, non-inductive, have lots of capacitance per volume, have ESR so low you don't even need to think about it, and fairly cheap up to several µF. for oscillators or some audio purposes, some of the disadvantageous qualities of ceramics that aren't relevant to a bypass application start to show up in your signal. so you start with cheap polyester greencaps and work your way up the price range of specialty dielectrics depending on your capacitance tolerance, anti-inductance, and leakage needs. for bulk capacitance at the end of a longish wire, use electrolytics. they're still cheap and the quite measurable ESR is actually a benefit in keeping long wires from behaving inductively and voltage-spiking your circuit at power-up. those are the main three. for specialized needs there are others but you rarely hear of them

this thread's digits brought to you in part by the obsolete Intersil CA3046 monolithic transistor array, now replaced by the Renesas HFA3046 monolithic transistor array

>>1743066
i personally have like 50 of them in a glass jar, together with a ton of 3 transistor version in a TO5 CAN package

Hey guys,

Do you know what's the role of condensators in circuit like that? protection from over voltage?

also, where can I learn about practical circuit making?

Way too up of my league, what a clusterfuck. If I need to modify or add a single thing after testing I'm fucked.

>>1743093

>also, where can I learn about practical circuit making?

By reading the OP. It also mentions this book which I personally loved as newbie.

>>1743093
What the fuck is a "condensator"? Try circling the part of the circuit you don't understand.

>>1743100
the word for capacitor in a bunch of european language is a lot closer to condensator

>>1743102
Alright, well the function of the electrolytic capacitor in that diagram is to remove ripple from the DC output of the rectifier. It also aids noise rejection. Nothing to do with overvoltage.

>>1743093
they are there to filter the ripple of the power supply.

>>1743066
>>1743087

Is pic related useful for any particular diy circuits, or did they just run out of pins so they decided to just share one and then try and market it as a feature?

>>1743126
that is a diferential pair

>>1743126
On a similar note, I could never figure out why they didn't package op-amps in extended to-92 packages like transistors, instead they left pins that were NC and used up valuable real estate.

>>1743094
Just add extra solder points and bridges to solder.

>>1743093
We usually refer to it as a Capacitor in english, looks like you're trying to englishize the french word.
On where to learn: Youtube/EEVblog fundamentals & go get ''The art of Electronics'' on #Undernet or a torrent...
The 1000uF Cap is more of a large ripple filter, I'm pretty sure there's supposed to be a smaller cap to help the L200 output smoothly.

>>1743189
Maybe because they already had DIP-8/14/16 packaging equipment?
It would have been better for breadboarding though.

>>1743189
you can always start a company to produce them anon

How do I balance theory and practice? I am way obsessed with theory and math behind EE and that gets in the way of my projects. I really want to know the workings of things like S&S and LTI and laplace transforms etc but then I ask myself why do I need it?? Why do people study that in college? How do they use it? It is not like you are ever going to be asked to use the laplace transform to solve DE because you have software for that and even if you had to, you don't need to understand how it actually works. Also
>lots of real life systems are not even LTI
>for example harmonics of a real life musical instrument are actually time variant
At best I can learn this stuff to be able to recognize the lingo and what they are talking about in the articles etc.

but hey I love math can't help it

>>1743224
>you don't need to understand how it actually works
you do though, because you need to recognize when and why you can or can't use certain techniques.
Also laplace and DEs are just the basics, they are building blocks to understand things like control theory, lagrangian mechanics, etc.

>>1743234
Well maybe you do need to know some things, but as an EE I doubt you'd need to know the complex analysis, convergence of improper integrals, etc. let alone delve into Lagrangian mechanics or control theory as an EE?? The math behind the fourier/laplace transforms can get quite complicated pretty quickly depending on the rigor required.

>>1743234
>>1743224
You don't need rigorous math to get very good at electronics. Just circuit laws and knowledge of how the devices behave. Complex numbers and fourier/laplace are a requirement for some comms and power stuff (three phasic power, power electronics).

>>1743266
And quite honestly, if you get a good grip on the circuit side of things, and start getting a lot of experience building things you can get really good. Math is really important when you can only get one shot at doing something/can't fuck up situations. And about LTI systems, they are a good base for most stuff, (as is classical control) and most sane people work with linearized systems or feeling/experience.
Do you need to know all the vibration modes of a guitar string to build a very good pickup, write a guide on how to amplify that into a usable signal? Not likely. You need common sense and some practical experience tho.
>why study that in college
Because college is not only about learning stuff, 80% of it is seeing if you can actually stand college. And desu systems and control get blown out of proportion by shitty gigantic textbooks and fear.
>How do I balance theory and practice?
Build stuff. If you are like me you will fuck up a lot, and while you are troubleshooting your brain is 100% "hmmmmm" and suddenly you know how a transistor works or "hmmmmmm so that is what they mean with bias level/harmonics/modulation/thing I read about"

>>1743266
>>1743273
Yeah I agree with all of that. I just have some sort of an itch, an anxiety about learning as much math as possible cause I hate not understanding something and honestly the way LT/FT are taught in most EE books they just give you the tool blah blah here is how you solve DEs, just convert from time to freq domain bro,
just be yourself, and here are some examples, look, we reduced a complicated DE to a simple algebraic polynomial, it is just a fraction, easy! and who the hell knows how it works and why it works. But then I notice some similarities between LT and FT and phasor analysis, but of course there are important differences that are not always obvious and you need to have a deeper knowledge of the underlying math to understand the differences, for example jw vs a more generic b+jw and why we need a negative power of e in terms of faster convergence, and s-plane to z-plane mapping and all that fun stuff. I think the math behind some of that stuff is more than I can chew. however while trying to get into that I've also learned some important basic stuff, like general functions, unit step, impulse, convolutions etc. I have the Oppenheim book and it is not the easiest reading so I feel some sort of a weird remorse since I feel like I am wasting my time trying to learn things that I won't ever need for any projects I'd ever work on so there is no way to reinforce that knowledge by applying it in practice.

>>1743296
How about doing it backwards? Making projects and learning what you need to build them?
The diference of the laplace transform and the fourier transform is that the fourier is only on the imaginary axis, not on the whole plane.

happy new year /ohm/

>>1743301
>The diference of the laplace transform and the fourier transform is that the fourier is only on the imaginary axis, not on the whole plane.
Yeah, I've read about this in multiple text books but I can't wrap my head around this. When I see an integral like f(x)e^-x dx I can imagine this integral in a cartesian plane as a decay function, i.e. e^-x gradually kills off f(x). But when I see e^-(jw) or e^(b+jw) in case of LT I am lost. I understand e^jwt is rotation on the unit circile whereas e^(b+jw)t can also modulate the magnitude so we are no longer strictly on the unit circle. But this stuff is totally abstract and I don't understand what it has to do with DE and harmonics and EE in general. I understand simple phasors and I can see how differentiating a phasor you get stuff like Xl = jwL. And I know you can arrive at the same result with the LT and get sL, where s can be b+jw or even the same jw as in the case of phasors. Then why the hell do we need an integral transform to arrive at the same result we get with the phasors that only rely on the eulers formula which is pretty straight forward.

>>1743303
happy new year, lads

>>1743344
M8, I did not get most of your specific points (because I'm a bit drunk) but I think you are overcomplicating things a little. Maybe you should play with filters and falstad to get a feel of things idk. About your question
>laplace transform
>classic control (there are new types of control but it's good for 95% of everything)
>filters
>same stuff for digital

>fourier transform/spectral analysis of data
>pretty useful to see how stuff behave
>filters of all sorts
>power electronics stuff
>basically all of comms

These transforms were made to make things SIMPLE, not complicate them. No one wants to solve DEs when analysing stuff. 90% of them don't even have solutions. They are fully reversible so you can get the same results with time domain analysis. (Solve DEs). Believe it or not, but most linear system textbooks are pure hot garbage. I've got better results when I was in that class by using random books on control, because they had a short, to the point and pratical introduction into systems ,transforms etc.

>>1743356

>>1743357
the right part is a intersection of a vertical plane that contains Im, and the function

>>1743358
>>1743357
Just like you can calculate a line integral along a random surface (evaluate the surface at specific poitns and sum them up), you can do it with a transform.

>>1743357
>>1743360
thanks anon, why the hell none of the textbooks visualize this stuff like that

>>1743126
this man died in WWII for you, filthy casual
https://en.wikipedia.org/wiki/Blumlein_pair

>>1743189
japs did make a lot of low-tier analog stuff in SIP packages, especially Sanyo and Matsushita
DIPs became the standard stateside, probably because more compatible with automatic insertion and more backward-compatible with TO-5 type cans

>>1743224
with more practice
>why study that in college
because engineers need to have the tools to do it right the first time at any scale, esp when lives are on the line. also it's a flake filter

>>1743242
if you're just dicking around with fractions of an ampere and a dozen or two volts at frequencies not exceeding several hundred kHz, and don't need to work over more than a couple of decades in any dimension, the math is superfluous to the qualitative relationships

>>1743363

>>1743363
the engineering mindset is about having tools with which to decompose any problem into appropriately sized/shaped, solvable portions. speaking as devil's advocate, the textbooks don't necessarily want you to develop an intuitive sense of the problem so early on, lest it limit your ability to abstract later on

>>1743422
>blumlein
ahh, that's the long tailed pair configuration! I will study Blumlein's achievements, there seem to be many.

>>1743437
he was a good man, cheers
the long-tailed pair is, of course, extremely and broadly useful. it's at the core of every opamp
not very often seen in hobbyist work anymore because the art of hand-matching transistors isn't as widely known as (due to then necessity) it once was, and commercial matched pairs let alone monolithic transistor arrays are still at least an order of magnitude more expensive than individual transistors or even monolithic opamps

>>1743447
We're well into the era of op-amps, and potentially past that and into the era of DSPs, but DSPs are still less accessible to hobbyists so I'll disregard those for now. Op-amps themselves can be used for nearly any voltage-based signal processing under the sun with very good performance, is there any need for matched pairs anymore? Let alone long-tailed pairs.

And I don't want this to be seen as taking a shit on the great man himself or his achievements, just making a contemporary commentary.

>>1743363
guy that answered you here, this basicaly >>1743424
College is being kinda weird for me. (I'm almost out not, 9th semester) because I'm a intuitive learner and had to wade through this bullshit.

>>1743522
I'm almost out now*, basically*

>>1743513
>le fourth turning
no thanks, I just ate
DSPs are plenty accessible to hobbyists. if the math instructions built into the medium-low-tier ARM MCUs aren't enough, FPGAs are also fairly cheap and come in reasonable QFPs
>voltage-based signal processing
but you're not always dealing with voltages
>matched pairs
easy example: precision current mirrors. complementary matched pair application: reducing even-harmonic distortion in a power amplifier
>long-tailed pairs
sometimes the extra gain of an opamp can be a liability e.g. inserts delay or poles you can't tolerate due to the instability they bring, amplifies thermal noise too much

>>1743522
same here, fren, except I skipped college and my senior year of high school entirely. I'd like to FUCK the academy. learned more from the bipolar guy down the street than from school

if i want to power a 2V led from mains do i need to calculate the resistor value for the RMS or for the sine peak voltage?

>>1743549
rms, but make sure you meet absolute max specs at the crest of the wave and all reverse specs

>>1743549
Most leds have small reverse voltages. You might damage them. Add two leds, one in each direction, or a grown up diode in series

>>1743593
you can tilt the pups to emphasize the bass or treble register and on really good ones you can adjust the individual pole pieces
I mean, it's a legitimate stylistic preference to want the top string a little bit cooler than the rest, esp on a lead-type guitar. you should think about replacing the pups with known quantities if you don't happen to like what's there

>>1743534
>>1743522
so how did you intuitively learn things like laplace transform or fourier transform? the only intuition i can think of is the spectrum of the signal in case of FFT. i at least understand what that is and why it is needed. and i can intuitively understand simple phasors for steady state sinusoidal signals. anything beyond that gets too abstract and confusing pretty quickly unless all you are asked to do is plug-n-chug problems by rote memorization based on provided examples, without any intuition. which is a torture for me.

>>1743643
>laplace
>control systems
>fourier transform/frequency domain
>control systems and communication/audio stuff
Basically try building, tearing shit apart or reading on how they work. I mean, if you get how time domain and frequency multiplexing work you should be good I guess.

>>1743191
>Just add extra solder points and bridges to solder.

That would make it look even more orkish. I try to keep it as simple as possible and with as little vias as possible. This wasn't a good example though, but I want to stick to 2 layers because money.

Why are all of the low noise linear regulators smt/smd parts. I need something like the lt3042 which has a breakout board, but I cant find anything else in its category. I need it for adc/dac setup.

>>1743667
because smd=low noise?

Anyone have a KSGER t12 iron with the shitty "V3.02" software on it and know how to updgrade/fix it? Its impossible to calibrate since it won't go over ~380C, even if the bundled PSU is perfectly capable. Reading around the net it seems the v3 software was some failed hacks KSGER did themselves, but nothing about how to fix it. They no longer sell it, so I guess they were aware, but messaging them on Ali does nothing it seems.

>>1743667
in stock

>>1743667
you can design and order a custom breakout board from oshpark if you don't mind waiting 2 weeks. i paid $1.55 total for three of these boards.

>>1743703
If you're having a board made anyhow, you can drop a couple of breakouts along the edge. (with appropriate routing and mouse bites of course) I did that once to get some SOIC-8 breakouts made.

>>1743046
Are there any gps modules that dont have any COCOM limits on them?

>>1743722
ICBMs are against the board rules

>>1743724
I just want to put one on a model rocket kit i bought. Do they exist or not?

>>1743643
my hook was getting deep into audio shit as a kid. learned that a sawtooth wave is even harmonics, a triangle wave is a particular mixture of odd harmonics, a square wave at 50/50 duty cycle is some other mixture, and promptly forgot all the details. now I just kinda know where to look and have a general understanding that sharper changes tend to have higher frequency components

>>1743651
vias are free

>>1743667
ok boomer

>>1743688
it's just an STM32 isn't it?

>>1743729
lern2sdr

I have a problem with this LED in my circuit. Rest of the circuit is working fine.
This LED is being turned on by a latching circuit made out of the two transistors. The signal comes from the output of a 555 timer.
Whenever I disconnect the circuit from the battery the LED naturally turns off. But when I turn it back on the LED goes on again even though there's no signal coming from the 555s output.
The 555 is activated by another unrelated switch called KIPP.
Now what confuses me is that this LED doesn't always turn on when I flip the switch. Sometimes it stays off as it's supposed to. Depends on how rapidly I flip the switch.
I suspect it has something to do with the large bypass capacitor I have in parallel to the supply.
What could cause this LED's "undefined" state and how do I reliably "define" it?

>>1743733
>it's just an STM32 isn't it?
Yup, but the only custom firmware I have seen (PTDreamer) seems hopelessly out of date (2+ years), so not sure that helps. Unless there is some secret repository of dumped KSGER firmware somewhere...

i'm a noob to electronics so excuse the newfaggotry in the post that follows. this is my first time putting together any sort of electronic thing so I'm understandably a bit retarded.

I wired up three 3.7V 18650 batteries in series in the nigger-rigged battery pack pictured. yes, I know you can just buy them but I'm on break from school right now and I leave for a different state tomorrow. the batteries are then connected up to an appropriately specced BMS board as per the diagram on the Ebay listing. On it's own, the output terminals from this BMS read about 12.4V, which tells me I did that part right. When I wire up the second chip, however, which is a small buck converter (https://ebay.to/2QgjSjM)) I read only about 2.2V on the output from the BMS. after a few seconds I hear a high pitched electrical whine sound coming from the buck converter, at which point I disconnect the batteries to avoid anything potentially dangerous/stupid. I wired the positive and negative ends from the BMS to the positive and negative ends on the "IN" side of the buck converter. why's it making that sound and why is the output of the BMS being limited to ~2V? what could I/have I already fucked up with this mistake?

>>1743750

try putting a cap, say 0.1 - 1uF from base of Q1 to ground. this will short the base to ground when power is applied, ensuring Q1 starts off turned off.

>>1743773
>a high pitched electrical whine sound coming from the buck converter

>>1743773

what would Sherlock Holmes do?
process of elimination: disconnect the buck converter and use the BMS with a regular load, like a vibrator, or car lamp, or whatever you got at the appropriate voltage.
if you can pull lots of current just fine, then your converter is dead or you mis-wired it. otherwise your BMS is bad.

you can test the converter on its own, by hooking up to a 12V charger.

>>1743759
Just because it's old doesn't mean it's out of date, if it works it works.

>>1743783
Does the converter actually pull current with nothing connected to the other end? all I really have to test the BMS is LEDs
for the record I have an extra BMS and 9 other buck converters but I don't want to hook them up if something I'm doing is wrong and I'll just end up breaking other parts

>>1743790
>Does the converter actually pull current with nothing connected to the other end?

lil bit.

>all I really have to test the BMS is LEDs

you prob have gadgets that need 12Vdc. router, modem, external hard disk, house phone, answering machine, police scanner, etc.

Local idiot here, what would be the best way to get the trailer I'm living in to be at least partly solar. Right now in connected to the properties electricity hook ups but would like to switch off. Only needs to be portable enough to fit in the back of a pick up preferably cheap and simple enough for my dumb ass to understand. Thanks

>>1743784
>Just because it's old doesn't mean it's out of date, if it works it works.
Yeah of course. But in this case the custom firmware is targeting an ancient version of the board (V1.5), and people trying it on the more recent ones are reporting bricking. And I really don't know enough about STM32 or its toolchain to ever try fiddling with that code...

>>1743797
>get 12v panel
>connect to battery

I hope hackers dont abuse this:

https://topclassactions.com/lawsuit-settlements/open-lawsuit-settlements/nyc-jail-visitor-invasive-search-class-action-settlement/

Deadline to file 4/27/2020

>>1743777
>this will short the base to ground
how? also checked

https://www.youtube.com/watch?v=78zYtjTm2yY

in this video, clive mentions multiple times how this circuit is "modular" as if that's a bad thing. I'm a novice so I don't know why that is. Can someone explain to me why this is an indicator of amateur/poor design?

>>1743832

it's the way caps work. when discharged, they act like a short circuit to DC, sucking in a lot of current. as more electrons congregate on the plates of the cap, fewer electrons wanna join, so the current eventually drops to zero, and it becomes an open circuit.

>>1743871
i can't listen to this dude for longer than a couple minutes. his thoughts are meandering in random directions and he keeps interrupting himself and mumbling stuff and keeps repeating modular err emm look at this thing err umm modular btw look at that let me umm oh now lets try umm so modular and going in circles.

>>1743871
but i think he means that this thing looks like a bread board where a bunch of components are slapped together just for the sake of quick testing. it could have been done in a more compact/integrated way.

How do I find out the charging voltage of a battery with no writing on it?

>>1743891
>How do I find out the charging voltage of a battery with no writing on it?

voltage is easy. just measure it with a voltmeter. for current, you can compare it with known units of the same size.

>>1743871
>modular

means made from modules. like this TV. if one module goes bad, you can just replace it, without replacing all the electronics.

>>1743893
I forgot to mention the battery is completely drained. It measures 0.05V

>>1743894
>I forgot to mention the battery is completely drained. It measures 0.05V

of course it is. so hit it with a safe current level (10.3mA) and watch it go up and up and then more-or-less stabilize at some fixed value.

>>1743879
I see it now thanks for the explanation but I still don't understand how current can even get to the base of Q1 if the circuit is powered up and IC1 doesn't output anything

I invented a circled sine function. What do you guys think? Will it result in a warm sound?
It may look like a really round sine, especially if stretched vertically. Here is how I generated it. Remember the challenge is to generate it as a function of time, NOT in the polar coordinates since I want to implement a soft clipper plugin in the time domain.
So here is my secret. Knowing the polar form:
y=r*sin(theta) and x=r*cos(theta),
simply solve for theta=acos(x/r) and then solve for y.

So this "sine" is just two half circles.
Here is how you generate them:
float circle (float x)
{
float m = fmod(x+M_PI/2, M_PI*2);
if (m < 0)
m += M_PI*2;

if (m > M_PI*2)
{
m -= M_PI*2;
}

float theta;

if (m <= M_PI)
{
theta = acos ((m-M_PI/2)/M_PI*2);
}
else
{
theta = -acos ((-M_PI+m-M_PI/2)/M_PI*2);
}

return sin(theta);

}

Pretty cool, huh?

>>1743924
Now. the question I want to ask YOU my high IQ people, what do you make of the spectrum of this function, see FFT on the right. It looks convex on a *logarithmic* scale. Considering the properties of logarithms, if that spectrum shape was a function, would it be some fractional power, such as the square root?

>>1743920

when you turn on a circuit, it doesnt immediately go to a known state. there are weirdnesses and transients as the voltage rises from zero to Vcc. also caps like C1 may have to charge to reach a steady state condition.

>>1743924
How would the FFT look like if you removed the negative offset of the function?
You could probably graphically approximate that function in excel and get a formula

>>1743560
>rms
Yeah, maybe if you calculate the RMS power dissipated in a shockley diode curve. RMS voltage assumes an ohmic load if used to calculate power, but the current curve won't be sinusoidal at all so this won't work. I'd add at least 50% on top of the calculated resistance if he's coming close to the current limit of the LED for this reason.
He should be slapping a rectifier and filter cap on it anyhow to stop it flickering, or even better, use a capacitive dropper, or maybe a neon.

>>1743688
Oh man, thank god I didn't get that one. Good luck!

Hi /ohm/, I'm trying to fix a mechanical keyboard, which seem to have burned. I was looking around and this is the only part that looks burnt. What is it? And how can I replace it? I read that it might be a thermistor, meant to protect the kb against over voltage.

Sorry for the lousy picture, it's the best I could take.

JunkScooteranon here with a status update
>motor controller relay flickers on and off
>try to get the throttle off, the setscrew is stuck
>my multimeter passes enough current to hot-wire the throttle open
>spin rear wheel with throttle open, nothing happens
>figure out that the front brake switch is an open-on-press circuit, so the motor controller knows if it's disconnected
>reconnect front brake
>stop hearing the throttle relay clicking on and off when I hotwire it
>fug.
So either I burned out the motor controller, I finally drained the batteries, or it's my multimeter.

Can you recharge 12-volt sealed lead-acid batteries with a normal
car battery charger?

>>1743893
>you can just replace it

Well, that's one theory. Of course home electronics "designers" have nothing like that on their mind when throwing it together. It's often cheaper to "farm out" the different boards to the lowest bidder (i.e. have delta make the PSU)

The modules themselves are often one-offed for a particular model or series and nigh well impossible to obtain after the run. I still have a bunch of sony TVs that are waiting for someone to harvest their tcon boards and put them up on e-bay. They're 720p... I've been waiting a while.

In fact, I can't even do a component repair because they don't make the chips anymore and... firmware. For all I know a "time to make money" timer went off and it doesn't want to work anymore.

>>1743534
>extra gain of an opamp is bad

Is the LM386 just a low (voltage?) gain op-amp with beefier drive transistors for higher current output?

>>1743729

Why? If you want to be able to find your rocket, you put a radio beacon on it.

I'm worried that you have the (crazy) idea that you think you'll get a GPS with 1-metere accuracy and hook it up to a cell based transceiver and you will get help me text messages with the exact coordinates.

>>1743797
based, but consumer product selection isn't really our thing. try >>>/out/ or >>>/g/

>>1743996
he has an LED and he wants to add a resistor and connect the lot to the mains. 98%+ of the voltage will be absorbed across the resistor he's selecting. LED Vf is negligible
it is correct that calculating with peak voltage would build in a nice 41.4% margin without having to think about it, but thinking about is good

>>1743925
what, it's a square wave with pretensions

>>1744014
very very approximately, but there are some extra specializations for below-V- operation, some internal feedback, a minimum closed-loop gain, etc.

>>1743998
>Can you recharge 12-volt sealed lead-acid batteries with a normal
>car battery charger

I do, no problem. I trickle charge them though, don't hook it up to a 100A wheeled charger.

Anyone know what kind of circuit board is inside on of these flashlights?
Found my old late 80s toy and would like to fix it

>>1744024
Did you try getting mom or dad to help you change the batteries?

The battery contacts might be damaged or missing from corrosion

>>1744024
take it apart and show internals
old toys are good candidates for circuit-bending

Cool thanks. I'll go bother them.

What's the cheapest way to get components in yurop?
Arrow was fee-free until recently, now they charge 12$ with every purchase under 50$. I rarely order more than 20$ in components at a time.

>>1744100
Europe has RS and/or Farnell, right? Not Mouser or Digikey-tier, but probably sufficient, and certainly comparable to Arrow.
None of them have warehouses down where I live so I'm sticking to alibay for now. Plus there's always LCSC, but I don't know what sort of minimum requirement to get free shipping they've got going on, last I remember Arrow had a pretty nice, low limit, at least better than Digikey's or Mouser's.

>>1743882
personally i can't listen to greatscott. It's just his german accent and there's nothing wrong with that, I'm ESL myself. But I just can't listen to it

>>1743046
>jkgamm041
is this guy dead or am i looking at the wrong youtube page

>>1744123
Really? I'm ESL as well and find his accent very easy to understand and slightly amusing. I actually prefer it to most of the US or UK similar types of Youtubers, but probably because they tend to ramble on all day about simple things for some reason.

>>1744157
i can understand it fine, it's just that he sounds like a german forrest gump and idk why but that bothers me a lot.

>>1744162
I get it. Indian accents bother me this way even though I understand them.

Is it a bad idea to do solder reflowing in my normal owen where i make my food?

>>1744162
https://www.youtube.com/watch?v=wNnOfF1NkxI
best accent

>>1744174
lmao wtf? there's no way he's not doing it on purpose

>>1744174
where the fuck is that accent from?
Also I think the weird thing is the cadences at the ends of sentences.

>>1744173
if you like being poisoned with chemicals, please do, if you really want to not die but still do it, use leaded tin, not the lead free poison.

https://www.solderconnection.com/specsheets/Lead-Free_Solder_Fumes_Increase_Need_for_Fume_Extraction.pdf

Is there any way I can charge this Li-ion battery with a charger designed for 18650's etc without making something bad happen? The battery in question belongs to a video camera and is normally charged via a DC input on the camera itself. The charger I have is a Liitokala Lii-500.

>>1744217
7.2v means it has two cells and you have to charge each cell separately, it will have at minimum 3 contacts on it to do this

>>1744124
he isn't very active lately

>>1744100
if you order chink brands from LCSC you don't need to worry so much about the shipping cost

>>1744208
>press release
fake news

>>1744219
>you have to charge each one separately
false
>at minimum 3 contacts
only a fucking idiot looking to get sued would consider the thermistor terminal optional

>>1744217
if you have to ask, no you can't

>>1744219
It has 3 contacts, but the third is for a thermistor, isn't it?

>>1744223
I know what you mean

>>1744229
You need to charge the cells separately to make sure they are balanced.
You don't HAVE to balance charge them if you want to go full retard but you should. If the battery doesn't provide contacts for each separate sell however then you can't balance charge anyway.
As for the thermistor, yes most lipos have it, but if you limit the charging current you don't need to use it, the thermistor is needed if you want to charge at full tilt like 2C or whatever, but if you do around 0.5C you will be completely fine. But just to be safe you should use lipo charging pouch so you won't end up with a face full of acid on fire

>>1744234
Also i should mention that a lot of batteries contain their own charging circuitry that has temperature and current limiting so all you need to do is hook up the correct range of input voltage

>>1744234
if you're a smart manufacturer you place a self-balancing circuit inside the pack. you've probably already put a lot of smarts inside the pack anyway to frustrate cloners
>don't need it
but you do, because the real world isn't necessarily a nice constant 23°C. charging below freezing is awesome if you like having energy storage devices with you that will internally short out and go catastrophic at some unspecified later date without warning

>>1744100
also you should really keep more lab stock

Bought some shit which turned out to be more diy than I expected. Can anyone explain how I wire the N and L? Assume I'm slightly retarded too

>>1744207
it's a czech accent

>>1744274
hot women, terrible accents

>>1744274
>>1744275
thats more than just the accent in that video. i think that's him overdoing it for whatever reason or maybe he has some speech impediment.

>>1744267
>Assume I'm slightly retarded too

you are, coz the drawing shows an orange and a blue wire, and you only got half of that right.

run a blue wire from (1) to (1) and an orange wire from (2) to (2)

then cut off 2 small sections of wire, and run one from(3) to (3), and another from (4) to (4).

looks like you need a soldering iron to do this unless you have some female spade terminals on hand.

>>1744295
>him overdoing it

This way of talking was popularized by the "Food Wishes" guy. A lot of people copy it (drawing the false conclusion that this way of talking was responsible for his popularity). The "Ibuildit" guy as recently started to talk like this a bit as well, although it's not as pronounced but I definitely noticed the day when he started doing it.

>>1744267
>>1744327
Thanks. Does the coloring matter though? I needed to purchase seperate wires for the pic, so I got a telephone cable containing four wires within which I extracted. Didn't think the color mattered

>>1744332

of course the coloring matters. for purposes of documentation and safety, the drawing should match the reality.

>>1744223
>he isn't very active lately
>last video over 3 years ago
yeah you could say that

>>1744332
If I tell you to connect the blue wire to a certain point, would you use the literal blue wire or the wire you assigned to represent the blue wire in the picture?
Things like this can easily be avoided if you use the same or at least similar wires. If you had only a set of red and black wires, that's another story but still you can easily assign one and remember: orange is red (close enough), blue is black. You correctly assigned orange to orange though.

>>1743729
Google is your friend, there is a wired article on this problem with model rockets.

>>1744358
Nevermind, should have read the article before posting that, it's crap, they got an exemption and special modules built.

>>1744336
>>1744353
Agreed, did it on the first time on a whim just to get it tested asap to see if it worked before doing it better next time. But better to just do it right from the get go though

>>1744267
>Neutral
>Live
I hope you knew this much already.
The colors are to let other people know WTF is going on if they have to deal with your shit. For instance, I can already tell that you aren't in North America, or they would be L=Black N=White G=Green or Green/Yellow.
If blue and orange are important mains wire colors where you live, you can probably take apart a scrap power cable and find those colors inside.

>>1744411
No, I don't have any previous electrical knowledge. I bought a "diy flatbed" printer base. If I knew I had to cut and wire everything together because the seller jewed out on all the non-finished components, I would have never wasted my money on this shit

>>1744446
To be fair connecting 2 wires to their assigned posts is the least DIY you can get while still branding it as DIY.

>>1744448
I bought one from china 5 years ago, which was plugging in three cables and thats it. This one from canada needed 20 wires cut and wired among other things which is ridiculous and not customer friendly at all

>>1744478
I understand. Many people buy DIY kits just to connect those cables or glue/screw some pieces together and then feel a sense of accomplishment - "Look mom, I made this myself!".
You were on the other end of that, you probably wanted something cheaper and now you have to waste your time, negating what money you saved. It was just bad luck.

I'm moving to another continent tomorrow famlam, it should against the law to separate a man from his junk pile..

I'm no brainlet, but everything with electronics feels like I'm staring an insurmountable 200 foot tall cliff of proficiency, and every time I try some new avenue of ascent I hit a hard dead end
god I hope it gets easier and clicks for me

I need regulated 5v from an unregulated 5v. I have a regulator with a drop out of about 300mv. Is there a way to increase the 5v source to 6ish volts so I can regulate it back down to 5v. Without a step up buck converter but Ill use it if there isnt an alternative.

>>1744545
Why do you need to regulate it if it is already exactly 5v? Does it sag under load?

What is the best way to start a career in electronics? I'm thinking of applying to college for an electrical engineering degree, but I'm not sure if the job opportunities would be different than from what I would get from an apprenticeship program. Does anyone have any advice for this? I'm wanting to do something that involves trouble shooting, or making robots. One last question that I doubt will be possible. Is there any way to have a job in electronics that would allow me to live on a ship?
My apologies if this is the wrong place to ask and I'm shitting up the thread.

>>1744061
I'll have some time this weekend to disassemble it. In the meantime i contacted Hasbro to see if they have any info on it. Will post finds soon

>>1744549
you can't go wrong getting an engineering degree. it's a guaranteed comfortable life if you don't have nepotistic connections for a sinecure or the chops for competitive fields like law or medicine. i've been doing paperwork in a windowless office for the last few months though so i can't point you to a career that will let you do those specific things you've mentioned.

>>1744549
Electronics engineer here. I went into it thinking university would be fun times building things in an electronics lab. In reality there was one project course per year and I spent 80% of my time cramming calculus textbooks. Professionally you will spend most of your time in a cubicle using CAD software. The people who get their hands dirty going to sites to fix and installing things are called technicians. You spend two years at a tech school for that.

>>1744545

What is the nature of the unregulated power supply? For example, you could replace the silicon rectifiers with lower voltage drop schottkies, and get another roughly .5 volt out of it.

>>1744599
what fresh hell course did that picture come from

>>1744446
Well you really need to know what Live and Neutral are if you're going to play around with anything involving mains power. The most important point is to not connect Live to a chassis or anything not inside a case, because it's a shock hazard.
That's why the standard colors are so important, because it lets another person know about the danger.

>>1744594
Thanks. Everyone has been telling me how engineering is great. I was hesitant if it was as competitive as medicine. Good to know it isn't.
>>1744599
If you don't mind me asking how much math do you use in your day to day job, and also since you use CAD do you design your own blueprints, or do you go over other peoples designs?
>The people who get their hands dirty going to sites to fix and installing things are called technicians.
Thank you for that. I'll do more research into becoming a technician. I think I'm more interested in getting my hands dirty than doing a lot of math.

>>1744351
I assume he's retired or moved on. dunno. the big breaking changes in hand soldering are mostly behind us

>>1744411
black-white color coding is mostly reserved for in-the-wall in burgerland these days. yanks no longer write the rules, thank Bob

>>1744506
>not identifying a protégé to inherit your treasury
sad

>>1744549
talk to your local electricians union or canvass the local factories for maintenance help. you don't need university for this
some vo-tech schools might have a mechatronics/robotics/industrial technology program
>jobs will magically create themselves out of my fantasies
you have to be 18 and to have never read The Secret to post here

Redpill me on motor controllers, bros.
>If I drop a 350/500/1000 watt controller in a system with a 100 watt motor, will it burn the motor up or cause problems?
And
>Will a controller work with lower voltages (24 volt battery pack with a 48 volt controller)?
I assume not, because under-voltage batteries are a safety hazard and the controller has no way of knowing what batteries you're using.

>>1744647
>read The Secret to post here
What's that?

>>1744647
>canvass the local factories for maintenance help
Thanks for the advice. I'll call them and some vo-tech schools in the morning, along with the university I'm thinking of attending to see what my options are.
>jobs will magically create themselves out of my fantasies
I figured I'd ask about working/living on ships since I know military personnel do that to work on turbines/motors and safety equipment I knew the chances of civilians doing the same was low, but I figured it was worth mentioning because I'd never know if I didn't ask.
>>1744662
The Secret is a self help book that basically says if you wish for something hard enough it will happen. Be it job opportunities, love, certain events, or friendships, if you imagine it happening and believe then eventually it will become reality. The whole thing rides on the belief that a person can change reality purely through thought.

>>1744563
>he's trying to get a service manual from fucking hasbro
my sides are in orbit, lad

>>1744662
>will it burn the motor up or cause problems
If your battery voltage is matched to your motor impedance, it should be fine. I wouldn't 100% trust a chink BLDC driver at its rated power anyhow.
>Will a controller work with lower voltages
Probably, but it might possibly have a 24V linreg running off what it thinks is 48V (only reason I can think is for an op-amp/other analog IC or maybe a low-side MOSFET driver) in which case it won't work.

>>1744604
Signal Analysis II

>>1744691
Reality as you will ever experience it is literally subjective and it IS shaped by how you think, by definition. Reality for you is completely different from mine and completely different from a person with dementia or schizophrenia.
That being said that book is a pile of shit cash grab and the writing is bad.

FUCK
i thought zero crossing triacs meant they turned off at zero crossing but as it turns out all triacs do that and zero crossing triacs also only turn ON during zero crossing which of course makes them completely useless if you want to cut off a part of a sine to make a led dimmer
So i need a random phase triac instead..
fuck... i ordered 20 of them from chinks.. i wish they taught this shit in school instead of useless trash like the powerhouse of the cell

>>1744612
I have a CS&E degree and at first I hated the math and didn't understand why I had to learn it, but it grows on you and you understand the world better.
As an engineer you can design complex systems from scratch to do whatever you may want - but day to day life is usually as other anons described it, in a cubicle doing CAD, software development or something similar.
As a technician you can pretty much design simple stuff and augment that by finding ready made solutions and implementing them. Your day-to-day would look like going to work, getting into a van full of tools & cables, and going to sites where you have to install or fix electronics, for example the projectors in a university, change a fuse or something in a household, install power cables & wall sockets on a new building etc. On a ship they probably need a technician more than an engineer because you will need to fix stuff, not design it en route.

why are some of them white?

>>1744753
I assume it's to let the light bounce about inside the optocoupler and therefore get a slightly higher gain than otherwise. It wouldn't make much difference, but it wouldn't be entirely insignificant.

Also can you use optocouplers as isolated varactors?

>>1744756
that would seem logical, but apparently not

>>1744760
I'm not sure what I'm looking at there, do any of those show gain? I assume gain would be a ratio of collector current to LED current.

>>1744756
CTR = current transfer ratio = gain. marginally poorer and somewhat more stable over temperature in the white packages

>>1744765
what is notable: turn-*off* time in the white packages is much faster with all else held constant

>>1744756
Oh so the white ones are superior to black ones just like in real life? Cool

>>1744765
but, what I think is the reason for it is that the lack of melanin makes the package isolate better esp against surges

>>1744756
>>1744760
Is it possible to use triacs in parallel to increase max current?

>>1744775
it seems really ill-advised unless they're zero-crossing switched. even then I'd be wary of using them on reactive loads where zero voltage and zero current don't necessarily coincide

>>1744527
it's the same as with eg programming. Just start nice and easy and build up over time. You can't expect to be designing oscilloscopes from scratch 1 week in.

what is the smd alternative of packages like DIP8 that is the largest in size? I am unable to solder most of the amoeba smd packages the sell chips in

>>1744819
SOIC8?

https://www.cnx-software.com/2020/01/03/espressif-systems-esp32-now-supports-bluetooth-le-5-0-5-1/

Thots?

>>1744612
Techs can earn a lot of money too. Atleast here in brazil, were sometimes the demand for eletro-mechanical techs is high

>>1744866
Where*

>>1744723
>If your battery voltage is matched to your motor impedance,
I don't understand, is that a function of the controller? Do you mean that if I have an X-volt system, as long as I have a motor that's designed for X volts the wattage doesn't matter?

>>1744880
Your motor has a power rating, a current rating and a voltage rating. You need to match
>controller power rating can be greater than motor rating
>voltage rating needs to be equal. Lower won't damage the motor, higher will cause problems

>>1744819

One of the best tradeoffs was PLCC, you could surface-mount it, reasonably easy to hand solder, and you could get thru-hole/wire wrap sockets for them! Instead of a DIP 8 you could get a PLCC 16 (since it uses all sides of the chip).

>>1744819
maybe it's not you, maybe it's your iron or tip
anyway SOICs for smallish devices up to about 18 pins, then you start looking at SOPs. good datasheets will include package drawings, mediocre datasheets will have links to package drawings. you should be able to get some idea of what's within your range by reading the prints, dimension "e" is almost always the pin pitch. DIP is 2.54mm ofc, SOIC is 1.27mm (customarily), most other modern packages are 1mm or less
another method to consider is hot air and solder paste, works great for finer-pitch stuff like QFPs if you just apply flux, squirt a long bead of solder paste across each side's pads, place the part, and evenly blow hot air on it until melt

>>1744886
can't breadboard 'em without a breakout board
if you don't have the right tool to remove the chip, you might break the socket

>>1744892
> can't breadboard, need carrier board

that's true for virtually every surface mount package, with the added handicap that there's no way to socket any of them — unlike PLCC where you can do both: something there for manufacturers and something for hobbyists.

> don't have tool, something bad might happen

might break soldered-on chips prying them off with a screwdriver if you don't have a soldering iron or hot air tools.

>>1744898
>soldered-on chips

is there such thing as a power op amp for kW applications?

https://www.sparkfun.com/jobs/276
would you?

>>1744929
I'm not smart enough and I dont have an EE degree

>>1744747
>>1744599
Just like software development... They sell you on cool stuff and even call it CS but it should be just called computer programming as there is zero actual science involved in your day to day activities unless you landed one of those rare jobs that is not web development. So I laugh when people on Stackoverflow claim you need to study things like algorithms/numeric methods/discrete math/ etc. Only if you are curious about it yourself.

can you recommend a good modern generic opamp i should stock up on?
I used up all my lm324 and surely in 2020 there is some better standard than some 50 year old tech right?

Where can I get those rubber buttons? Preferably suited for SMD backlighting.

>>1744951
https://www.sparkfun.com/products/7836

>>1744949
http://www.ti.com/lit/ds/symlink/lmv358a.pdf
this maybe? Literally just go on TI's website

>>1744920
what the fuck

>>1744920
>>1744958

There's no TTL 74xx chips with outputs for kW operations either. There is this thing called "power amplifiers" though.

Back in the day, there were tube-based op amps though, I suspect those could be made to drive substantially more power.

>>1744920
There probably is. But you can build your own with a long tailed pair.

Is it possible to make a HTTP(S) server based around FPGA? Will it be any faster, than normal server?
Are there any FPGAs with SATA support?

>>1744986
>Is it possible to make a HTTP(S) server based around FPGA? Will it be any faster, than normal server?
Sure, if you put a CPU core in it, and have it run Linux and Apache, then it will be just as fast as whatever the equivalent speed CPU would be.
>Are there any FPGAs with SATA support?
Probably, but I doubt they have anything more than the electrical and bit-level stuff, so you would need to add more HDL to support block-level protocols.

how do you even wrap your head around these kinds of schematics? Do you just recognize subcircuits and estimate how they work together?

>>1745017
I am typically lost too when I see unfamiliar complicated schematics since it just seems like a spaghetti of connections. But the key is to recognize familiar building blocks. In this case, look from left to right and google: differential amplifier, the long tailed pair and how you bias it, current mirrors. Pretty standard stuff (this is probably a simplified schematics, not even kidding), push-pull output, feedback (it is external in this case but typically you'd see 6 connected to 2 via some sort of a resistor network). And should be able to recognize the stages. Looks like 4 stages here. (Probably)

A good buddy of mine has an 808 bluetooth speaker. Some drunk faggot busted the female mini usb port. I tore it down and tripled the battery life. I also took the bullshit charge port off of it. But that was 2 years ago. I movedand lost it and just now found it again. All of my electronics shit is locked up 4 hours away but i have a soldering gun and a shit 10 foot usb cable. I know the bare wire is ground. Should i also solder the shielding with it? And i have green, white, and red. Pretty sure red is positive and the other 2 are data.

So, do i solder the shielding in with and also what wire goes where?

>>1745088

I have kids, and I've fixed many a USB and headphone cable. There may be a standard on the colors, but many don't bother to follow it.

>>1745017
It says right in the file name what it is

>>1745017

The lowest red box is the classic widlar current source with a classical 5kΩ value.

>>1745088
Google the pinout. And you should have a black wire too for ground/0V.

>>1744986

Yeah, you could. The first thing you'd need to do is implement a TCP/IP stack in an FPGA. After that the web server part is trivial. Are you going to embody the content in the FPGA as well?... anyway you see where that is going, the web server using traditional CPUs is not the bottleneck.

A couple of large storage vendors (i.e. EMC) went to battle and implemented the RAID parity calculations using ASICs and FPGAs (basically large vectored XOR operations) but over time dropped all that and just started using ordinary processors since the cost/benefit didn't pan out.

>>1744920
sure. you put a linear power stage on the output of your favorite opamp, take the feedback from the ultimate output. lots of examples in datasheets. cool your final trannies well
or you scale up a class-D driver/controller with some big fuckin MOSFETs

>>1744929
>marcom
no

>>1744949
nah, LM324 is pretty much still the state of the art for low-power nothing-special opamps
if your V+ - V- were always less than 5V, you have a shitload of options and I'd recommend the MCP6001 as bretty gud

>>1744986
>HTTP(S) server built around FPGAs
sure but it would be lame. that said it was far from uncommon to see HTTPS accelerators in server farms before server CPUs got crypto accelerator instructions
>FPGAs with SATA
many of them. it's just a two-lane differential point-to-point connection. you only need a fast serializer/deserializer, a couple of caps, and a good board

>>1745017
basically, yes

>>1745088
in USB and other digital protocols, ground the shield only on the upstream side. use the real GND lead for the circuit

>>1745000
>apache
>not even 2019 anymore

>>1744943
To be fair I worked as a software developer during my last year in uni to make some extra cash and while I can certainly say I'd have made it without knowledge of discrete math/algos, it helped tremendously and is probably why I was valued by the people above. Apart from not making things unnecessarily complex, I could work on problems way faster and on one occasion I actually wrote a proof (not because I'm retarded, it was a very complex algo and we needed it to work every time without fail) that an algorithm I wrote would always find a solution if one existed - after about a month there were so many changes that the proof needs to be redone but that's beside the point.
I genuinely hate people calling software development curriculums "CS", after all the fucking work and research I had to put in during my thesis (in machine learning).

>>1744957
lm358 is the same gilf as lm324 the only difference is tha 358 is 2ch and 324 is 4ch

If i want to reliably filter out anything non bass from an audio wave (bass is up to 300hz) can i do that with a single stage pass filter or do i need at least 2 stages possibly even more? The filtering doesn't have to be perfect, i simply want to know if the audio wave contains bass or not

>>1745241
that's better determined experimentally, along with the optimal corner frequency
I think a second-order Sallen-Key filter (12dB/octave) would be enough for a first go, and easy to tune once built

>>1745241
depends how sharp you want the filtering to be

am i stupid or is the transformer on the left wired wrong??

>>1745252
that's a common mode choke

>>1745249
>>1745247
i want to pipe the output into an opamp and have it output high if there is base present or low if there isn't, and i will use a voltage divider to set a voltage cutoff for the opamp so it will be adjustable, so literally just has to be good enough to tell if yes there is base present or no there isn't

it will look like pic related because i like the simplicity

>>1745254
i did more reading and looks like the sallen key filter suggested is way better, and it's actually very simple as well as it turns out, basically you just change a single wire
i will give it a shot

Is SW1 with R1 sufficient to bypass the flipflop? I want a circuit where SW2 can be switched from momentary to latching.

>>1745253
Like he said, L2 and L1 are not transformers. They are filters. T1 is the only transformer in that schematic.

>>1745258
Sallen Key filters are pretty good, they're probably the workhorse of any nth order active filter made of discrete components.

>>1745263
Why 1k? That's like 150mW gone to waste. Also five points:
-You'll want a pulldown resistor after SW2, otherwise it U1A's CLK input could be floating.
-You'll be shorting U1A's Q output to Vcc if both switches are on at the same time, this could be fixed by replacing R2 with two resistors, one on U1A's Q, one on SW1's pin 2.
-I'm not sure if 4000 series can handle 12V, so double check that.
An optocoupler isn't a good choice of a device to switch a load with.
-Why not just use a T or JK flip-flop?

And I'm pretty sure there are better ways of bypassing the flip-flop, like using resistor-transistor logic. Do you even need a bypass button?

>>1745287
Is it possible to chain output from low pass into high pass sallen to get band pass sallen filter?

>>1745289
Yes, but if you just want a first-order band-pass you can use an IGMF (infinite gain multiple feedback?) arrangement instead. Cascading these may be better than cascading alternating high/low sallen-keys, I'm not sure. For a band-stop, a twin-T notch filter is probably the best way to go.

>>1745287
>Why 1k? That's like 150mW gone to waste.
Arbitrary number. Just thought there should be a resistor there so current will go through the switch instead.
>You'll want a pulldown resistor after SW2, otherwise it U1A's CLK input could be floating.
Pulldowns were omitted for clarity of the example.
>You'll be shorting U1A's Q output to Vcc if both switches are on at the same time, this could be fixed by replacing R2 with two resistors, one on U1A's Q, one on SW1's pin 2.
Both switched can be on at the same time. I didn't think this mattered but I will fix it.
>I'm not sure if 4000 series can handle 12V, so double check that.
3-18V says the datasheet.
>An optocoupler isn't a good choice of a device to switch a load with.
Opto was just thown in there as an example. I think I want to use ULN2804. There will me multiples of this circuit and loads will be small automotive style relays.
>Why not just use a T or JK flip-flop?
No real reason.
>And I'm pretty sure there are better ways of bypassing the flip-flop, like using resistor-transistor logic. Do you even need a bypass button?
I need the switch. The idea is to configure the board for the individual circuits to be either toggle or momentary before power on.

>>1745263
>>1745305
>momentary
>switch connected directly to edge-triggered clock
It's gonna be real momentary when the bounces hit it.

bros i have an interview with abb for motor controls in a couple of days and i don't know shit...

>>1745320
Depending on the position you might be fine. Lots of people I deal with every day don't know shit. If you can learn shit then they might hire you. If you have absolutely no qualifications and absolutely no peripheral knowledge then you'
re probably screwed. Good luck.

>>1745287
Okay, new schematic. The ULN2804A has a 10.5k internal resistor for a 6-15V CMOS input. So driving it direct with the 4013 is okay, yes?
I assume I could drive it directly with my 12V source as well but if I need a resistor between U1-Q and SW1-2, and assuming a 1k resistor, the resulting 1/10.5 voltage divider won't matter for the ULN2804 input?
SW1 is a DIP switch. SW2 is a momentary button for a human to push.
>>1745312
I added a capacitor for you.

>>1745322
idk i always fuck up the technical questions. I interviewed with Toyota for what i thought was again electric powertrains but then the guy started grilling me on hybrid vs diesel emissions. I just gave some generic answer about stoichiometric ratios resulting in more or less NOx, trying to reason through it, but it was obvious that it was going terribly. Got the rejection email shortly after.

Everybody kept saying that for your first job people don't expect much at all but i'm having a really hard time.

>>1745332
Don't pretend you know things you don't. They're not asking you because they don't know.
Instead of guessing tell them you don't know but are interested to find out, or explain how you would begin figuring it out.
People get hired into entry level positions for their ability to learn, trustworthiness, and simply not being a total weirdo. They do not expect you to have intimate knowledge of their product or market.

>>1745332
dude you have no idea. when i got out of college, i had a bunch of really weird interviews, one of them was an online running time test with a bunch of math/logic questions. they were not super hard, but you have like 2 min per question and the time is ticking right in front of your eyes and you can't go back or skip questions. it later turned out it was their weed out strategy since they already had a bunch of entry level positions filled and were probably thinking if we find one out of a dozen who can do this test we may hire him. needless to say i didn't get that job. maybe someone else did. but fuck this shit. if i were to do it again, i'd try my luck elsewhere instead of practicing doing math equations in real time. wtf

>>1745332

Volkswagen knows all about emissions. Maybe that's what he was getting at.

>>1745332
Hey friend, my advice is to be confident and don't look like you're uncomfortable (it makes it seem like you don't believe you can do the job. It's fine if you don't, but don't let them sense it).
Also, try to make it seem like you don't really need the job but you'd be interested in working with them (key here is "with them", as opposed to "for them").
If you show passion and belief in yourself people just hire you.

I'm looking to buy a kit with a few general mosfets just to have on hand. Does this seem like a good assortment for $7?

>>1745428
those are regulators, and most of them will collect dust
just get some xxx30N06 MOSFETs, they should be able to turn on nicely at 5V logic levels with a good low Rds(on). also maybe some xxx30P06 just in case you want to switch high-side

Here's the model lighthouse I made last summer. When placed by a window, an LDR senses the light level and starts flashing when it gets dark enough. It's built with 4093 logic IC, LM317 lamp driver + BD136 switch transistor.

>>1745523
Cute. Good job.

>>1745523
nice

>>1745428
Those are all linear regulators. Blame the chinks.

>>1745621
There is a transistor in those regulators somewhere...

>all electronics stores on ali ship with cianao or yanwen by default
>costs more than the product to get China Post Registered Air Mail
Can I reduce shipping costs by ordering from multiple listings from the same seller? IIRC they shove them in the same package if possible, but I haven't seen the shipping costs decrease at checkout.

>>1745473
>most of them will collect dust
I tend to use more 3.3 v regulators nowadays more than anything else.

>>1745523

a flashing light just means your mom's bedroom is open for clients.
if you wanna give the impression of a lighthouse, it should rotate, just like a text should rhyme to suggest that it's emulating poetry.

>>1745627
yes, sometimes the system does it automatically, sometimes you might have to build a cart and ask the seller to reprice it. there's a lot of chink merchants who front-end for one or more drop-shippers. those probably can't be swayed

>>1745726
>babby poetry
sure
kid
he said
colon dee
t.(c) spurdo 2020

>>1745861
>drop-shippers
All this shit's shipping from china, so I don't think that applies. Not in US or Europe or Russia.
I'll build up a cart with Fantasy Electronics or one of those guys and see what I get.

Still haven’t used it

>>1745258
So to make the sallen filter work i realized the opamp needs to have a negative voltage supply since it doesn't work if the negative opamp rail is GND since it can't provide negative feedback back into the filter

I have this thing in picture, so if i understand this correctly i feed 5V in and get 2.5V (pin 8) and -2.5V (pin 5) out with gnd being the common center.
But there is a problem, lm324 can go down to gnd rail but not to + rail, so i will probably have to add an additional boost converter to provide the IC in the pic with say 10V so the opamp has big enough working range of -+5V as i assume it is not possible to connect the 5V going into the pin 1 to the + rail of the opamp and then connect the negative voltage on pin 5 to the opamp - rail, is that correct?

>>1745882
Yes, many op-amp circuits require a split-supply of some kind. While there are dedicated ICs for this, if you're just running with what you've got, either of these can be good circuits. There should probably be a cap or two on the second voltage divider. I usually pick the right-most one if I've got a spare op-amp. You can add or remove caps where you see fit, I think you technically only need one cap across a voltage divider, but I haven't rigorously tested it.

In essence, you want the 2.5V to be buffered by either the op-amp or the capacitors without carrying any power rail ripple through either the op-amp or the resistive divider.

>>1745865
it definitely does happen. I've had about 5 orders out of 100 come back with the response similar to "These items are being shipped from separate warehouses, don't mind the delay or fake tracking number". can't remember if Fantasy is one of them

>>1745882
just make up a virtual ground and ac-couple the signal with a cap. you don't need bass down to dc, 20Hz should be plenty

>>1745892
you really should place one cap across each resistor unless you can guarantee that the supply is stable, solid, and physically nearby

>>1745882
Wasn't there something about simplicity? Single supply design: You can use your Sallen & Key filter and replace RB with RB1 and RB2 in series (as a divider) such that RB1=RB2=2RB. Now the output has 2.5V for a 5V supply.

>>1745882
naw, just make your signal smaller. there's no fucking reason to try to push your signal all the way to the rails, geez

>>1745902
Yes there is a third method, replace any resistor to ground with two resistors, one to 5V and one to 0V. Shorts to ground should be replaced with the two resistors but also with capacitors in parallel with the resistors. The new resistors should be twice the original resistor values.

This isn't necessarily simpler, unless you really don't want to spare an extra op-amp or only have 1 or 2 "ground references" to make.

>>1745258
So i tried to build it, i fed in a 400mv peak clean sine (which is amplitude devices like phones output at max sound volume) about 250hz and the filter is set to be roughly at 300hz
I tried to power the opamp from +5V and GND to see if i really needed the negative power supply for the opamp, and it looks like it actually works, which is pretty cool. (yellow is input wave red is opamp output), when i increase the frequency the amplitude of the output wave shrinks
The only issue is that the red wave is deformed for some reason, but as i only need to detect a certain frequency and not actually process it as an audio it's fine
I wonder if the opamp had say a -5V rail available if the red sine would be nice and clean

>>1745905
You know there's over 9000 free schematic tools available?

>>1745963
this is 227hz vs 1khz
yeah i think it's good enough, i will try to hook it up to my leds and play some base to see how well it works

>>1745963
you need to bias your stuff or add a negative voltage. Really this is getting so common here it should go in the OP.

>>1745969
I know the wave would look better and originally i wanted to do that, but then i found out it actually works well enough without the extra complications that the negative voltage would bring
this is an actual song and the high frequency shit is beautifully filtered out

>>1745971
why do you filter out high freqs? do you have
a hiss in your source? wouldn't it sound bad since you are filtering out percussion basically?

>>1745978
the output is not intended for speakers, it will be sampled by a micro to see how much base is present in the music and then it will use that ti turn on and off some lights

>>1745988
ah, cool, sounds fun

>>1745971
>
I know the wave would look better and originally i wanted to do that, but then i found out it actually works well enough without the extra complications that the negative voltage would bring
You can BIAS the opamp inputs. Literally TWO resistors. And that level of distortion>>1745968
>>1745963
Is not ok

>>1746036
Since all you are doing is looking for presence of bass, I guess that would make the other polarity unnecessary, and the distortion from ignoring it irrelevant.

>>1746036
there. i did it. negative rail for the op amp.
happy?
and now LEAVE ME ALONE

>>1746061
it did one annoying thing tho, when the freqeuncy got really high that left hump started growing instead of shrinking, it wasn't too bad, still suable, but a bit meh, can't really ship it into mass production like that

>>1745971
you need to make sure that the inputs of the opamp don't go below -300mV or it can blow. "works" is irrelevant if the thing will *stop* working forever as soon as you accidentally overload the input

>>1746074
Maybe he could put a diode or full-wave rectifier on the input? I'm sure a full wave would add some harmonics, but he dgaf about high frequencies anyhow.

>>1746074
>opamp don't go below -300mV
oh shit, good point, i didn't notice lm324 has input limits this small, but that's fine, at absolute top volume consumer devices i am using produce audio waves with 400mv amplitudes and it didn't blow the op amp, but i am simply going to reduce the volume to keep it at 250mv and be well within the safe margins while everything will still work as intended

>>1746079
i can't do that, the filter won't work correctly if the opamp isn't getting the full sine

>>1746082
>>1746074
wait i just realized, they mean -0,3V below ground rail, and if the ground pin of the opamp is -5V then that would be below -0.53V so i should have be fine

>>1746083
If you're not using a dual supply, the "ground pin" of the op-amp is at 0V, not -5V. The input from an audio signal would be (probably) at 1V peak-to-peak, which means it goes from -1V to +1V. The input will be negative relative to the -V rail of the op-amp, which is less than 0V and means trouble.
If you use the two resistors trick to connect the "ground" of the input relative to Vcc/2, then there should be no problem.

>>1746096
I am using a chip which does some black magic with two caps and gives me -5V,
originally i wanted a single supply but it is working much better with the dual

>>1745473
>>1745621
Damn im an idiot, I almost got tricked. Thanks

>>1746083
oh, yeah, now you're fine if you give it some legroom with the the 7660 or something like that

>>1746104
>I almost got tricked.
How? Lm317 is the most famous linear regulator on the planet, everyone knows it's not a mosfet
plus it costs more than a fet anyway

>>1746108
but to have to buy a bunch of fixed regs alongside that? sad
>costs more than
but what if you wanted the FET?

>>1746099
Then you're just fine, that's the right way to do it.

My DAC outputs a stereo signal but I have a pair of active speakers and also a powered sub.

How can I split the stereo signal to both speakers and sub? Do I need a receiver with 4 outputs and rewire the speakers and sub to make them passive?

>>1746215
the powered sub probably has a crossover and high-impedance input. probably just connect the sub's inputs in parallel with your other speakers

>>1746222

Yes it has that input but when I connect it that way the speakers sound in mono. The sub was a part of a sony sub + speaker system, I don't know what magic went into the speaker circuitry.

If I plainly connect the sub speaker to the amp with just wires then it acts as a normal speaker, losing all the low end frequencies.

>>1746227
look up the manual online by part number, I bet there are connection diagrams

>>1746066
Now all you have to do is correct the phase

>>1746250
>phase
it doesn't fucking matter. did you read the application brief or are you just trying to make work for other people?

>>1746134

probably won't need the mosfets, you can substitute a voltage regulator instead (in some cases) with a few minor design changes:

Audio amplifier using LM317: https://www.youtube.com/watch?v=FxHu5c4r89I

>>1746255
what's the Rds(on)? you don't know? then why the fuck are you shitposting?

>>1746250
how do you do phase correction anyhow?

>>1746250
retard

>>1746252
>>1746265
>>1746270

If you're looking for the difference in amplitude (in order to detect the filter effect) it helps to have them in phase. Both a difference amplifier and differentiator (to get them back in phase) are quintessential op-amp applications, if you're using an LM324, you still have one to spare.

JunkScooter final update
>spend hours trying to diagnose the controller
>think the throttle or controller is fucked
>prepared to drop $30 on an aftermarket throttle and controller
>finally get around to charging the batteries
>hook everything back up
>everything works perfectly
>mfw I don't know whether to be pissed or relieved

>>1746333
Sometimes taking it apart and putting it back together inadvertently fixes it. Happened with a television of mine once.

>>1746290
>a difference amplifier and differentiator
I don't understand how this would work. A differentiator has a constant phase shift of 90° (±, cant' remember which) above its corner frequency and has a gain that increases with frequency after the corner frequency. And a difference (differential?) amplifier has no phase characteristic at all.
Do you mean I should be shoving one of these in the main filter's feedback loop?

Say you've got an inverted notch filter (0dB everywhere except around the peak, where it goes up to 20dB) and the phase diagram looks like a differentiation of that (above 0° on the left and below 0° on the right, approaching 0° at either extreme). Like a wah-wah effect. What would I add to get a similar gain-frequency profile but with a phase diagram that's 0° everywhere?

ok what the fuck is this circuit, i've tried it with all sorts of different frequency and amplitude sine waves, with different waves at each input
i thought i was making a mixer or summer or subtractor or whatever, but none of those fit the bill, so i thought i had an extra degree of freedom somewhere
then i tried to calculate the transfer function but i got
>Va + Vb = 0
so it turns out i've got one less degree of freedom, do i need to do the calculation assuming non-infinite input resistance?

>>1746351

If you wanted to bolt an active node onto the output of something that was out of phase, a differentiator is one way you could do it, it looked 90º out-of-phase to me. You could also change the phase of the unfiltered input waveform. What matters is that they match when you feed those into a difference node which will subtract one from the other. The last op-amp in your LM324 can be used as a comparator to detect the presence/absence of whatever you are looking for. You could place a capacitor (or RC) in the input to the comparator to "hold" (as in sample-and-hold) the detect threshold. At that point, you might not need the µC.

This kind of reminds me of those 2D VU meter applications where it will give you the relative power across the spectrum, except you're looking at the lowest frequency band, and only one led high. Anyone remember chaining lm311s together to make a VU meter?

All this is to avoid a lot of complicated stuff on the µ controller side (obviously you could store a running window of values in memory, figure out how far back you need to look to compare, etc.

Not trying to make this complicated, it's all been done before.

>>1746396
Yes a differentiator has a phase shift, but it also has a gain that varies as a function of frequency. I want to isolate these features, as slapping a differentiator after or before my filter is equivalent to an additional high-pass filter on my signal.

>>1746406

That's true, so instead just use an allpass "filter" on the input waveform and the phase will be the same when you do the ∆

>>1746380
looks like the V3 sine wave is being summed with the inverted V4, and vice versa. The outputs should quickly die down to zero.
Only noise should be amplified, I'm guessing.

>>1746443
>The outputs should quickly die down to zero
Doesn't happen
>Only noise should be amplified, I'm guessing
No, it's a regular pattern at a frequency consistent with one or both of the input waves, though at a far higher amplitude (hence the /100). The output of one of the summers is equal to the inverse of the output of the other summer. The whole thing is burning 4.5mA too, but that's about normal for these shitty op-amps.

>>1746447
This is the closest I've got to it, the V(q) is an output of the wacky circuit, V(diff) is the result of feeding the difference two inputs into a differentiator. I'll try swapping the op-amps and seeing if I get a different result. Then I might build one, but that would require busting out the oscope and fgen.

do diodes work with tiny currents? i want to protect an arduino digital pin from reverse voltage since i will be feeding ac into that pin, so i want to slap a diode there to cut off any negative voltage.
but since the pin impedance is very high like 100mohms or something, will the diode still work as intended or do i need to connect a resistor between diode output and ground to create an artificial current drain for say 2ma to make sure the diode works as intended?
it's an ordinary cheapo chink diode

>>1746509
>feeding ac into that pin
What the fuck

>>1746511
do you not know what a diode does?

>>1743046
Hi everyone, I'd like to get started with electronics, and someone already recommended me the book Make: Electronics 2nd ed.
I need some components to get started, and I found this kit.
https://www.elegoo.com/product/elegoo-upgraded-electronics-fun-kit/
I found it for 15 euros, is it a good choice?
Thank you.

>>1746527
>another, different style of breadboard psu module
Fuck, I guess I got the one shitty one, because it overheats really badly even with no load.
I hope that somewhere out there there's a split-rail breadboard psu module for audio circuits.
Note that that particular one doesn't have a USB input, just a USB output, which some of them have.

It's not that bad for the price if it ships reasonably quickly, though not exactly stellar either. But you're not going to be doing much with an optocoupler and a 74HC595 latching SIPO shift register and all those male-female jumpers. And I've no clue what you're supposed to be doing with those pin headers. Those listed parts are typically arduino things, which the box doesn't come with. Not saying it should though, we've got enough hobbyists with arduinos on our hands. I'd prefer to see some op-amps and comparators, complimentary BJTs and FETs, and perhaps an inductor or two on there but that's just me.
Also that breadboard is a bit shitty because at link 30 there's no continuity between the rails, I'd advise getting one with full continuity through there. Some smaller breadboards for smaller projects is also nice.

I'd consider perusing aliexpress or ebay for cheaper and/or better packs, or even making your own of multiple orders. If you have the luxury to without suffering from long shipping times or high markups, buying parts as you need them would likely prove more convenient.

>>1746527
Doesn't the Platt have a shopping list? I see no tools and no meter.

Holy shit guys:
6-pin optocoupler (acting in photovoltaic mode as a photodiode via bc junction, e.g. 4N25) + JFET to make a variable resistor.
Like a more consistent (and more expensive) LED + LDR + heatshrink, assuming ~1V is enough to bias a JFET significantly. Could always stack the fuckers in series, but that's getting a little ugly.
Not sure if you can get photodiode-based optos.

>>1746527
Isn't that the one with the common-collector transistor for no reason and other transgressions against reality? We recently removed it from the OP for that reason.

>>1746538
Shit, never mind, someone beat me to it:
http://pdf.datasheetcatalog.com/datasheet/fairchild/H11F1.pdf
Not that cheap, but neither are JFETs in the first place. I'll be sure to grab me some of these, because OTAs and mixers are for shitlords.

>>1746539
H11F3 is cheaper still

>>1746527
As some other anon said, Make: electronics has shopping lists in every 5 chapters or experiments. I suggest getting multiple lists' components/tools at a time to avoid shipping costs and waiting.
Also I'd unironically advise you to get an arduino kit as a beginner since you can do fun stuff with it right away.

>>1746509
just add resistors to pull it up ffs

>>1746538
yes, that's the very one. disrecommended by /ohm/

>>1746542
>suggest getting multiple lists' components/tools at a time to avoid shipping costs and waiting.
Well, that's the idea. I realize that that kit doesn't contain everything I need, but I wanted to get as much as possible in a single kit, and buy as few parts as possible separately.
>>1746538
>Isn't that the one with the common-collector transistor for no reason and other transgressions against reality? We recently removed it from the OP for that reason.
I don't know what you're talking about, but I had already confirmed that the transistors in the kit are the ones recommended in the book:
>Part numbers 2N2222 or PN2222 or PN2222A are okay. PN2222 has become a more common designation than 2N2222, but either will work.
(the kit has PN2222)
>>1746533
Right, I forgot to say that I already have the tools and a (basic) digital multimeter.

Basically what I would need to buy separately are test leads, battery connectors, relays and fuses.
I have no idea about that potentiometer included, it looks different from the one pictured in the book but I couldn't find a better picture of the one in the kit.
I'm starting to think that I should just go to a physical electronics store and buy exactly what I need, without wasting time and money with separate shipments.

>>1746547
>I'm starting to think that I should just go to a physical electronics store
People here always seem to order the cheapest components from China they can find and then complain a lot about it. A physical electronics store if you have access to one is always going to be a way more pleasant experience.

>get a broken ceiling led light
>look at it
>it had some transformer and shit to power it which i don't have
>the light itself is 20 leds in series
>*big think*
>assuming the leds are probably 12v then 20*12 is 240V which is what local mains is, therefore i don't need the transformer, which was probably for isolation purposes only and i can use a full bridge rectifier instead to power it directly
>plug the light into the rectifier, stick in in the outlet and turn it on
>most of the leds instantly turn black
FUCK
no i don't understand this was supposed to work, how could this happen to me?
the math checked out

>>1746550
LCSC strikes the right balance between low cost and genuine components

>>1746570
>thinking instead of reading and testing
stick to pottery

I'm confused over circuit outputs and how they can be used.
From what I can ascertain, the output of a circuit can be either a voltage OR a current. I don't know what you should actually expect, though, for when designing your own circuits.

For example, let's say I want to use an OTA in/as a VCA.
I'm not trying to ask for implementation-specific help for an OTA, this is just the example I came up with to explain my thought process and confusion. Well, they output a current, not a voltage. Does that mean I _can_ drive a speaker using just a current? Do I need to convert that current back to a voltage? If I do, how? (I know you can use a resistor but my intuition tells me that's the retard's way of doing it.) If I convert it back to a voltage, why not just use an Op-amp in the first place?

How am I supposed to know what type of output is or isn't correct for the next (or final) stage of the signal chain?

>>1746627
voltage is benis, current is cum
How will you cum into vagene with no benis?

I recently got a HiFi amp from the 80s for cheap, but when listening at low volumes (with headphones), the sound keeps cutting out on one channel. It can be brought back by turning up the volume or even just the bass EQ.
After some googling I came to the conclusion that the fault may be the relay, so I went to measure it.
I applied 20V (it says 24VDC on the top, but I don't have a power supply laying around that gives 24V) and the relay clicked. If I understood things correctly this closes 2 switches inside the relay, so measured the resistances.one of them was 8 Ohms and the other one was 70 Ohms.
Can this be the cause of the problem or should I look elsewhere? And if it were, how would I fix it? the relay says "TYPE AP2U 24VDC" and "DEC japan" on it. I had a look online but couldn't really find anything about it.

TL;DR: Two sets of contacts inside a relay measure at 8 and 70 ohms respectively. Does this mean there is something wrong with it? How can i fix it?

>>1746636
my bull will do it for me

>>1746636
>>1746639
memes aside... Adding another stage to the chain inherently introduces some form of resistance, converting some of that current into a voltage?
Up until now, I had been modelling it with the output just tied to ground like in the pic, which technically made there be no voltage. But...

>>1746645
...replacing the short to ground with a model of a loudspeaker, it seems there's an oscillating voltage.
Which makes me wonder, what is driving what? Is it the current that drives the speaker (they're an inductive load), or is it the voltage making the current? It always feels like a chicken-and-egg problem with this shit

Is there a way to easily work on a circuit (or project) from two computers using Eagle?

>>1746659
git

>>1746627
If you have something that has a voltage output it means that if you want to get usable information or power you should design whatever it connects to it based on that voltage value. The current itself doesn't matter. (Within the power rating of the device ofc). The oposite is true for current sources/sinks.

>>1746674
You have to keep in mind the basic laws:
>V=R*I
>P=V*I
>https://www.youtube.com/watch?v=DovunOxlY1k

Someone ITT suggested I look at Ott's Electromagnetic Compatibility Engineering for information about EMF and the like. That was very helpful.

I'm now wondering if there's any standard reference or recommended references for ruggedizing electronic device designs for particular situations (heat, humidity, corrosive atmosphere, long-term unattended operation in any of the above environments, explosion-proofing, etc.).

>>1746696
Ingress Protection (aka IP ratings) may be the search term you're looking for

>>1746699
I think IP rating is mainly about enclosures, not components or functions.

>>1746748
not him but about the other things the dude asked, I think most of them are solved with engulfing the board in something or/and enclosures. Aside from radiation hardening and wider temperature operation ranges of military chips.

>>1746748
fair point. I second your request for Ott-tier bibles about electronic assemblies packaging

>>1746761
>electronic assemblies packaging
That may have been the keyphrase I needed to start digging at the library (and on libgen). Immediately found the following:
* Ali Jamnia, Practical Guide to the Packaging of Electronics: Thermal and Mechanical Design and Analysis (3d ed. 2016)
* Area Array Package Design: Techniques in High-Density Electronics (Ken Gilleo ed., 2004)
* Michael Pecht et al., Electronic Packaging Materials and their Properties (1999) (part of the CRC Press "Electronic Packaging" series)
* John H. Lau et al., Electronic Packaging: Design, Materials, Process, and Reliability (1998)
* High Temperature Electronics (F. Patrick McCluskey et al. eds. 1997)
* Gerald L. Ginsberg, Electronic Equipment Packaging Technology (1992)
* Donald P. Seraphim, Principles of Electronic Packaging (1989)

I'm gonna stop retrieving results there and start skimming. Might be some gems among those. Apparently quite a few international symposia on electronics packaging are held every year as well, with assloads of proceedings publications every damn year.

>>1746570
>assuming the leds are 12v
>no current limiting
Single-chip (i.e. common) white LEDs are usually about 3V each, due to 3eV being the energy of each photon they emit, which is in the ballpark of a 450nm blue that's used to excite the phosphor. 2V > 2eV = 650nm red LED. 1.5V = 1.5eV = 1000nm IR LED. Not like they'd bother with a flyback converter if they had multi-chip LEDs in there, they'd just have a little linear current regulator to drop a dozen volts or so.

>>1746570
Led drivers have a transformer not only for isolation, but for control sake. If you wanted to scale 20V using a buck converter or a non isolated topology for example, you'd end up controlling things in very narrow duty cycle range, the transformer adds a 1/n or n scaling factor to it, can be used as a "constant current source" with some power electronic magic and other stuff.

Alright lads, after some skimming of the sources found in >>1746789, I found some pretty based-looking references mentioned in (Jamnia, 2016).

* Patrick D. T. O'Connor and Andre Kleyner, Practical Reliability Engineering (5th ed. 2012)
* Eugene R. Hnatek, Practical Reliability of Electronic Equipment and Products (2003).
* Military Handbook: Electronic Reliability Design Handbook, MIL-HDBK-338B (1998)

The last one is nearly 1100 pages long and frankly looks like a one-stop shop, even if the info and methods discussed are undoubtedly not cutting-edge.

>>1744746
Live and learn (laughing and loving optional).
Zero-crossing triacs are useful, you'll find something to do with them eventually.
Keep in mind, some of these things sold as arduino-compatible triac "dimmer modules" actually have those zero-crossing triacs in them, too- and so won't actually function as a dimmer, just a switch.
So congratulations, you now know more about triacs than many Chinese resellers!
Always Read Datasheets

I have a devive that runs on 4.75~5.25v. My power source however is 4.74v. Does that 0.01v matter?

>>1746899
>Keep in mind, some of these things sold as arduino-compatible triac "dimmer modules" actually have those zero-crossing triacs in them,
but aren't triacs without zero crossing detection cheaper? wouldn't it make more sense to scam the people by using random phase triacs instead and claiming they are zero crossing ones?

>>1747014

ahahaha, no. the 4.75 and 5.25 ratings are guarantees. real devices will work outside of those guarantees.

>>1747014
What the other guy says is probably correct, but it sounds like a bit of a shitty power supply if it can only put out 4.74V. Does it droop significantly under load? What's the ripple like?

Anyone got good literature on sonar imaging and the electronics/signal processing involved?

>>1747061
Not really, but I know time-digital converters are used in some sonic applications.
As far as imaging goes, if it's anything like an X-ray, you can use spatial fourier transforms in order to reconstruct an image from a series of sections, which is something I learnt as part as an electronics+imaging paper that was fairly interesting.

>>1747014
>>1747042
>shitty power supply
it's almost certainly a "charger" made with 5% tolerance resistors throughout because who cares when you're just gonna buck it all down to 4.1V for the charger or less than that for the system power bus

>>1747081
Yeah I know all that. I'm just looking for sources that show specifically how it's done.

>>1747014

Use slightly thicker and/or shorter wire to beat the .01v voltage drop.

How does one create a bidirectional current buffer?
I'm creating a modular a/v switch which is constructed from block units. Each block unit has two video inputs on it, controlled by a pair of SR latches. In order to make sure that only one video input is enabled at a time, the reset line of every latch (across all block units) is tied together.
But if you were to connect a really large amount of blocks together, you would need a really strongly driven logic signal to reset every latch. So is there a bidirectional current buffer I can implement on each latch to prevent this issue?

>>1747147
sounds complicated. why don't you just use voltage-mode on the control bus to locally switch a current

I have a 12V horn rated for 50W
I measures its resistance and it's about 2 Ohms.
But when I measured the current it only drew around 3A.
How can this be? The resistance should be higher I think

>>1747165
>around 3A
yes, that's 12V/2Ω = 6A at a 50% duty cycle because it's probably a mechanical switch based oscillator = 3A

>>1747168
That would make sense
But how did the 50W rating came to be?

Are there good "grab bag" or lot deals on electronic components online I should pick up as a beginner? I keep seeing this 1390pc kit, and "1lbs grab bags" etc. On eBay, but I don't want to get something that has a disproportionate amount of useless crap and still have to buy tons of other components anyway.

>>1747171
approximate rms power, maybe

>>1747183
that's the point of grab bags, zoomer
>as a beginner
if you want an assortment of values of a particular flavor of component, look for an assortment or a kit. first find out what you want to make, then buy to fit needs plus a few extra

this circuit says it requires +12 volts and -6 volts.
If I understand virtual grounds correctly, I could instead replace the +12 volt rail with, say, a +24 volt rail, then have the grounds of all of the components connect to a +12 volt rail. Then, all of the components originally connected at -6 could be connected to a rail at +6 volts.

Then, the +12 volt rail would be the virtual ground, and the circuit would function correctly?

>>1747164
How does voltage mode work with logic signalling? Wouldn't typical TTL be considered voltage signalling? Clearly there's a concept I don't understand here, is there a place I can read about this?

>>1747206
I'm not a pedant, but I would consider it voltage signaling. You need a certain voltage to "turn on" bipolar devices. TTL also requires a certain amount of current as well... much more so than a competing technology that uses FETs such as CMOS... Comparatively speaking CMOS would be considerably more "voltage signaling"

>>1747171
a)
>it is wrong
b)
>how have you measured current/voltage?
Ideally you'd use a osciloscope with a current probe/resistor. multimeters have several ways of calculating a AC voltage and some of them do not behave well for non sinusoids.

>>1747147

if you need to drive a whole lot of inputs you can just parallel several gates to increase the fan-out. like in the example pic, 3 NAND gates can have the outputs joined together to triple the fan-out, or they could be used individually, each one driving one third of the many inputs. typical values of fan-out are 10 for TTL, 50 for CMOS, broadly speaking.

>>1747203
I can't really see the rest of the circuit, so it's entirely possible that you'll need to sink and source current from your 12V rail, which could prove difficult if you're using an ordinary switcher. If it's just a linearly-buffered divider, go ahead, sounds fine.
Makes me wonder how difficult it would be to sink current through a boost or buck converter. Just another MOSFET and inductor (with freewheel) after the diode to ground? The feedback network would need to be able to handle it, but I don't think that would be too much of a stretch, just two triangle waves on either side of a middle rail, and a comparator for each one

>>1747359
Why not use inverters? They often have "inverting buffer" written in their name, after all, plus you lose less pins, which should be a prime concern if you have enough pins to drive to be worrying about fan-out in the first place. A single high-speed totem-pole would probably work fine too.

Is fan-out ever really an issue? Purely in terms of input and output impedances I can't ever see it being a problem. At high clock-speeds I suppose it could result in latency, or at the very least a less-than optimal rising or falling-edge profile.

>>1747445
>>1747359
I've considered inverters but I think that in my application they'd latch each other. Duplicating the output driver is not strictly going to work either, since the issue is that I have no idea how many video blocks are going to be connected and each one needs to drive all the others.
This might illustrate what I'm trying to do a bit more clearly. Pic related is my current design, which begins to falter if I have about four latch blocks connected.

>>1747359

Yeah, that looks fine. But you know, you could replace all that NAND crap with a single transistor or FET and get some vast fan-out numbers at very high speeds (depending on the transistor).

>>1747464
>>1747445
>>1747359
You know what? I'm explaining this all retarded. The whole project is better visualized as a solid-state replacement for those cool mechanical switches that you find on a standard tape player where only one button can be pressed at a time. The project is a bunch of identical circuit boards which snap together. Each circuit board has a pushbutton. Pressing the pushbutton will turn on the one circuit board and turn off all the other ones. The trouble is finding a way for each circuit board to amplify that reset signal so that the parasitic resistance of the bus never leads to a situation where any given circuit board can't reset all the other ones.
That makes it a lot more clear, right?

>>1747471
>better visualized as a solid-state replacement

Ahh.

What is the penalty for more than one being on at a time? If there will be loss of life, then you will probably need a clocked solution.

Why are you using TTL? I don't see this as being a problem with CMOS (which has very high input impedances). So, I'm assuming you have more than 10 of these things chained together.

If you use CMOS, you can even directly integrate a capacitive single-wire touch switch into the "flip flop" you've got there... that's how much these things amplify, you could reset/access the "bus" through multi-MΩ resistors and it will still work.

>>1747471
Oh neat, those things are expensive, as are a lot of switches and pots. If it can be replaced with shitty 5c tactile switches, I'm interested. Problem is all the memory is volatile, while something like a digital pot is designed with non-volatile memory. So if there were non-volatile D-FF ICs that would be really neat. Wait now I have to search Mouser.

>>1747480
Well, all I learnt is that you can get nonvolatile digital programmable variable capacitors.
Magnetic cores anyone?

So does anyone have good recomendations as to where I should look for rechargeable Li ion batteries? Should I just get any random replacement phone battery and design my circuit (charging circuit/protection & buck converter) around the battery of my choosing?
It's going to be used to power a BLE controller.
I've skimmed though the recommended Li+ Battery reading as per the OP, not that much applies to me right now outside of safety considerations (which is obvious when dealing with Li batteries).

>>1747493
Where to buy them? eBay and ali may well have cells that are actually made by Samsung or Sony or whoever, but it's near impossible to tell. I'd go for the more reputable hobby sites, the kind of place you'd go for those blue heat-shrink guys you see on RC copters and cars. I imagine they have 1S guys too, but they'll likely tend to be of the higher-current lower-capacity variant.

>>1747471
how concerned are you about someone pushing two buttons within nanoseconds?

Am i calculating this shit correctly?

If i want to power this octocoupler from mains, 230V rms,
https://www.tme.eu/Document/575892aa6df21bf2432c9b77664b6bc5/SFH628A-4.pdf

Then at about 1.3V the led draws about 5ma, so to power it from mains i need a
(230 - 1.3) / 0.005 = 45k resistor
which will be radiating 228.7 * 0.005 = 1.14W of heat
Is that correct? That seems like a lot of heat for just 5 miliamps, which would mean i can't use a normal blue resistor since those can do like 0.5W tops and get hot as fuck while doing it

>>1747638
that is correct, but why are you running it at 5mA when you don't need to? look at figure 3 on p4 of your PDF, it's specified down to 0.5mA. you need only make the receiver a bit more sensitive

>>1747644
So if am reading it correctly then at 20°C if the led is getting 0.5ma, then the max current through the tranny on the other side will be also 0.5ma since the ctr is around 100%?

>>1747645
correct. so if your output-side pullup voltage is 5V, your output-side pullup resistor should be (5 - 0.5) / 0.0005 = 9k if you want an output voltage of <=0.5V when the LED is illuminated

I have the ads1256 board which is a high precision and fast adc. 24bits at 30ksps. It runs at around 5v and I need to power this from USB which works, but I'm looking to eliminate noise. The board has a 2.5v reference ic on it.
My question is will the USB power be a source of noise. Or rather is the reference the only thing that matters.

>>1747658
>will the USB power be a source of noise
yes. the 2.5V reference is not perfect, look up its "power supply rejection ratio"
if you add some filtering at the USB port and the input of the board, and connect all grounds generously, you might be okay

>>1747664
I was afraid so. I have a few low noise ldo regulators. But how do I remove noise when I can't step down the voltage.

>>1747675
Just thinking. If I had an external 2.5v reference powered from a regulator. Would that remove noise even if the adc is USB powered.

>>1747675
>But how do I remove noise when I can't step down the voltage.
fC=1/(2*pi*sqrt(L*C))

Anyone here who has built guitar pedals? What did you make and what are some of the easiest ones to make?

>>1747773
Yes. In order of complexity
>fuzz face
>distortion
>tremolo
>(after this you need basic filter knowledge)
>phaser
>octave up

>>1747682
I feel like thats not going to work to clean up something with as broad a spectrum as a usb 5v line.

>>1747777
I kind of have filter knowledge, but only on the mathematical side (I had a signals & systems class). I would say I'm completely clueless when it comes to actual circuit filters, but it might be easier to learn.
Can you point me to a resource to learn that stuff and/or pedal specific resources you think are good?
What do you think of the Electra overdrive pedal (pic related) as a first build? I have all the components - except the 1N34A diode which I read I can swap out for another 1N4148 - ready and I'd like a warm OD in my chain.

>>1747773
built this one a while back. Decent tone. I think if you google lil fuzz you should be able to find the pdf guide

>>1747779
it'll help with the high-frequency stuff tho. you can get meaningful filtering at the low tens of kHz and still use reasonably priced components. if you really need super clean power, maybe run the system at somewhere less than 5V. maybe you don't even need to run at a "standard" voltage

>>1747790
Googled "Lil fuzz" and it brings up a rapper called lil' fizz so thanks for the kek. I found the pdf, will check it out. probably gonna build the electra first though (see >>1747781)

>>1747797
https://www.youtube.com/user/chromespherecom/playlists
this guy has bretty good videos and tips

What specs should I be looking for in a JFET? Got my eye on 2N5457, 2N5460, and 2N5485.

>>1747863
depends on what you want to use them for, of course. it's pretty aplication-dependent

>>1747863
> looking for? popularity.

All very popular:

2N5457 - this is like the 2n2222 of the FET world.
2N3819 - the 2N3904
2N5458
2N5459

>>1747890
Audio stuff, as variable resistors for AGC circuits, just general-purpose ones for audio frequencies. I see the 2N5460 is a 40V-max model (and is also a P-type), while the other two are 25V models. I imagine a ±12V supply is about all I'll ever use, but as far as biasing them, I need to pull the gate to -5V or so to get them to turn off, which might effect what power supplies I'm able to use. The 2N5485's datasheet says it's an RF transistor, and I already have some RF JFETs from my mini-whip project, but what spec determines how good a JFET is at higher frequencies? The capacitances are pretty similar, and there's nothing like a gain-bandwidth product to look for.

>>1747952
I'll look over those datasheets, thanks.

Final questions, is there any reason to go for complimentary JFETs? Are N-chans still better to go for than P-chans, even though they're technically depletion-mode?

>>1747967
>whatabout complimentary JFETs, P-channel?

Rare. We don't see much of that in designs.

>>1747973
So no JFET totem poles? H-bridges? Double-ended current-mirrors?

>>1747974

Of course all these things exist (like CMOS which essentially a kind of FET and use complementary devices on one substrate...there is many quadrillions of those around) but I haven't seen much of that using the kind of devices under consideration.

Feel free to throw out some discrete FET part numbers....

>>1747556
Not at all. If someone is trying that hard to break the device, they can have it.
The design you posted is extremely interesting to me. I'm going to look into this further.

I have resistor in a breadboard and i connect mains cable to this resistor and then turn off the outlet with a switch so there is no power passing and when i touch the resistor with my finger i hear buzzing noise in my headphones that aren't connected to the breadbaord at all, what the fuck?

working with 230V on a chink breadboard that isn't rated for mains is scary as fuck
is it on? is it off? nobody knows
is the red wire 5V DC or 230V AC? if i stop posting you will know the answer

>>1748087
>not placing a neon lamp in the circuit so that you know for sure
poorfag detected

>>1748087
yay it's working
fartduino is correctly counting all zero crossings in the 50hz mains AC.. well mostly correctly, i wonder if that 1% error is due to fartduino or due to shitty power plant

>>1748137
arduino, almost definitely, lern2interrupts

>>1748138
Son, I have been using interrupts when you still thought fortnite was cool

>>1748139
nigger, I was counting NOPs and writing demo screens when you thought INT 18h was cool
anyway def a coding problem. printf is slow esp on 8-bit processors, maybe you're missing interrupts in the main loop because of that. also chinkduinos shouldn't be counted upon to be built with the best quality parts. try adding a buffer variable

>>1748087
> AC mains on breadboard

better idea: you know you can just tap off a tiny 230 to 5v transformer and still detect the zero crossings right? Maybe even your arduino is powered off such a thing, but before it's rectified and smoothed.

>>1748088
> neon lamp
another better idea: the neon actually flashes with each half cycle, so you could pick that up with a photodiode (or even a solar cell)

>>1748155
>230V to 5V transformer
>tiny
unlikely and expensive
>read the neon lamp
>better
the point of anon's project is to find the zero crossing, not to merely count cycles. the hystersis of a neon lamp would be undesirable

>>1747556
This design is unnecessarily complicated. You only need one of those 10k resistors pulling to VCC. Of the diode pairs only the one on the right is necessary. Only one schmitt inverter can run to every clock. Inverter isn't even neccesairy if you put the LEDs on Q instead of !Q.

THIS WAY PLEASE
>>1748210
>>1748210
>>1748210
>>1748210

>>1747781
I had a teacher in EE who was a musician and got into EE school because of music and he gave a very good filter/pedal building class. I'm afraid his books are in portuguese :(. That aussie sound site is a good start.
https://sound-au.com/articles